Excerpts - Scope-drive list

Home Last Updated - 9/9/06 ben at davies . net

Excerpts from the Scope-drive List

Go to Page 2 Jump to Troubleshooting

Alignment Altitude Offset Analysis ASCOM Asteroids AutoGuiding Axes Backlash

Balance Browser Circuit Cold Weather Comets/Satelites Computers Connectors Coordinates

Current Limiting Resistors Current Compensation .dat Files (+ others) Drift DOS Encoders Equatorial Platform

Error Correction Flywheels Focuser- FieldDerotator Gearing Gear Ratio (calculate fullstep size) GEM Flip

Guiding Go To Halfstep Handpad Handpad, RF Home Position Hot Keys Hour Angle

Initialize Invert Output KeepAlive Circuit Limits Longitude Latitude LX 200 Mechanical

This site is a member of WebRing. To browse visit here.

Link to latest version of Scope.exe

ALIGNMENT: Alt-Az: Initialization Equatorial: Polar Alignment

The inits are for ALT-AZ style mounts. If an equatorial telescope is well polar aligned, all you have to do is point to one star, enter its equatorial coordinates and reset to those coordinates (push r on the keyboard). Scope can then figure out the relationship between alt-az and equatorial coordinates because one axis is already known. If you are not well polar aligned, the inits can help with local pointing accuracy. If you are, they will probably make pointing worse.

The terminology can be confusing here because any telescope's pointing coordinates are alt-az. So we are using the term alt-az in two different situations. For an Alt-Az mount, the telescope's Alt-Az coordinate system is coincident (hopefully) with the altaz horizon coordinate system. For an equatorial it is not. The equatorial mounts' Alt-Az coordinate system has the Azimuth plane coincident with the plane of the celestial equator. Think of an alt-az mount tipped over to an angle above the horizon equal to he latitude, so that the azimuth (Right Ascension) axis is pointing at the north (south) pole. 0 and 360 degrees azimuth are at the meridian. The altitude then is measured from 0 at the equator to 90 at the pole. Scope.exe must convert back and forth between sky (equatorial) coordinates and telescope (alt-az) coordinates. With an ALT-AZ mount, Scope establishes the relationship using the 2 star initialization. With a well aligned equatorial the relationship is established by pointing at one star and doing a 'reset to equatorial'. Ben Davies

+++

...if you are going to do something like initializations that requires a high power crosshair eyepiece, you need accurate coordinates to go with it. The surest is to use Guide.exe to send the coordinates (see Lx200). Scope's data files (Bstars, etc) are now corrected for every thing except proper motion. If you stay away from high proper motion stars like Arcturus, these coordinates are also very accurate.

ALTITUDE OFFSET:

1) Q. Can someone explain what the altitude offset does and why it is important to do?

A. Dave Sopchak first suggested to me that we could make the altitude reading as 'unknown' at startup time as the azimuth reading. As you know, the starting azimuth is irrelevant as it is the difference between successive azimuth positions that counts. But the altitude must be known precisely, or at least had to be known precisely until the alti

tude offset was developed.

The altitude offset uses the fact that the angular separation between two equatorial coordinate sets and two altazimuth sets must be the same. So I iterate until the altitude readings of the two positions in question yields the proper angular separation. The difference between the derived starting altitude and the claimed starting altitude is the altitude offset.

You should use the altitude offset when you are uncertain of your starting altitude (you no longer have to level your tube, for instance). Do not use it if you are certain of your altitude. Mel Bartels

+++

2) To understand the altitude offset, you should understand the azimuth offset first. Take's coordinate conversion routine only needs the difference in azimuth between positions, not the absolute azimuth position. So, an azimuth offset can be calculated based on the arbitrary starting point where azimuth is set to zero. That is, after completing the coordinate conversion initializations, the true azimuth can be determined of the starting point. This value is the azimuth offset. Mel Bartels

+++

Q. On my Alt/Az scope, when I do inits, I often forget to do an Altitude Offset. This results in good tracking, but terrible goto's. Since I believe its the same bit of code for both Alt/Az and EQ mounts, does this mean if you have an EQ mount, and you do an Init 1 & 2, that you will need to do an Altitiude Offset as well James Lerch

A. Less necessary for equatorial mounts as the simple 'reset to equat' usually guarantees accurate altitude, ie, declination... Mel Bartels

+++

Altitude offset is the error in altitude/declination. If you point your at the local horizon, an altitude of zero, but tell the software you are pointed halfway up the sky, an altitude of 45, there will be an altitude error of 45. Since it can be challenging to get altitude exactly correct, several of us a few years ago worked out an altitude offset calculation that derives this error. It works a little better if the 2 stars for inits are about at the same altitude.

But that's not necessarily what best for 2 most accurate inits overall. Calculation and accurate altitude offset does not necessarily show up in the latitude measurement. Do not make a tight coupling between these two. Altitude and latitude are two different beasts. The 2 star init may reach an apparent accurate latitude but then show a large altitude offset error. That means that the scope will point accurately in some parts of the sky but not in others. For most accurate results, the altitude offset should be very close to zero and the latitude error should be very close to zero.

How close to zero depends on how accurately you wish to pursue the Z123 analyses. For a practical matter, I'd recommend getting latitude to within 0.1 degree. Getting it to an arcminute or so makes a big improvement. Mel Bartels

+++

I do an alt-offset after all is supposed to be wonderful, just to verify the inits. The 'test' offset should be very close to zero Mel Bartels

+++

The altitude offset is typically done once each evening, as soon as the two stars are initialized. It is best to use two stars at about the same altitude with widely differing azimuths. Mel Bartels

+++

(The) Optimum conditions for determining altitude offset not necessarily optimum for overall initialization accuracy! Mel Bartels

+++

The declination in a perfectly aligned equatorial will equal the altitude reading. Mel Bartels

+++

You can look for optical vs mechanical mis-alignments by running a series of 2 star inits, and after each init 1 and init 2, do the 'a' altitude offset option. Do NOT adjust the offset - just jot down the value. If the value is fairly consistent as you re-do both inits, then this tells you that you like optical:mechanical misalignment. The alt-offset will fix one direction of this. The other is more complicated. If you see consistent alt-offset values, then post back to scope drive list and we can talk more about the perpendicular component. Mel Bartels

ANALYSIS FILE:

Analysis would help of course, but make sure you solved all random, mechanical (imbalance) and backlash errors in the system first. Otherwise you will basically analyze random errors of the system. Your flexible rubber shaft couplings would be one source of random error for example. Also make sure your scope is collimated accurately before attempting analysis. That is one important source of pointing errors that is often overlooked. Berthold Hamburger

+++

There is a handpad mode called analysis. After the object is centered, push the mode button momentarily to record the position. What gets recorded in the analysis.dat file is the actual altitude and azimuth along with the object's equatorial coordinates, and the sidereal time. These values are later used to construct an error graph, which is essentially the difference between the recorded alt-az and what the alt-az really should have been. This will work regardless of mount type, but the type and degree of error will likely be different.

Z1 and Z2 and Z3 will continue to function regardless of mount type. All mount types will have alt or dec, and az or ra, offsets, and will have axis misalignment.

If you do have an equat and you are not precisely polar aligned, it will show as a relatively smooth sinusoidal curve on the error plots. This can now be fairly compensated for by the new pointing model corrections (PMC), which is under active development to say the least! Mel 136

+++

To create a series of analysis points. Put the handpad in analysis mode, goto an object from a data file, then precisely center the object using crosshair eyepiece. Record the analysis point by using the handpad's left or right button mode keys. This records the target position, and the error from that target position. Now, pull up the Z1Z2 analysis via menu option. You can play with different values to minimize the rms error. Try to stay on one side of the meridian flip - don't add analysis points from the other side - to do so is not sufficiently tested.

Since you will likely be doing the analysis multiple times, you can build a scroll file to move to an object, record the analysis point, and move on. See the webpages on the handpad and scroll files for details. Mel Bartels 10139

+++

input the coordinates of the target object via a data file (hopefully a star to be pinpoint about it, and use a crosshair eyepiece at high power), then move to the star and center the star on the crosshairs, and record the analysis point with either a left or right mode button press. Input the coordinates of the next target star, and repeat the move and centering. After you've recorded one to many stars, you can go into the menu option to display the analysis, and play with adjusting the Z1 and Z2 errors.

Z1 Z2 Z3 errors are defined as: axis misalignment (one side of the rocker is higher than the other) (called Z1 or axis misalignment) discrepancy between the optical and mechanical axis in the horizontal axis (called Z2 or azimuth offset) discrepancy between the optical and mechanical axis in the vertical axis (called Z3 or altitude offset)

Do not reset coordinates (reset equatorial)- it ruins the analysis points because the original alignment used to generate the analysis points is destroyed. Mel Bartels 6711

+++

Remember you can analyze for Z1 Z2 Z3 errors by doing 3 inits and also by doing the analysis file. Mel Bartels

+++

Q. A question regarding building the analysis.dat file. Do you build it while tracking the scope or should tracking be disabled?

A. I recommend leaving the scope in track mode as this will make for the most accurate handpad button press when the object is exactly centered. Mel 100

+++

Q. ...could explain the procedure to fix up Z3 for an EQ and fixed mount (not a GEM)? Is there a different method between EQ mount and alt-az mount ?

A. No difference. Take a series of analysis points varying one axis while keeping the other rather constant. Then use the plotting function to see if you can find a Z3 that minimizes the rms pointing error. Mel Bartels

+++

Before you do analysis of Z1/2/3 and other pointing corrections, make sure you have slippage, backlash, encoder resets, etc. under control so that you have a system that moves very repeatable...can you slew between your 2 init stars a couple times and have errors of a few arcminutes or less? If not, then try and troubleshoot the drive/gearing/encoders/etc. Also, are you sure your fullstep sizes are accurately determined/defined in config.dat?

If everything is pretty tight, then try this:

- 2 star init (I've gotten into the habit of my init's being in the same two parts of the sky...helps with repeatability later on) - alt offset/re-init

- check derived latitude versus actual latitude...if within a few arcminutes you are now good to go (these days I tend to only be off 1 arcminute in latitude after a 2 star init...but until you get all your pointing analysis complete, you probably won't get that accuracy unless your mount is very well built.)

- start the analysis mode and point due south - analyze stars at/near 180 az, at elevations of 20, 25, 30, 35.....85 degrees

- Then look at the pointing error data graphically. Is there an azimuth error as a function of elevation? If so then manually tweak Z2 and get the graph of az errors vs. elevation to stand straight up, not be skewed.

- Then start over again, but this time do an "elevation analysis" at several azimuths, such as 110 degrees, 250, 320, etc (analyzing points at 20, 25, 30, 35.....85 degrees elevation at each azimuth)..then graph again...did you get straight az vs. el error bars at all azimuths?

I don't know how it is in other scopes, but once I nailed Z2...my pointing accuracy was pretty good...and then I had to make corrections for my potato chip shaped base (did an analysis run at constant elevation, 40 deg, at every 10 deg. of azimuth), and corrections for my slightly eccentrically mounted elevation worm gear (by analyzing alt errors vs. altitude.)

Other hints: Make a couple scroll files to help in this massive analysis effort...it saves time and keystrokes. Also, use a large star database like 0000XL.DAT that has thousands of stars in it, and use the option to find the closest object to current scope pointing...the scope seldom has to move more than a degree in either axis to find a desired star. Tom Krajci 1593

+++

Z1 and Z2 are intertwined so that you need a sufficient set of measurements to distinguish the two. Z2 is just like the Z3 error but at 90 degrees or the horizontal difference between optical and mechanical axis, whereas Z1 is the warpage between the altitude/azimuth axes or the declination/rightascension axes. Typically you will have almost zero Z1 error, and the Z2 and Z3 errors will be a few arcminutes or tenths of a degree. Z2 and Z3 can be fixed by collimating your scope so that the optical and mechanical axes coincide or by shimming the tube a tenth of an inch or so. The Z1 error can be fixed easier in some mounts by shimming the bearing locations on the declination plate.

I consider Z3 separately as it is a constant that can be added/subtracted and does not influence the other error correction values. Z3 is the altitude offset in a dob, or the declination offset in an equatorial mount, often pictured as the difference in angle between the mechanical axis of the tube and the optical axis of the tube. Obviously the mechanical axis of the tube is its idealized axis, and may not coincide with however the saddle and tube is actually built. Z3 is the compensation factor to bring your axis into alignment so that if your optical axis points too high then Z3 is negative. Mel Bartels

+++

The GEM flip messes up the analysis so do all analysis points for the time being on one side of the meridian. Mel Bartels

+++

For the analysis, do init 1 and init 2. You should do them manually if you have an equatorial mount, unless you are precisely aligned on the pole to 1 arcminute or better. Do not do init 3. For equatorial mounts, the Z1 parameter is probably most important. Mel Bartels

+++

Next, you may want to spend a clear, full moon evening testing 2 star alignments (with all mount correction features switched off). Make every elevation at 30 degrees, but try a pair of stars at 180 and 270 azimuth, then try a pair at 200 and 290, then at 220 and 310 etc.....each time, pressing the ' I ' key to see how close the derived lat/long is from your actual observing site. You may find one particular pair of azimuths gives you the best agreement. Stick with that pair from now on. (My pair is elev. 30 az 200 and elev. 30. az 320) (All the while, occasionally check your PEC zero marks on your steppers versus what the software is keeping track of.) And, always set the mount up in the same way...whichever leg is pointing due south, etc....keep it that way every time. Also, make sure the encoder inputs are being ignored.

After you find your 'sweet spot' for 2 star alignment you are ready to perform detailed mount pointing analysis. Point due south and take mount analysis measurements of stars at elevations from 20 to 75 degrees, every 10-20 degrees or so. (Don't let the encoders reset/override the software...disable/ignore them.) You will learn 2 things here. You will check if your gear ratio is properly calculated for your elevation gear (elevation error as elevation changes), and you'll see if there is any Z2 error (error of azimuth as elevation changes). You can then adjust the Z2 value to straighten out the error plot of azimuth versus elevation. And if you need to, recalculate the gear ratio and put it in CONFIG.DAT

Then make a series of slews around the sky (at the same elevation, such as 30 degrees, perhaps lower) every 10-30 degrees and record the pointing errors. This will show two errors. It will show if the rocker box and ground board are 'potato chip' shaped (which shows as an elevation error as azimuth changes), and the second error is if your azimuth gear is not perfectly centered (which shows as an azimuth error as azimuth changes). If your errors are repeatable, then you will make big improvements to pointing accuracy with this approach, but it may take a couple hours of clear sky time to measure all this stuff. Tom Krajci #6074

+++

Before you undertake mount error analysis...make sure you motors don't stall or lose steps, and that you've got slippage (if you use a roller drive) under control. If not...your analysis will be inconsistent and imprecise.

I recommend you be consistent with your 2 star init at the beginning of the observing run...always use the same azimuth and elevation for the two stars. (In my case I go to AZ 200 EL30 for one star...and AZ 320 EL 30 for the second star...using the nearest 6th mag. star from 00000000.DAT...which is within a degree or so....and I always set the same foot of my ground board pointing due South, and level the base. Only then am I ready to do good mount analysis.

After my two star init I call up a scroll file to move to azimuth 180 and make analysis at stars of various elevations...then I plot the data and tweak Z2 manually to eliminate the error. (At this point I also save the elevation versus elevation error data for use in the appropriate .DAT file for mount correction.) Then I run a scroll file to analyze pointing at elevation of 30 degrees, but at varying azimuths around the sky. I then plot this data and see if I need to tweak any of the Z values. If not, then I merely record this analysis into the .DAT file for elevation errors that change with changes in azimuth. (This corrects for my potato chip shaped rocker box bottom.) Tom Krajci 6723

+++

This http://www.asahi-net.or.jp/~zs3t-tk/ is a link to Taki's site describing mount pointing calculations and correction for construction errors Chris Roland

AUTOGUIDING: http://www.bbastrodesigns.com/autoguide.html

+++

I did a setup assembling that works great both, for stepper system as well as for Scope II. The idea is not to be dependent on a specific 
camera or with a specific mount electronics but to be based on a standard protocol. The way to do that is to use LX200 based commands for 
autoguiding and a ASCOM driver. Take a look on the following web (http://www.technoplus.nl/astro/lx200.htm). Albert Van Duin is selling a 
LX200 emulator that connects it to a computer, accepting LX200 commands and driving four relays.  
My setup is as follows: The autoguider is based on a Philips TouCamII modified for long exposure and a monochrome CCD sensor. I'm currenly 
usign GuideDog, but it is also possible to use MaximDL, K3CCDTools, ......I have installed an ASCOM driver and I tell to the software that I 
have a generic LX200 telescope. The computer is linked to the LX200 emulator with a serial cable and the LX200 emulator relays are linked to 
the ScopeII handpad buttons.  Roger Artigas

+++

Q. Is an auto guider necessary if the GOTO system is polar-aligned and calibrated accurately?

A. You can do short exposures up to 1-2 minutes with no guiding, if you have a smooth and accurate mount and drive system. Beyond that you need some guiding. You can do offset guiding where you pick off a portion of the edge of the field of view to find a star to guide on. You don't need a separate guide scope. If you do use a refractor for a guide scope, it can be very cheap. Popular are 2-3 inch refractors bought used for a few dollars. Quality of star image does not matter.

+++

Scope.exe can accept guiding input from either a relay box that simulates pressing handpad direction buttons, or from lx200 commands send on a serial link. The guiding input works on all types of mounts including altaz and equatorial. What you have to do is to rotate the guiding CCD so that the outputs coincide with the correct motor direction. Mel Bartels

+++

Q Can two handpads be connected in parallel?

A. Yes, as the pullup to +5vdc can be triggered by either handpad or relay operated autoguider. Mel Bartels

....use a 6 pin modular Y connector in the handpad plug on the circuit board to provide two handpad ports. Be careful, you want a connector that is wired straight through. The type for normal telephone use crosses pins and will give you nothing but grief. Get one labled for "network or data" use. They are available from Jameco www.jameco.com for less than a dollar, though they may have a minimum order. Marty Niemi

+++

For autoguiding, the system must be moded to �Guide+Stay� mode BEFORE the autoguider commands start being sent. HAS

+++

Q. autoguide-software which is directly compatible with scope.exe. Does anything exists? Is it free?

A. try free guider.exe (if you have a b&w quickcam and a cheapo DOS laptop), and see:

http://www.ghg.net/cshaw/quickcam.htm http://www.ameritech.net/users/mniemi000/auto.html Paul Shankland

A. I too have tried using all of the programs you mentioned (K3CCD Tools, GuideDog, Startrack) but found they all connected ok to scope but motors didn't move. I gave that up a year ago and tried Guider by Marty Niemi, this worked (made for a B&W webcam) but wasn't compatible with the way I wanted my computers setup, so I went for AstroArt which works fine.

A You could try Teleauto program... it's free and is very very good! You'll have no problems using it with a ToUcam. Rich Bowden

A i'm using http://www.technoplus.nl/astro/lx200.htm It does really great job ! I have my best results with this autoguider in paarallel with the button on the handpad. i have sometimes problem with autoguiding software crashing with the RS232 commands, but never with this emulator. Vincent Steinmetz

+++

Q Will autoguiding make PEC obsolete? (Or will an autoguider corrects for PE?)

A I agree with Chuck Shaw and other seasoned imagers that PEC helps the autoguider. However, other advanced imagers disagree. Of course, the amount and rate of change of PE probably factors into people's experiences. In other words, how you feel about this may depend on your mount! Mel Bartels

+++

I routinely do what you are talking about using either my Greyscale Quickcam or my CB245. Check my website at: http://www.ghg.net/cshaw/14newt.htm I have changed the guidescope rotator to one slung on the side of one of the altitude trunnions, but have not posted any pictures of it yet.

You will need to build the autoguider interface, which is documented in the Cookbook manual. I have an updated version of the circuit I can send you to properly interface it to Mel's circuit. Basically it just looks like another handpadle to Mel's circuit.

I will be writing up a presentation for the Image-The-Sky conference in Oregon later this year , and will also have it on my website, to capture the details of how I do this. However, the essentials are that I use guide+save mode, and once in mode, I put the autoguider into autotrack mode. When the mode ends I immediately kick the autoguider out of guiding since scope.exe will be in ms mode at that point. The you simply use the menu driven capabilities in the EC menu to extract the PEC data from the guide files into preliminary pec files that another menu command can average together to build a new PEC.dat file. Sounds more complicated than it is......

I use CBwinCam (free windows based program written by Veikko Kanto, available on the WillBell website) for CB245 acquisition for autoguiding with my CB245 (Its OUTSTANDING!!). I use Guider.exe (freeware by Marty Niemi) for autoguiding with my Quickcam. Both cameras use the Cookbook autoguider interface circuitry.

For just trimming out drift, you do not need to autoguide to get it accurate. Simply get into guide mode (not guide+save) and maybe sure that drift update is enabled (on) via the handpaddle menu. Center the star on a crosshair, turn on guide mode, and periodically recenter it. After a couple of minutes, exit the mode and the drift will be calculated and updated. Chuck Shaw

+++

Q. I was searching the web for info on building your own autoguider. Using a quick cam with cookbook software???

A. Do you have the old B&W Quickcam? If so, take a look at my web page link in the Sig. I have a program that uses the Cookbook style interface to control a mount. I can't really add much for any other camera besides this one. Martin Niemi

http://www.ameritech.net/users/mniemi000/index.html

+++

Q. is anyone using Guider.exe with a Philips ToUcam?

A. Guider was written using low level hardware control commands for only one camera, the old B&W Connectix Quickcam.  I would try using some of the autoguider programs such as "guidedog" that use the video for Windows interface to talk to the camera.  I believe that would work with your ToUcam.
Marty Niemi

+++

Q...Can anyone tell me where the information is in the files section on how to hook up an STV to Mel's system?

A. The STV is easy to connect as it uses it's own internal relays like the ST-4. There is a schematic in the files, circuits area the shows how to hook up a ST-7,8,.... You do not have to go to the trouble of using that circuit which has it's own relays. The easy way to connect the STV is to use a 6 pin modular Y connector in the handpad plug on the circuit board to provide two handpad ports. Be careful, you want a connector that is wired straight through. The type for normal telephone use crosses pins and will give you nothing but grief. Get one labled for "network or data" use. They are available from Jameco www.jameco.com for less than a dollar, though they may have a minimum order.

Use the NO (Normally Open) pins from the STV relays. The point here is that you are trying to provide the same connections that the 4 direction handpad buttons do. This means that you are switching +5 volts as the common connection to three different parallel port pins. You need to use two small signal diodes such as a 1N914 (easily found at Radio Shack or the like) to provide the connections to activate two of the pins at once. The schematic on Mel's web pages shows this very well. I would not trust the color codes shown on the schematic in the file area. There are two many variables involved such as the way the cables were assembled. Be sure that you have it right by ohming things out or testing with a continuity tester. Marty Niemi

+++

I look at it like this: Guiding is readied per the handpad being put into guide mode, then buttons are activated to cause guiding commands when the handpad left mode switch activates the actual guiding. Now you can have drift update on or turn it off -- I usually like it on especially with autoguiding as it makes autoguiding a little easier, though some feel that a slight drag to the scope motion improves autoguiding by forcing all corrections to be in a single consistent direction - hence yet another feature called 'drag' that you can turn on and specify the amount of also. If you wish, you can record the guiding button presses for generating PE curves -- that's where the pec synchronization becomes critical as the pec cycle has to start with the same motor shaft or gearing (or whatever you are doing the PE for oriented consistently). So PEC has to be turned on and you have to get the motor shaft oriented consistently for the generation of PEC to make sense. That's why the fullsteps per PEC and so on have to be entered in config.dat; so that the software knows how many motor rotations will comprise a PEC (typically one, but if you are trying to get PEC for a 4:1 gear reducer and the output of the gear reducer is most important to you, then the answer will be 4x the # of fullsteps per motor revolution) Mel Bartels

+++

Q. I'm trying to get "Scope" application configured so that, using the LX200 protocol, I can use the autoguide function with "Astro-Snap". So far I can get it to center, track and move... but not autoguide. I guess that "Scope" needs to be configured somehow to have it do so,...

A. Put handpad in guide mode, then initiate guiding by pressing the left mode key. Adjust guiding speed as necessary via menu option or config.dat file. The reason for the additional left mode key press is convenience at the eyepiece. You can turn on/off actual guiding by using the left/right mode keys while the handpad stays in guide mode. This means that the 4 direction buttons toggle from coarse centering when in altaz mode (guide=off) to guiding control (guide=on) depending on your use of the left/right mode buttons. I find this a real convenience at the eyepiece. Mel Bartels

+++

Marty Niemi and I have been comparing notes, and one of the things that we ended up wondering if it was possible to do is to have scope.exe automatically enter Guide+stay mode when it starts receiving autoguider commands?

We both admitted we have had panic moments after carefully framing the target in the CCD FOV, when we take our autoguiders to "track" mode, and have forgotten to put scope.exe into guide+stay mode 1st! Scope.exe will dutifully start slewing around like crazy at ms speed as the autoguider goes nuts sending commands at a rate that assumes the scope is responding at gs speed. Kinda humorous after you get over the panic and foolish feeling!!! <grin>

Your change to have scope.exe STAY in guide+save mode but simply stop recording after 3 pec cycles has been GREAT!!!! I would get the same type of runaway commanded slewing madness when I would not be quick enough to kick the autoguider out of track mode when scope.exe would exit back out of guide+save mode before! Thanks again!!!!!!!!! Chuck Shaw

....Chuck and I were comparing the number of times we have fired up the autoguider without changing to guide rate first. Guiding via LX200 serial commands works great as the rate is usually set first by the controlling program. Marty Niemi

+++

Are you sure the guider is rotated properly to correspond with movements in the X, Y, axis? If it is not it will cause some strange things to happen. If you are guiding only in one axis, RA, and are not polar aligned, you will see drift occuring that is not corrected for. Generally, if you are guiding in both axis and are off in polar alignment and the guider is keeping the guide star centered, what you will see is a gradual rotation of the image with the stars at the edges becoming elongated during long integration times. If the guide scope is not really, really rigid in relation to the imaging scope, flexure between the two will cause this drift to occur as the guide scope keeps the star centered and the imaging scope flexes as it moves in Elevation both in Ra and Dec. Also, as the telescope moves and the orientation of the mirror relative to the force of gravity changes, if the mirror cell is flexing (spring loaded collimation screws are good for this) the result will be what appears to be drift. So, see if you can determine if the guider is keeping the guide star centered..if it is..then look at the mechanical aspects of the mount/guider configuration as being the probable culprit. Bob Kirschenmann

+++

Q. have never been able to get good results with auto-guiding.

A. Most often autoguiding problems come down to the delay factor in the autoguiding program. Scope.exe responds instantly to guiding commands so you need to set a very low value in the autoguiding software for the delay time. Mel Bartels

A. I have used Marty Niemi's Guider.exe with the stepper version of scope.exe for years, and it has worked REALLY well for me. I never used the LX200 interface. I always used the hardline interface (i.e. it looks like a 2nd handpad to the PCB).

Do not use backlash when autoguiding. In theory it should work fine, but in practice its hard to get it to not over correct. You don't need it in RA anyway... And in Dec, if your Polar alignment is not perfect you should be getting a tiny dec drift so you should not need it there either. You want to run the calibration routine twice. The first time backlash in the system may fool it, but the second time it will work well.

After your calibrations, get the guiding window as small as possible to allow faster downloads. Do not use a star that saturates the detector. If the star is bright, use a short integration and use the averaging routine to average several shots. For bright guidestars I usually shoot at 0.5 sec integrations, and average 3-4, and have the trigger set at one less than the number of images I am averaging to allow a little "persistance".....

Most poor guiding is due to chasing the seeing. In poor seeing use a dimmer star to allow a longer integration if at all possible. The longer integration smoothes out the seeing a lot. Once you run the calibration, start messing with the aggressiveness, dialing it down till you are not getting commands that bounce back and forth. I usually have the aggressiveness set at about -3 or so..... Lots of tiny short corrections... but let your mount tell you how often you need to send a guiding correction.... An old rule of thumb is to send a guiding correction no more often than about every 1/10th of the integration time your mount can image unguided. If you can shoot 30 sec unguided, then don't send corrections more often than every 3 to 4 seconds. That allows long integrations and no seeing chasing. If all you can go is 15 sec unguided, then 1.5 to 2 sec. Rules of thumb are just that however.... You have to let your mount tell you how often it needs correcting, but 90% of guiding errors are over controlling (i.e. aggressiveness is too high) or chasing the seeing. The fix is lower aggressiveness and longer guiding integrations.

AXES:

equatorial telescope: altitude is same as declination, azimuth same as hour angle (degrees east/west of meridian however, when you do the gem meridian flip declination is now mapped to altitude so that

dec = 90 = alt = 90
dec = 45 = alt = 135
dec = 0 = alt = 180
dec =-45 = alt = 225
notice how dec now travels in reverse compared to altitude and azimuth is mapped to hour angle with displacement of 180 deg

BACKLASH: http://www.bbastrodesigns.com/operate_backlash.html

10/13/03 Adopted new backlash take up algorithm. Previously, when motor reversed direction, the move motor software code detected the changed direction and caused the backlash to be taken up in the new direction. However, this was done in a standalone routine where any backlash to be taken up by either motor was made into a standalone movement. The software was not tracking the object during this backlash take up. This is not satisfactory for equatorial users whose autoguiders command a Declination guiding movement and expect the Right Ascension motor to continue precision tracking. Hence, a new algorithm.

Now, the tracking function adds backlash to take up to the tracking move to make. Backlash take up is thus one continuous combined motion with the tracking motion. The other motor continues to track without entering a special backlash take up routine. It is important to set good MaxIncrMsPerPWM and MsHsToggleIncrMsPerPWM values, as these determine how the microstepping takes up the backlash as quickly as possible. Set MaxIncrMsPerPWM to 2, 3, 4, or 5. Higher means faster take up but at some point the motor will not be able to get up to speed and will jitter. Set MsHsToggleIncrMsPerPWM to MaxIncrMsPerPWM unless you want to experiment with halfstep tracking.

Added menu option to the motor category to zero out backlash. Added shortcut key '%' to zero out backlash. Mel Bartels

Backlash behavior should be as follows

**************************************************

tracking with autoguiding on: motor commanded to move takes up backlash per the fastest possible speed as allowed by the microstepping parameters in config.dat other motor continues to track while above motor takes up its backlash (new behavior)

**************************************************

tracking using handpad to center object (no autoguiding turned on): any motor commanded to move stops tracking for both motors, as before motor commanded to move takes up backlash when handpad direction button released (new behavior) other motor pauses during the handpad commanded move, BUT, when the handpad button is released and tracking resumes, the backlash takeup in the other motor no longer pauses the unaffected motor

**************************************************

tracking off: handpad moves motor, no backlash takeup done until tracking turned on (previously, backlash would be taken up at start of handpad commanded move) Mel Bartels

+++

When not tracking, the motor will not move the extra amount to take up backlash. I debated whether to code this behavior (scope.exe used to do this) but decided against it for a couple of reasons: extra coding and debatable as to its worth. Mel Bartels

+++

Q. I just found that somewhere along the way, "BacklashMsArcsecSec" got dropped from the config.dat file.

A. Backlash now works at MsArcsecSec speed. The reason for this is that backlash is now capable of being taken up on one motor while the other one is still tracking. This for critical guiding reasons. Consequently, to code a different microstepping speed and to recognize when backlash takeup is in effect for one motor but not the other, was rather difficult in the code since there is but one microstepping speed that is applied to both motors, so I simply made the backlash speed the same as the microstepping speed. Mel Bartels

+++

You can adjust the backlash by using the submenu in the control section.

+++

negative backlash. This unusual option means that the motor will move in the opposite direction than expected in order to take up backlash

+++

Q. ...the backlash correction with small movements seems to occur AFTER the button is released. I would have expected to have occurred when the button was first pressed. Is this normal?

A. Backlash takes up after the movement. Rich Bowden

+++

If you can go "out and back" with good results but have trouble with "one way" slews, here's something to try...Do the "one way" slew to an object and consider the error as backlash and enter it as such then try again. You may find that you've had more backlash than expected. Ken Hunter

+++

AbacklashSignalPPortPin17: if non-zero, then parallel port pin 17 used to signal the direction of the 'A' motor backlash: logical high if backlash direction is CCW, logical low if backlash direction is CW (can be used to control motorized counterweights); starting value is logical low. Mel Bartels

+++

The most backlash I have seen in an RS components gearbox is ~12, not including the 250:1 and their 500:1 models. Lenord Stage

+++

I have never had any success with backlash active when autoguiding. You do not need it in RA (motor is always running in the same direction, just faster or slower per guide commands, and for Dec, if the mount is not perfectly Polar Aligned Dec will also always move in the same direction unless you are chasing seeing. If that's the case, you might decrease how often you are sending guiding commands in Dec? Chuck Shaw

BALANCE:

Lots of things can change the evenness of the (micro) steps, such as increasing the torque load by unbalancing the tube assembly. Mel Bartels

+++

I use a turnbuckle between a pin that is on the exact altitude axis and another one down on the side of the rocker box. You can clamp down a very high pressure then, and eliminate virtually all slippage.... You want to have the OTA balanced as well as possible also.....even though I found I can be pretty far off in balance if I have enough tension set up with the turnbuckles. Chuck Shaw

+++

I've just fitted balance weights to my 10" Dob, and it now balances at all altitudes.

My weights are long lead rectangles with a slot in the centre parallel to the long axis. Each weight has one edge filed flat, and rests on a wooden guide on the lower edge. It's secured by a single bolt and washer thro' the slot. The key to getting the balance adjustment is to stop the two weights interacting. One is fitted to the back of the tube, with its movement axis parallel to the tube. The other is just in front of the pivot, with its axis at 90 degs to the tube axis (I have a hexagonal wood tube). Thus when the tube is horizontal, you adjust the rear weight to balance. Now put the tube vertical, and adjust the front weight. In each case the weight you're adjusting moves horizontally. The other weight (if you moved it) would move vertically, and hence wouldn't change the balance at that point. Hope this helps. (Strictly there should be a third weight mounted horizontally across the tube to keep the balance over the pivot, but this would only be needed if you found the tube wanted to tip sideways on the base) David Hucthinson

+++

My microcontroller based telescope driver is an autonomous system which has everything on board. Yesterday I used it for the first time on my 14" telescope ! More details can be seen at http://martin-cibulski.de/atm/mount_controller_1/index.htm

My controller supports the basic functions which I need for visual observing:

- Alt/Az moving in 5 speeds - Two star alignment and Alt/Az<>Ra/Dec transformation - Display of current Ra/Dec position - Tracking

- Slew to any Ra/Dec position - Object databases, currently Messier, NGC, IC It does not support: - PEC - Lx200 commands (is planned) -Encoders - Autoguider interface Martin Cibulski

BROWSER:

Microsoft's Internet Explorer: To make this Web page available offline:

1. In the Favorites menu, click "Add to Favorites". 2. Select "Make available offline".

+++

Internet Explorer: You can save a Web page for offline reading by clicking the File menu and then clicking Save As.

COLD WEATHER:

If you decide to add a heater, try using a dual-element, 12V tail light bulb with the elements wired in parallel circuit. Make a clamp o hold the bulb and socket against a robust part of the gearbox case away from the shaft ends. Use thermal grease between the bulb and gearbox to increase thermal transfer. Add a thermistor or other Thermometer (a cheap indoor/outdoor or candy thermometer works well). Then, wrap the gearbox in insulation and seal out the light with Aluminum tape. Connect the bulb circuit to 12 V supply with a rheostat (dimmer switch or large variable resistor) in series with the + lead.

This heater should be capable of producing up to 30 watts of heat, it is reliable & redundant (i.e. if one element blows while observing, the second should be able to keep you going until you can change the bulb) and simple to build. The drawback is that depending on the temp and your insulation, it could draw a couple of amps. Larry Bell

+++

Local observatory uses low temperature lithium grease in bearing and gear assemblies. Mel Bartels

+++

I use fishing reel lubricant (its a very lightweight grease) that has Teflon in it. I get it at sporting goods stores. It does not seem to get stiff in the cold, but it DOES attract dirt. (WD-40 will also wash it out of the gear quite easily when you need to clean things....

You DO need some type of lubricant....... I have never found silicone spray to be enough of a lubricant.... WD-40 is NOT a lubricant, it is a solvent and a water displacement material (thats what the WD stands for actually....its the 40th version of a mix of stuff initially used to coat Atlas rocket's outer aluminum skins to protect them from corrosion) Chuck Shaw

+++

I too use a 486SX-25 CPU in my old Compaq LTE-Lite4/25c. I do notice a slight dimming of the LCD screen when I allow it to become cold, but I keep a small cover over the laptop when not looking at the screen, and that seems to help. Chuck Shaw

+++

I put the laptop into a box, more precisely a solid plastic laundry basket that is on edge. The laptop generates enough heat to keep itself warm. In the observatory on nights down to 0 to -20F, heating blankets might be necessary. Mel Bartels

+++

I use my laptop in an upturned solid walled plastic laundry basket. I always start it warm, that is, keep it in the car until I am ready to turn it on. On the coldest of nights it has functioned on its own heat generation.

At the observatory where machines are started cold, heating pads are used underneath the equipment when it is outside of the warm room. Mel Bartels

+++

I have a couple of 486 laptops that I have used down in the 20s (Fahrenheit) with good results. They do have active matrix screens rather than the dual-scan variety. I find that they generate enough heat on their own if you take some measures the retain the heat. I generally sit the laptop on a piece of closed-cell foam and sometimes lay a towel over the top when I don't need to see the screen. Normally, I let the hard disk timeout when I first set things up, then turn off the power saving functions and run strictly off floppies. The floppy seems to tolerate the cold well. Dan Wilder

+++

I use an old 486 for my autoguider and the screen would get slow in the cold. I also found that the screen intensity was too bright in all settings. So I taped a red transparent plastic in front of the screen ( 1/8" air gap). It solved both problems !

You could also insulate the back of the screen (glue some thick felt ?). But I haven't had the need to go that far. Juan Herrero

+++

have been using the Bartel system for 4 years now in Canada in very cold temperatures. I had problem in the beginning in using P133 COMPAQ computer which had problems starting just below freezing. I than bought a Toshiba 486 50 mhz T-1920 (I thing) and I have no problem... no heating needed. I have used it down to temperatures to -35C and it works very well. The screen is very slow to respond though. Alan Rahill

ASCOM:

ASCOM is a loose collection of developers, led by Bob Denny, who develop and maintain the various drivers - most, if not all, in their spare time.

For the most part the driver sources are published and are available for anyone to download, this includes the LX200 driver. See http://ascom-standards.org/ for the ASCOM initiative.

There are two groups: Ascom-Talk http://groups.yahoo.com/group/ASCOM-Talk and Ascom-users http://groups.yahoo.com/group/ASCOM-users/ Chris Rowland

+++

Q. So, how does scope interact with ASCOM? or is the ASCOM driver built into scope-drive? I'm having a terrible time trying to sort out what drives what and who talks to whom with ASCOM, and just where the heck the code resides and which program is responsible for what!

A. It can be very confusing. ASCOM is astronomical common object model, built in VB6 for MS Windows 95+. It offers a driver interface to controlling programs such as planetarium programs, imaging programs, and dome control programs. Any of these programs can pick a ASCOM driver and be assured that the driver will operate the telescope that it is designed for. Naturally we want scope.exe to be a telescope that ASCOM can operate. The easiest approach is for scope.exe to pretend to be one of the telescopes that have an ASCOM driver. Scope.exe pretends to be a generic Meade LX200 telescope.

So, you get your controlling program(s) and configure them to control a generic LX200 telescope. Now, many of these programs have their own LX200 driver in addition to supporting ASCOM. So you actually have two choices here. Either should work. Connect a serial link to the PC/laptop running scope.exe and configure scope.exe via the config.dat file for lx200 styled external control. Fire it all up. The planetarium/imaging program talks to the ASCOM driver which talks to scope.exe which talks to the motors via the circuit board; scope.exe then talks back to the ASCOM driver which talks back to the planetarium/imaging program. Mel Bartels

+++

Q What is the advantage for using the ASCOM interface for a planetarium program to control scope.exe rather than simply having the planetarium program interface with scope.exe directly using LX200 protocol? I can see the ASCOM interface being VERY useful when something like FocusMax is controlling everything (sending camera commands to MaxImDL running on the same machine as FocusMax, and then also sending focuser commands to scope.exe based on the data FocusMax receives from MaxImDL from the images it takes during the autofocus routines.... LOTS of interfaces going on during all of that (mesmerizing to watch also!)

But for simple control of scope.exe by a planetarium program, and feedback from scope.exe back to the planetarium program, seems just the direct LX200 interface is simpler?

A. more and more programs, it appears to me, are going with the ASCOM driver and foregoing their own LX200 drivers. Mel Bartels

A. The main advantages the I see are that the ASCOM suite can act as a hub or central connection to the scope so that multiple programs like MaxIm and ECU can control the same scope while executing on the same processor. Currently you would have had to logically disconnect the COM port connected to the scope from the first program and then connect to the same COM port in another program like ECU. There is something in ASCOM called POTH (Plain Old Telescope Hub) that can permit multiple programs executing simultaneously to use the same physical serial connection to the scope. Very handy to use.

Another advantage is that the author does not need to maintain his own extensive set of commands for multiple scopes. All he needs to do is connect to the ASCOM chooser dialogues to let his user pick his scope type. From that point forward he is assured of a consistant set of commands to control the scope. Gotta ease the work load, I'd think. Marty Niemi

COMETS - ASTEROIDS - SATELITES:

+++

Mel and I have a develop a new feature for Scope. It soon will be release. This new feature is the ability to track a comet. I have created a program calculates the position of the current comets. Its outputs are a regular DAT file and a CDF file. The CDF file is a new data file format. It contains not only the position of a comet but it also has its drift rates. This allows you track a comet. Both programs can be download from my website. http://www.geocities.com/kindellism/ Rex Kindell

+++

Q. If the comets.exe (or Solarsys.exe) is run properly with current elements before loading scope.exe, does this mean when I load a given comet and perform a slew/goto to it via scope.exe it will not only locate the comet, but also track the comet at it's current drift rate? If so, how do I revert back to sidereal rate tracking after I'm done? Brian Sherrod

A. Use the menu option to set drift to zero. Mel Bartels

TRACKING 2002NY40

I want to use the offset equat command in a scroll file to track the asteroid tonight. I've plotted it's position for each hour from 20:00 EDT to 02:00 EDT. I intend to find the target and center it then start the scroll file in auto scroll mode with an offset equat command determined for the positions for each hour (3600 seconds).

I used The Sky to compute the coordinates based on the location where I'll be observing. I did it for the start of each hour after I downloaded fresh predictions this afternoon.

Do I need to use a minus sign for the RaMin and RaSec since the asteroid will be heading westerly? The DEC is increasing so no problems with the direction of movement. Don D'Egidio

I am approaching it slightly differently. I built up a .dat file of the RA/Dec data I got from the MPC (5 minute intervals). I also included Vega and Beta Lyra and M57 in the data file to be able to do an offset slew for best accuracy (although I did a 15 deg slew right to it last night). I took a lot of shots last night where I had NY40 starting at the lower left of the frame, and progress up to the upper right. I was takling 15 sec images each 30 seconds. The result was a series of images of the same starfield but with NY40 in a different spot on each one. I will build a movie from them (I hope). It took 40 images to have NY40 go diagonally across my 10x13 arcmin FOV last night. Since it will be moving about 4 arcmin/min tonight , I will probably only fit in 5 or 6 frames before it is out of the FOV..... Be sure and use topocentric coordinates for the RA/Dec you use tonight since the object is so close. Your location on earth will make a noticable difference in where it appears in the sky, and can throw you off if you are only looking thru a ccd camera's FOV. Chuck Shaw

I used a completely different approach:
I was using "Guide+Stay" for about 20 minutes tracking with an ST6 on the asteroid itself. Handpadupdate "ON". When the adjustments were within the seeing I stored the values. The drift values became about - 80 in DE and -100 in RA at begin of the night and -180 / -250 at the end. Then I took 80 each 15 sec. exposures (one hour). With this operation the asteroid was almost an hour stable on the pixel. Because of it's acceleration the drift values had to be updated from time to time (still in guide+stay mode). A .mov file out of that operation has been uploaded (2002 NY40.mov). Was a great experience with scope.exe. Thank's Mel. No commercial product allows that. Rolf

+++

Satellite Tracker http://www.heavenscape.com/ has been driving my LX200 8" for years and has tracked hundreds of satellites Ray Talipsky

+++

I recently downloaded a copy of "Predict" at http://www.qsl.net/kd2bd/predict.html

CIRCUIT: circuit boards the stepper circuit Getting the stepper windings properly hooked up

pcb1.gif - Mel Bartels' drawing of his circuit board Stepper circuit parts list (bottom of page)

Mikel Berrocal and Carmello Fernandez Board modification for bipolar motors

+++

...the software expects the winding sequence to be A B A- B-, not the A A- B B- that some other controllers ask for. ...diagram contributed by Bob Rubendunst Chris Rowland

+++

Mel's board is basically an amplifier, with regards to the main stepper motor drive portion. Whatever output is coming from the parallel port gets amplified and sent over to the stepper motors. Since the parallel port pins are digital, the signal sent to the stepper is either on or off.

Mel's answer to current control is Pulse Width Modulation during Microstepping. By Rapidly turning the coil on/off with varied amount of 'on' cycles vs 'off' cycles, the current can be controlled very accurately. Currently his software can support up to 40 microsteps using this method. For things like Field De-rotation and Focus control, Mel did rely on independent IC driver's James Lerch

+++

The 25 pin "D" connector is what links the board to the computer that provides all the intelligent control. A printer cable connects the 25 pin D connector to the parallel port on the computer (where you would normally plug in the printer). Most people use an old laptop because it's portable, but there's no reason you can't use a desktop PC if you don't need the portability. Most old, cheap computers work well - a 50 MHz 486 DX2 based computer (or anything newer and faster) will work fine. The board simply serves as an interface from the computer, converting the low voltage, low current control signals from the parallel port to the high power signals feeding the stepper motors and providing interface from the handpad to the computer. The small 9 pin "D" connectors go to the motors on the two drive axes. The RJ45 connector (the one that looks like a large phone jack) goes to the handpad. The red and black wires go to your power supply (red to positive, black to negative). The "standard" supply voltage is 12V, although many people (me included) use higher voltages to get better slew speed. I would recommend you start with 12V, that may work fine for you; alot depends on the motors you use. Steve Rivet

+++

The PCB can support 2-3 amps per winding. If more than that, then use the current limiting circuit to limit current to this level. Mel Bartels

+++

I have found that there is approx. 5 volt difference between the controller ground and the laptop frame ground. As soon as I inter-connected the two grounds, the problem stopped. No wonder the motors were rattling as the transistors would have been permanently switched on, so they would have energized all the fields at same time. Just goes to show that not all laptop power - ground designs allow a "true" ground. Steve Ross

+++

a single transformer used to generate two isolated voltages is fine - but be very cautious as the grounds and the supply voltages cannot touch each other! You can also consider those DC to DC converters - use a single voltage source from the transformer, and use a DC to DC converter to step down the voltage and operate the driver board.

The issue of tying the grounds together is like this in my thinking: a common ground is usually OK, as it is unlikely that a voltage spike will travel through the ground into the PC or laptop, and it is unlikely that noise will be a problem on the ground. However, it is possible for voltage spikes and noise to be transmitted on the ground, thus some advocate the totally isolated grounds. In my thinking, in tinkering with circuits for 35 years, it is more likely that you will mess things up by having isolated grounds and forgetting this and accidentally touching something between the two grounds, than it is that a voltage spike or other problems will occur along a single common ground.

In other words, nothing is impervious to destruction, as a virtual shelf full of blown transistors and other components bear silent testimony to at my home!!! Mel

+++

..... check your board grounds to ensure the computer ground and 7805 grounds are either properly isolated (if you chose to go that route) or tied together (simplest approach) Larry Bell

+++

Make sure you have a solid ground (power return) line to the supplies, Mel's circuit board and the motors. DON'T daisy chain your ground wires. A single point ground with beefy wires will be less prone to ground loops and switching noise. Larry Bell

+++

The 6 holes for power and ground go as follows (board face up with the 6 holes to the lower left): 1. gnd 2. +5 vdc 3. gnd 4. +12 vdc 5. gnd 6. +12 vdc

for complete isolation, use separate computer and motor grounds, and supply an external +5.0 VDC source for the computer side or use the +5VDC output from pin 1 of the joystick port DB15; the vast majority of us will not require this total isolation, instead, tie computer and motor grounds together, and supply the computer +5 VDC from the 7805 power regulator (U6 - it will get hot supplying both sides of the board so consider a heat sink if not well ventilated or used in hot clime): do this best by jumping 1. gnd and 5. gnd together, then jump the bottom lead of the 7805 to 2. +5 vdc, and finally bring out two wires from 5. gnd and 6. +12-24 vdc for the ground and positive power connection respectively;

Be very careful when testing so as to not risk your computer's parallel port. Use a 6 volt drycell battery for motor voltage during initial testing. Mel Bartels

+++
With the same concerns about my laptop's parallel port, here's how I tested my board. Starting with the driver board powered on and disconnected from the laptop, motors and handpad connected:

Starting with the driver board powered on and disconnected from the laptop, motors and handpad connected:

#1 Using volt meter, Black wire of multi-meter to Ground probe parallel pins with red wire looking for voltages higher than 5v, there should be none.

#2 Using a milli-amp meter, ground -> 100ohm resistor -> Black wire of meter, probe parallel pins with red wire recording milli-amp reading for each pin

A) Pins 1 -9 should show very little to no current (Depending on invert option) * See Note 1

B) Pin 10 should show 50 milliamps only when Hand pad Left, Upper left, or Upper Right option button pressed * See note 2

D) Pin 11 should show 50 milliamps only when Slew Speed switch on

E) Pin 12 should show 50 milliamps only when Hand pad Down, Right, Or Upper Right option button pressed *see note 2

F) Pin 13 should show 50 milliamps only when Hand pad Right, UP, or Upper Left option button pressed *see note 2

G) Pins 14 -17 should show very little to no current (Depending on Options)

H) Pins 18 - 25 should show no current

#3 using a milli-amp meter, +5vdc -> 100ohm resistor -> Red Wire of meter, probe parallel pins with black wire recording milli-amp reading for each pin

A) pins 1-9 should show very little to no current (Depending on Invert option)

B) Pin 10 should show 15.6 milliamps with no hand pad buttons pressed * see note 3

C) Pin 11 should show 15.6 milliamps slew speed switch off

D) Pin 12 should show 15.6 milliamps with no Hand pad buttons pressed * see note 3

E) Pin 13 should show 15.6 milliamps with no Hand Pad buttons pressed * see note 3

F) Pins 14 -17 should show very little to no current (Depending on Options)

G) Pins 18 - 25 should show 50 milliamps of current

* Note 1 While probing pins 2 - 9 in either step #2 or #3 depending on invert option, Also check to see if you stepper motors are turning on. In my case I could here the stepper motor 'Twitch" and it became very difficult to turn the shaft of the stepper motor.

* Note 2 With the diodes in series and button pressed, you may only see 44 milliamps of current (If I recall correctly)

* Note 3 Again, with diodes in series and Key pressed, you may see 6 milliamps of current. No key press should still be 15.6 milli-amps

James Lerch

Zener diodes http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/zener.html are devices that are designed to conduct current in ONE direction and ONLY when the voltage is above a certain level, and to remain "open circuit" at lower voltages. In Mel's circuit, the zener is used to handle the current generated by the stepper motors when the current is removed from a winding. The program uses pulse width modulation (PWM) to control the stepper motor. PWM basically means that the motor windings are turned on and off very fast. Each time the winding is turned off, the collapsing magnetic field induces a current in the winding circuit that flows opposite to the direction it flows during normal operation. In order to ensure smooth operation, its necessary to "dump" that current. A diode is usually used for that purpose; however, a diode will conduct currrent in one direction from 0V all the way up to its rated value, thus dumping all of the power from the collapsing field. The motor can be made to run faster/smoother if you dump SOME of that power, but not all. That's where the zener come in. By replacing the traditional diode with a Zener which will stop conducting when the voltage reaches its cutoff rating, not all of the collapsing field power is drained away.

The 1W zeners can handle 30mA continuous at 30V. As the voltage drops, the current can increase (P=V*I) and still stay within the power spec. If you're running your motors at 12V like many people,

the 1W zener will take 80mA continuously. The 5W zeners can handle 5 times as much, so about 400mA.

The since the zener is used in a pulsed circuit, the current flowing is not continuous, so you can probably increase the current you put through it without hurting the device. Its a difficult calculation to determine the peak current flowing in the circuit as the field collapses--I used to be able to do these calcs, but wouldn't want to try it today. I'd say that if the zener isn't getting hot during a heavy period of slewing or tracking then the 1W is adequate and you shouldn't worry about it. Larry Bell

+++

Q. but has anyone actually tried various Zener voltages and noted how slew speed is affected?

A. Higher zener voltage ratings means quicker decay times as the current's exponential decay to zero is truncated prematurely. The downside is the higher voltage spikes. Doubling already high zener ratings improves slew speed very slightly, doubly very low zener ratings can improve slew speed up to 2x or so. Mel Bartels

+++

verify that the outputs from the TIP120's are working by disconnecting motors, then putting the software in parallel port test, and using a voltmeter, verify that logical high and logical low work by turning on the output voltage and turning off the output voltage. Mel Bartels

+++

Use the parallel port test function and measure the results at the Collectors of the output transistors. Select the "H" key and all the Base inputs should go HIGH causing ALL of the transistors to conduct and making ALL of the Collector voltages go LOW. (a voltage around 1 volt is EXPECTED). Make these measurements fairly quickly as the transistors will be ON HARD for the duration of the "H" test and may overheat if left in the ON HARD state (Collectors LOW) for very long

Select the "L" key and all the Base inputs should go LOW causing ALL of the transistors to NOT conduct and making ALL of the Collector voltages go HIGH. (a voltage very close to the power supply input). You can leave the transistors in this state forever without worry. If this is not the case, then the problem is in the portion of the circuit that is different from the rest. (All 8 channels are identical). You may find that the behavior is exactly the opposite as regarding the "H" and "L" due to circuit differences BUT ...

ALL 8 TRANSISTOR COLLECTORS SHOULD BE IN THE SAME LOGIC STATE DURING THE PARALLEL PORT "H/L" TEST AND SHOULD CHANGE STATES TOGETHER WHEN YOU CHANGE FROM "H" TO "L" AND VICE-VERSA.

During the PORT TEST the voltage is applied to all pins continuously. A bad situation for the motors if the voltage is above the safe dissipation rating of the motor. You want to test quickly if you are using overvoltage on the motors. ken hunter

+++

board connections: There should be no connections to the three connectors nearest the DB25 connector on Mel's board. That is for use if you are taking a 5 volt supply from the computer for powering the logic on the board. The supply voltage to the motors connects to the connector furthest away from the DB25. The 7805 voltage regulator will supply the 5 VDC for the logic and the signals to the parallel port mn

+++

...accidentally hooked up the power backwards. To my surprise, the 47uf electrolytic capacitor exploded. Luckily for me the board was upside down.

Anyway, I was intrigued and sent off for a supply of them. When they arrived, I opened up one end of a small cardboard box and poked the capacitor wires out through the closed end. 10 out of 10 exploded when 12volts reverse parity power was applied. I bounced the aluminum cans off the ceiling. I shot them along the floor. The greatest distance I managed was 23 feet.

All this surprised the hell out of me, so I thought I would pass the results along and put in a plug for Bob Norgard's protection circuit. http://groups.yahoo.com/group/scope-drive/files/circuits/Protection_ckt.jpg bd 9573

+++

Regarding the circuit protection discussion; there seems to be two camps of folks. The first puts a fuse in line with the motor power supply circuit (usually the positive lead), then connects a beefy doide between the positive and negative leads in the reversed bias direction so that when power is connected properly, no current flows through the diode. If the leads are inadvertently connected backwards (i.e. positive to negative and negative to positive as could happen if you're using alligator clips on a car battery), massive amounts of current flow through the diode and the fuse blows open; thus "protecting" the circuits. This kind'a works, but until the fuse opens, the circuits are subjected to the reverse voltage. This is because fuses blow "slow"--even a few milliseconds can do serious damage. Further, since the circuit can't be easily tested (you must hook up the power backwards to test it) you're never really sure how well protected you are until its too late. Such a circuit is often referred to as a crowbar circuit.

SECOND CAMP: As several folks pointed out, a better protection method is to simply put the diode in "series" with the positive lead such that current flows when the leads are properly connected. Of course, in this configuration no fuse is required because the diode prevents any reverse voltage from getting to the circuit if the wires are connected backwards. This method will result in a very slight voltage drop at the source (between 0.3 and 0.6 volts depending on the diode used). In most applications this is insignificant. For those who absolutely must have the full voltage, a better way to protect the circuit is to key the connectors using a harness/connector arrangement that mechanically protects against connecting the positive/negative leads backwards. Larry Bell

+++

I do strongly recommend a Shottky diode in series with the circuit. I personally use 8A 100V diodes in TO220 case and due a low voltage drop (0,3 to 0,4V under load) it even does not need a heat sink (driving two 3.5 Ohm steppers from 16V). There are even double Shottky diodes in the same case - 2 x up to 16A / 100V. If connected in parallel, you can get a very low voltage drop close to 0,2 - 0,3V. Pawel

+++

I install the diode the way it is used from the diagram in the files area (Bob Norgard). LOTS of users have hooked up things backwards and none have had any trouble with the board. One guy did it multiple times in one evening with no harm. All are on 12 volts too...? Is this change really needed ? Lenord Stage

+++

Lemme propose a time tested and no compromise . Like no diode drop/blown fuse etc. The simplest way is to use connectors that can never go in otherwise...the Only FOOL PROOF method is to use connectors that can never be connected the reverse way Sanath Kumar

+++

The scope.exe program has provision for auxiliary control. These are two control lines for you to use to operate an external motor. For instance, you might operate a dome motor via two solid state relays that take 5 VDC TTL input Mel Bartels

+++

I am building a dual axis, 4 speed, quartz controlled driver for 2 steppers, it is so simple that it requires no laptop nor PC, and no microcontroller at all. Basically, it is an oscillator, two 74HC4060 divider-counters, and the sequence of both stepping motors is stored in a single EPROM. From it, 4 bits of data output drive one stepper, and the other 4 bits drive the other stepper, by using one 2803 darlington array.

It is as simple as 4 IC, one oscilator, and a remote. As the Eprom can store a lot of info for almost $3, I will store on it 4 speed: tracking, X2 and X8 corrections, and a X128 high speed for rewind my unusual tracking gear system. The tracking speed is achieved by microstepping (4 microsteps per step).

If anybody wants the schematics, I am drawing them in ORCAD, so I can send it, or convert it as an image in a fews more days. Also I will prepare a hex file with the contents for the EPROM, so anybody could record one As we cannot attach files here, please send me an email to ariel.desseno@m... (you have to send this through the Yahoo scope-drive group) Ariel Dessino 4122

+++

Q. ...one thing I've found confusing - the diodes for the handpad. No rating/type are specified, can someone provide me with a type/rating for diodes that would work for the handpad? There's a bewildering assortment.

A. Just about any generic diode will work. I use the 1N4001 from radio shack. 1N914 also works -- just pick a standard rectifing diode. Current ratings between 0.5 and 1 amp are readily available and are overkill for the application. Larry Bell

+++

My controller and all the circuit diagrams that I've seen are meant to be powered indefinitely. That is, the motors can remain powered up for as long as you wish. I've left mine powered for up to 3 days at a time at large star parties so that I can find objects in the daytime based on my nighttime initialization efforts.

The key is that the software remain in active control of the parallel port. The parallel port serves as input into the controller. If you have high amperage motors and something else other than scope.exe activates a parallel port control line, it is possible for the motor and possible for the controller to suffer damage from the sustained current flow. This never happens when scope.exe is in control as the software never allows full current to flow for more than a fraction of a second.

It is good safety practice to turn off the power to the motors when not using the system. It is typical but should be tested for each PC/laptop that if you are running in straight DOS without any command line drivers that when you shut down the machine the parallel port will continue with the last commanded state, ie, no current flow. However, it is very likely upon machine restart that the parallel port will come alive in some unknown state, at which point you may suffer damage. Mel Bartels

+++

Q...if I could use steppers rated at 1.65V / 4.7A (clearly with current limiting!) Can the board be run on 6V so that the overvoltage would not be so great?

A. Yes, I've run at 6 volts. Any lower though and there will not be enough for the 7805 to output the 5 volts needed for the ICs on the board. Mel Bartels

+++

Q. Has anyone tried the "Rafael Gonzales Fuentetaja" version of Mel's circuit? There is a link at Mel's website to his detailed Spanish article on his Mel Bartel's Scope circuit implementation. (http://es.geocities.com/astrorafael/motorscopio/motherb.html)

From the computer translation, basically what Rafael has done, is to replace the TIP-120 transistors, with "pin-compatible" MOSFETs, IRL530's (International Rectifier), and replaced the 470 ohm the base resistors R21-R24 and R33-R36, with 10K ohm. To quote Rafael (from a barely understandable ALIS-GIST computer translation of his website document), let me paraphrase:
The original design of Mel Bartels uses TIP120 Darlington power transistors, that are adapted for this use, and are inexpensive. Nevertheless they suffer from one defect: with high currents, they get hot. The heat problem is easily solved with thermal heat dissipators. Nevertheless, it is a pain to turn all this electrical energy into heat, especially if it is also a large drain on your 12v field battery. For that reason, it was decided to use MOSFET transistors, since they are more efficient, and they do not require heat sinks. The one chosen is the IRL530N, easily obtained, but more expensive, at twice the cost (in Spain ?) of TIP120's, but I believe that it is worth the trouble. Although it isn't critical, there is a certain controversy on the best value for the gate resistors (R21-R24 "base resistors" in Mel's circuit), which should be between 1K and 10K ohms. For those that wish to switch between power transistors (TIP120) and MOSFETs (IRL530N) - since they are "pin-compatible", and both are TO-220 format - all you need to do is simply change the base resistors & transistors. I think Rafael is right on this slight modification. Anyone else try his approach? It would certainly go a long way to ease the battery drain used to heat up the TIP120 heat sinks. May also solve this particular problem of easily blown power transistors. Joe Zelinski

A. When I implemented the overvoltage circuit for slewing, I changed the TIP120s to IRL530s and used 1k base resistors. It has worked well at slew voltages up to about 28 volts. Mike Wade

A. Those MosFets are pretty sturdy and well protected internally - they even have fast, heavy-duty internal protection (avalanche diode) against overvoltage, something the TIP darlingtons haven't - and as has been mentioned before on the list, 1N400x type diodes may be much too slow to turn on an external overvoltage protection circuit to save the TIPs. The MosFets also have very low voltage loss when on - resistive, at around 0.15 ohms or less at a few amps, so at e.g. 2 amps, the voltage is less than 0.3 volts, as opposed to somewhere around 2 volts for the TIPs. The difference means more voltage across the motor windings (and a bit more heat in them, but much less heat in the transistors) - total current drawn from the supply will increase slightly with the MosFets.

(reply from Joe - I think you might have made a typo, - the heat generating power (P=I^2*R), in the MosFet in your example would be (3.25^2*0.15) or actually about 1.6 watts, which is only about 1/4 the power lost to heat if the TIP120's were used. That is very significant - the MosFet approach is definitely the one to use, for that very reason alone.)

(Nils again) - In my limited experience with MosFets (other types), using too small a resistor to the gate can cause parasitic oscillations - check with an oscilloscope if you have one, but my guess is that 10k is quite safe.

My guess is the MosFets are a good deal more difficult to damage (than TIP120's). The built in avalanche diode can dump a lot of energy at minimum 100 volts (maximum not specified). I'm not even sure there is any need for overvoltage protection. I would on theoretical grounds expect no (or only extremely brief) voltage spikes above 3 times the supply voltage, and with small steppers I have not seen (on my oscilloscope) any - is there anyone out there with a (storage) oscilloscope who can veri/falsify this (setting the protection circuit appropriately, checking if it turns on at all!)? Nils Olof Carlin

+++

Q. Just another question: are the TIP 120 transistors replaceable with other ones (2N3055 ...?)

A. You'd be better off replacing them with a suitable HEXFET that has very low "ON RESISTANCE", LOW LOSS AND nearly ZERO heat generation. The 2n3055 is a low gain transistor usually used as a linear amplifier which implies linear circuit losses and heat generation while the HEXFET's are fast operating switches... Just the ticket for a PWM Circuit. Ken Hunter

+++

Some notes on putting the kit together and testing it. Hopefully this will be helpful to others. \

I unpacked the kit and verified the contents. Everything looked good and the kit was complete. Even before starting to put things together I made some notes on the layout diagram. I labeled all the parts on the diagram so that it would be easier and faster to "populate" the board.

I started soldering with the sockets, then all the resistors. Next I soldered in the diodes. After this the taller parts, including the transistors and voltage regulator. Lastly the connectors. The handpad connector was the worst as it had locking plastic tabs that didn't want to line up. I prefer this order, but others may prefer another order. I recommend that novice solderers start with hard to damage components like sockets and resistors first.

My method for installing and soldering parts was to put the part in the correct holes, then bend the leads a little to keep it from falling out. I inserted every component of one type, then soldered all of one type. For example, I installed all the 220 Ohm resistors in one step, then flipped the board over and soldered all them in the next step. Novice builders may prefer to work with one part at a time.

Check things as you go along. After installing a part, I verified it was the correct value and that it was in the correct spot before soldering it. Note that some parts have to be installed in a particular orientation. Verify this as well before soldering. Inspect every solder joint once it is done. I checked every solder joint of a component type after soldering all of that type. Really look close with lots of light and a magnifier if needed. I only messed up one thing, I accidently broke a trace and had to solder in a jumper.

... I would add that using too little solder is much better than using too much.

I decided to not isolate the parallel port components with a separate power supply. I used the output of the 7405 regulator to power the "computer side" components and tied the grounds together.

All together, it took me about 2.5 hours to finish soldering the board. I bought the preassembled handpad, so I didn't have to make one.

I then did some tests for shorts and continuity. I verified that there was significant resistance between the power and ground leads. I checked for continuity of power and ground with the power inputs of the digital chips.

Then I performed a "smoke test" even before installing the socketed parts. I did this without having a PC or the motors attached. I hooked up a 12 volt gel cell. Of course I was very careful about using the correct polarity. I verified the 5 volt supply and checked that the regulator didn't overheat. This was good so I checked the digital parts ('ls08s and optos) for correct voltages. I then removed the 12 volt power from the board.

This all checked out so I put in all the socketed parts. Next I powered the board up again and verified voltages. With no computer attached all the motor outputs are on. I observed this with Mel's LED motor tester. Once again everything looked okay, so I again powered the board off.

It took me a few minutes to find a male-male cable. I have a large junk collection but I still had to make a male-male cable from a couple of m-f cables. I would recommend getting this part ahead of time as it would be frustrating to get the board done and not be able to test it.

I hooked up the board to my notebook computer and started scope.exe.

I first did the parallel port test and verified that the LED tester showed the expected results. You don't want to do this test with the motors attached as it will turn all winding on.

Next I ordered a move of a few thousand microsteps in each axis. The LED tester showed pulses and brightness changes meaning that things seemed okay.

Finally I hooked a motor up to azimuth and commanded a move. The motor moved as expected. I repeated this with the altitude motor. The motors aren't yet connected to anything on my scope. They are just running on the table.

After this I tested the handpad. It worked fine as well.

Now I have to package the board and put in a protection circuit to prevent problems if the power is ever attached the wrong way. I also have the "little details" of all of the mechanical work. I am planning on making JB Weld gears ala Tom K. I need to change teflon for bearings and some other things as well. I should have enough to keep me busy for a while. Joe Larkin

+++

CIRCUIT BOARD DESIGN SOFTWARE

There's some excellent PCB software available for free at www.expresspcb.com. Expresspcb is in the business of making custom boards and provides their software so you can design your board and then send them the design for manufacture. However, after creating the design, you can do a cut (ALT-PrintScrn) and paste to Photoshop to create an etch mask. Software limited to 2-layer boards, but can create boards up to 29" X 28.5". You can vary the grid spacing, line width, etc. and comes with a large database of components.

Doesn't do automatic routing, which is the major downside, but you can't beat the price. Don Ware

+++

Circad is a very good electronic (design) program but, being rather expensive, it's hard to find justifications enough to buy it only for our very reduced need of telescope hobbyists, and therefore not for making money. Me too, I don't use anymore that Dos version I bought many years ago. On the contrary I downloaded the freeware reduced demo version for Win98 at the following address: http://www.holophase.com/dleval.htm, which has been running perfectly since a couple of years. Of course this demo does not offer many of the powerful features of the commercial version, but it is largely sufficient for our needs of hobbyists and it should not expire after 30 days like you supposed (at least the free Win98 demo I downloaded time ago, that I am still using now). Marcello Cucchi

+++

There is anoth