Coleco Telstar Combat! No Video Repair HELP!

[eDoc]

Hello all, long time reader, first time poster on the forums.  I recently picked up a junk Coleco Telstar Combat system made in ‘77 for a couple dollars.  It was pretty roached, but it’s a rather obscure system so I couldn’t pass it up.  However, after getting it all cleaned up, I plugged it to my TV to test it out, but all I got was static and no signal.

 

I tried many times turning the system off and on but mostly no signal would appear at all.  Sometime a staticky signal would fade in and out, but it was very inconsistent. Once, I even got a vague image of the actually game screen to appear for a few seconds.  I’m fairly confident that the video cable coming from the system itself is not the issue and I have it hooked up to a known working RF switch that works well with my Atari 2600.  So, I opened the system back up, removed the RF shield, and began to adjust the RF inductors (I’m using a modern tv).   

 

( I attached pics for reference) There appears to be 2 RF inductor adjustments under the shield, and a screw adjustment at the top right corner of the board.  I started adjusting the orange (bottom) inductor but that only caused the video to go from no signal, to staticky image, back to no signal.  So I tried to adjust the white (top) RF inductor, but it wouldn’t budge.  Upon further inspection, I saw the slot inside this inductor had several cracks.  I desoldered the inductor from the board to examine it further, but couldn’t get it to budge.  After resoldering it back on, I still have the same issue, however, the static signal seems to become more and more scarce, leaving me with no signal as I fiddle more and more with the inductors.  I also tried adjusting the screw adjuster to no avail, but I’m not really sure which RF adjusters I should/need to adjust.

 

Would the cracking in the inductor be enough to inhibit the video signal completely?  Is there anything else that could be causing this issue besides the inductors?  I have a rather limited knowledge on electronics as I have only been attempting retro repairs for about a year and a half, so any new nuggets of knowledge are greatly appreciated!  This is a really neat and rare system that I would like to have working in my collection.

 

Feel free to ask questions if something isn’t clear.  Thanks in advance!

 

(Note: I have the same problem regardless if I’m using batteries or the power supply).

 

 

[TheProgrammerIncarnate]

I would suggest bypassing the modulator altogether and composite modding it. I'm actually in the process of doing so for my unit, so it runs on more modern TVs and has better quality. The composite signal is in the picture below (blue circle around choked signal.) It needs some form of amplification to be visible though. I can't remember the resistor values I tried with it, but the picture did come out pretty sharp. I'd be happy to help if you decide to do this.

 

 

[Rob Mitchell]

Please keep posting of progress .. I would like to modify mine as well.

[eDoc]

6 hours ago, TheProgrammerIncarnate said:

I would suggest bypassing the modulator altogether and composite modding it. I'm actually in the process of doing so for my unit, so it runs on more modern TVs and has better quality. The composite signal is in the picture below (blue circle around choked signal.) It needs some form of amplification to be visible though. I can't remember the resistor values I tried with it, but the picture did come out pretty sharp. I'd be happy to help if you decide to do this.

 

 

@TheProgrammerIncarnate I’m so glad you brought this up!  I thought about modding the system, but i’ve never done an A.V. mod before.  I know it’s a pretty common thing to do on Ataris, but since this is a more obscure system, I didn’t know where to start.  I believe I will try to do a composite mod on this, especially if you are willing to help me out along the way

 

Forgive me for what probably sounds like a rudimentary question, but what all would I need to do an A.V. mod on this system?  There are several similar mod kits on eBay which are marketed towards Atari 2600/7800 (I attached a pic of what most of the kits include, and some of them come pre assembled), but would I need something in addition to one of these kits to amplify the signal like you said?

[eDoc]

4 hours ago, Rob Mitchell said:

Please keep posting of progress .. I would like to modify mine as well.

I absolutely will!

[eDoc]

3 hours ago, eDoc said:

@TheProgrammerIncarnate I’m so glad you brought this up!  I thought about modding the system, but i’ve never done an A.V. mod before.  I know it’s a pretty common thing to do on Ataris, but since this is a more obscure system, I didn’t know where to start.  I believe I will try to do a composite mod on this, especially if you are willing to help me out along the way

 

Forgive me for what probably sounds like a rudimentary question, but what all would I need to do an A.V. mod on this system?  There are several similar mod kits on eBay which are marketed towards Atari 2600/7800 (I attached a pic of what most of the kits include, and some of them come pre assembled), but would I need something in addition to one of these kits to amplify the signal like you said?

@TheProgrammerIncarnate As a side note, this kit includes a resistor that outputs a 75 ohm signal to the TV.

[ChildOfCv]

The important point about the use of the above kit is that the output needs to be 1Vp-p.  I'm pretty sure from what little of the Telstar schematic I've seen, that the resistors given with that kit are not suited for the Telstar.  You'd have to find the correct combination for the Telstar to adjust the output voltage swing.

 

Anyway, there are already threads on this board about doing the mod. 

 

Edited August 1, 2019 by ChildOfCv

[eDoc]

32 minutes ago, ChildOfCv said:

The important point about the use of the above kit is that the output needs to be 1Vp-p.  I'm pretty sure from what little of the Telstar schematic I've seen, that the resistors given with that kit are not suited for the Telstar.  You'd have to find the correct combination for the Telstar to adjust the output voltage swing.

 

Anyway, there are already threads on this board about doing the mod. 

 

Thanks for the advice!  I actually read through that entire thread prior to starting this one, but they mod an entirely different system than the one I have.  How different would modding my combat system be compared to the system in the thread?

[TheProgrammerIncarnate]

The difference *should* just be a different bypass point and different resistors for the voltage dividers. The Telstar Combat! uses 6x C cell batteries for an optimal voltage of 9V. That means you'd want a pair of resistors making a 1/7-1/9 divider, although I don't know what the theoretical best value would be (the batteries won't provide a perfect 1.5V) Could someone confirm if this is correct? @slydc

[eDoc]

Ok, so the chip on combat is an AY-3-8700-1.  After some research I (finally) found a schematic for an AY-3-8710 chip (attached).  The page describes the combat game and I’m 99% sure that the 8710 chip is just a 8700 chip with a small update, but has the same pin layout.  The second page I attached also identifies the video and audio outputs.  So, based on the other thread, I theoretically should be able to solder wires to the 6 video outputs and 6 audio outputs, solder the video outputs together and the audio outputs together, solder them to their composite Cables, and solder the grounds to the Vss on the chip, right?

 

I’m still a little confused what the resistors would be for since there was no mention of resistors on the other Telstar system.  That said, I don’t know much about resistors and voltage dividers.  I do know resistors will lower voltage, but I wouldn’t know how to/where to create a voltage divider if I need to.

[ChildOfCv]

5 hours ago, eDoc said:

Thanks for the advice!  I actually read through that entire thread prior to starting this one, but they mod an entirely different system than the one I have.  How different would modding my combat system be compared to the system in the thread?

Well, first it would help to know what chips are on that board besides the AY microprocessor and the 4011 over there.  I can't read the rest.

 

Also, for the modulator, I can't see the circuit board on the bottom there, but looking at the visible components I'm guessing that it's an oscillator to produce the carrier signal, and a modulator that looks a lot like the one on this page.  https://www.instructables.com/id/Amplitude-Modulator-and-Demodulator-Trainer-Kit/

 

That likely makes the yellow tuner the channel fine tune, and the white one is likely to be the modulation filter.  BTW, what channel were you attempting to watch it on?

[ChildOfCv]

8 minutes ago, eDoc said:

Ok, so the chip on combat is an AY-3-8700-1.  After some research I (finally) found a schematic for an AY-3-8710 chip (attached).  The page describes the combat game and I’m 99% sure that the 8710 chip is just a 8700 chip with a small update, but has the same pin layout.  The second page I attached also identifies the video and audio outputs.  So, based on the other thread, I theoretically should be able to solder wires to the 6 video outputs and 6 audio outputs, solder the video outputs together and the audio outputs together, solder them to their composite Cables, and solder the grounds to the Vss on the chip, right?

 

I’m still a little confused what the resistors would be for since there was no mention of resistors on the other Telstar system.  That said, I don’t know much about resistors and voltage dividers.  I do know resistors will lower voltage, but I wouldn’t know how to/where to create a voltage divider if I need to.

The resistors are probably to give each component a different brightness.  The data sheet says they're open drain, so it's possible to short them all together.  But the end result is that they would all have the same brightness.  Also, the sync pulse does need proper voltage output.

 

They are also nice enough (at least in the sample schematic) to use an amplifier on the output.  That part in the lower left corner just before the modulator is exactly what the composite mod kit is, but with appropriate resistors.  I think the first thing I'd do is see how much of that sample schematic seems to match the real board.

[eDoc]

35 minutes ago, ChildOfCv said:

Well, first it would help to know what chips are on that board besides the AY microprocessor and the 4011 over there.  I can't read the rest.

 

Also, for the modulator, I can't see the circuit board on the bottom there, but looking at the visible components I'm guessing that it's an oscillator to produce the carrier signal, and a modulator that looks a lot like the one on this page.  https://www.instructables.com/id/Amplitude-Modulator-and-Demodulator-Trainer-Kit/

 

That likely makes the yellow tuner the channel fine tune, and the white one is likely to be the modulation filter.  BTW, what channel were you attempting to watch it on?

I was on channel 3.  I tried channel 2 and 4 just to see if that would have any results, but I only ever got a static signal on channel 3.

[eDoc]

20 minutes ago, ChildOfCv said:

The resistors are probably to give each component a different brightness.  The data sheet says they're open drain, so it's possible to short them all together.  But the end result is that they would all have the same brightness.  Also, the sync pulse does need proper voltage output.

 

They are also nice enough (at least in the sample schematic) to use an amplifier on the output.  That part in the lower left corner just before the modulator is exactly what the composite mod kit is, but with appropriate resistors.  I think the first thing I'd do is see how much of that sample schematic seems to match the real board.

Ok, the chips in blue are all 4011 and the one in green reads 40138.  Well, when I closely inspect the board vs the schematic, they unfortunately don’t seem to match

The schematic even shows a channel selector for the video output which mine doesn’t have.

[ChildOfCv]

They must be using a cheaper modulator, but much of the schematic appears to match.  The CRxx's to the left of the AY chip seem to be for player inputs (see page 2 of your data sheet).  The 4011 in the upper right corner must be the main oscillator.  That would mean that tuning coil in the upper right corner changes the clock frequency into the AY.  If you had a good oscilloscope or a frequency counter that can read up to 10MHz, you could probe the output of the gate on that chip to make sure the game is running at the proper frequency.

 

The 40138 is probably a 4013B.  But at the same time, that does deviate from the schematic.

 

So do the remaining 4011's

 

But, you can still trace pins 2,3,4,18,27,28 of the AY to find out where they go, and what the resistor values they connect to are.

Edited August 2, 2019 by ChildOfCv

[eDoc]

11 minutes ago, ChildOfCv said:

They must be using a cheaper modulator, but much of the schematic appears to match.  The CRxx's to the left of the AY chip seem to be for player inputs (see page 2 of your data sheet).  The 4011 in the upper right corner must be the main oscillator.  That would mean that tuning coil in the upper right corner changes the clock frequency into the AY.  If you had a good oscilloscope or a frequency counter that can read up to 10MHz, you could probe the output of the gate on that chip to make sure the game is running at the proper frequency.

 

The 40138 is probably a 4013B.  But at the same time, that does deviate from the schematic.

 

So do the remaining 4011's

 

But, you can still trace pins 2,3,4,17,18,20,24,25,26,27,28 of the AY to find out where they go, and what the resistor values they connect to are.

Alrighty, I’ll see what I can make out by tracing these pins and examining the resistors. Thanks for all your help tonight!

[TheProgrammerIncarnate]

The Telstar Combat!'s design is a little different from the example AY-3-8700 schematic, for one it has stereo sound (on two internal speakers) instead of all sound signals going to one speaker. The reason you don't want to solder directly to the 8700 is because of those resistors, which are there to make different components different "colors" (one tank is black, one is white, background is grey.) All the resistors you need to add the signals together are already on the board though, so you just need to amplify the output (I've tested this already)

 

There is actually probably an amplification transistor already inside the modulator, problem is there is a hard-to-remove panel on the other side of the circuit board (on the left side of my picture)

 

Also I got some good pictures of the Combat! that I plan on gutting here, including it's PCB.

[ChildOfCv]

Well, I did a little bit of reverse engineering, and now I have to ask:  Is there a mode where tank 2 is invisible?  Because Q6 seems to be connected in-between tank 2's output and the collection resistors.  And the transistor seems to be turned on or off by the something around the game mode switch.

Edited August 2, 2019 by ChildOfCv

[eDoc]

6 hours ago, TheProgrammerIncarnate said:

The Telstar Combat!'s design is a little different from the example AY-3-8700 schematic, for one it has stereo sound (on two internal speakers) instead of all sound signals going to one speaker. The reason you don't want to solder directly to the 8700 is because of those resistors, which are there to make different components different "colors" (one tank is black, one is white, background is grey.) All the resistors you need to add the signals together are already on the board though, so you just need to amplify the output (I've tested this already)

 

There is actually probably an amplification transistor already inside the modulator, problem is there is a hard-to-remove panel on the other side of the circuit board (on the left side of my picture)

 

Also I got some good pictures of the Combat! that I plan on gutting here, including it's PCB.

Thanks for the photos!  That all makes sense.  So that point which you circled in blue is where the video outputs meet after going through their resistors.  I soldered a yellow A.V. cable directly to that spot and soldered ground to the Vss on the chip and nothing appeared on screen (likely due to the amplification issue).  I already desoldered that back panel when I was examining the broken RF inductor, so it’s all open.  What kind of thing am I looking for to be the amp transistor?

[eDoc]

10 hours ago, ChildOfCv said:

Well, I did a little bit of reverse engineering, and now I have to ask:  Is there a mode where tank 2 is invisible?  Because Q6 seems to be connected in-between tank 2's output and the collection resistors.  And the transistor seems to be turned on or off by the something around the game mode switch.

I believe so.  There’s a mode called night battle where it looks like the black tank is hidden.

[TheProgrammerIncarnate]

7 hours ago, eDoc said:

Thanks for the photos!  That all makes sense.  So that point which you circled in blue is where the video outputs meet after going through their resistors.  I soldered a yellow A.V. cable directly to that spot and soldered ground to the Vss on the chip and nothing appeared on screen (likely due to the amplification issue).  I already desoldered that back panel when I was examining the broken RF inductor, so it’s all open.  What kind of thing am I looking for to be the amp transistor?

The amplification transistor *should* be the first thing after the signal comes in. Also, could you get a picture of the circuit board hidden under the back panel? I never got mine removed.

[ChildOfCv]

Ah, so tank #2 is black.  That explains why it's wired so differently.

 

I'd need some more reverse engineering, but my current theory is that the transistor Q6 provides the required DC input voltage to generate the white components.  That DC voltage is pulled down by the black tank when it's being displayed.  It's strengthened by explosion sound output.  Varying levels of whiteness are provided by the other inputs.  I do wonder what pin is supposed to be responsible for the mines though.

 

I don't think there's going to be an amplifier on the modulator.  Remember that it's sending its signal to a TV that expects antenna-level inputs, ie microvolts.  Even a 1V input would look to the TV antenna port like a lightning strike in comparison.  So you really don't need amplification.  On the contrary, the signal must be weakened before it reaches the TV.

 

But for composite, you need the 1V output.  What I'd do is look at the signal wherever it's finished building, and attach it to a Darlington transistor voltage follower.  That should allow the transistor to leech off the existing signal without interfering with normal operation.  But this depends on getting proper voltage swing from the summing circuit.  It would be easier to measure those levels with a scope than to fully analyze the circuit, though.

[TheProgrammerIncarnate]

10 minutes ago, ChildOfCv said:

I don't think there's going to be an amplifier on the modulator.  Remember that it's sending its signal to a TV that expects antenna-level inputs, ie microvolts.  Even a 1V input would look to the TV antenna port like a lightning strike in comparison.  So you really don't need amplification.  On the contrary, the signal must be weakened before it reaches the TV. 

Hmm the example schematic shows a (single) transistor amplifying the composite signal before it enters the modulator chip (again, only an example.) It may be possible that there's some signal throttling on the other side of the modulator, or that it goes into the TV without weakening.

 

16 minutes ago, ChildOfCv said:

I'd need some more reverse engineering, but my current theory is that the transistor Q6 provides the required DC input voltage to generate the white components.  That DC voltage is pulled down by the black tank when it's being displayed.  It's strengthened by explosion sound output.  Varying levels of whiteness are provided by the other inputs.  I do wonder what pin is supposed to be responsible for the mines though. 

The mines are shared by the right(?) player out pin (black.) Also, based on my analysis of other PC50x chips, it's likely that the background signal is always active except when another signal (sync, left player, right player, colorburst) is. So up to one video output pin is active at a time. I know for a fact that the video output pins in other PC50x games can be either grounded or floating (no-pullup resistors.)

 

26 minutes ago, ChildOfCv said:

But for composite, you need the 1V output.  What I'd do is look at the signal wherever it's finished building, and attach it to a Darlington transistor voltage follower.  That should allow the transistor to leech off the existing signal without interfering with normal operation.  But this depends on getting proper voltage swing from the summing circuit.  It would be easier to measure those levels with a scope than to fully analyze the circuit, though. 

 I managed to get the picture below with only a single transistor attached, although the voltage divider wasn't perfected. I never did measure the voltage swing though.

 

[ChildOfCv]

One possibility related to output is that you may not even need a voltage divider.  If the voltage never falls below about 1V, and if it has a 1V swing, you can direct-couple it to the voltage follower with a couple hundred ohms on the emitter.  Using a simple transistor like in the Atari composite mod kits gives you about a 20K load through the transistor with that setup.  Using a Darlington pair will shrink that load to a meg-ohm or so.  Then you just need the 68-ohm series resistor and a coupling capacitor to the composite plug.

[ChildOfCv]

14 minutes ago, TheProgrammerIncarnate said:

Hmm the example schematic shows a (single) transistor amplifying the composite signal before it enters the modulator chip (again, only an example.) It may be possible that there's some signal throttling on the other side of the modulator, or that it goes into the TV without weakening.

It's also using a different modulator that seems to be a bit more modern, from pictures I've seen of it.  The transistor for that circuit guarantees a certain voltage swing into whatever RF modulator you choose, while otherwise you'd need to redesign the resistor ladder to match the modulator that you plugged it into.  So the sample design is more tolerant of changes.  But Coleco probably didn't want to pay for anything that they couldn't just build themselves, so they used the simplest modulator circuit they could and designed the rest of the video output around it.

 

You still see that attitude somewhat in the Colecovision, where they did go ahead and use an LM1889, but otherwise still designed the RF output themselves.  Maybe someone at Coleco hated Astec

 

Is your picture above in "night mode" or is that the remaining issue you're trying to solve?