Colecovision Black and White Image. Current Progress to rectify the issue.

[Gaetznes]

A friend dropped off a Colecovision that had known issues. As the title suggest, the video is B/W. Which I have not dealt with before.

 

After hooking it up on a know working and tested power supply. I began my work.

Normal things like testing voltage a different key points at and after the switch. All is well.

 

What I have done so far.

 

1. Installed composite mod test rig, in attempt that this may bypass the issue.

Result: No Change, Same b/w issue as on the rf line, using a F connector.

 

2. Replaced the TMS9928a from another colecovision and had socketed that ic on the mobo.

Though this TMS9928a came from the mobo parts pile. All the traces had lifted. So its unknown if it was fully working or not.

Result: No change.

 

3. Replaced, the top half of the RF unit. Again coming from the same mobo as the TMS9928a.

Result: No Change.

 

4. Replaced all the aluminium electrolytic capacitors on the mobo.

Result: B/W image is sharper and cleaner.

 

5. Tested Both Crystals (One from the mobo i'm trying to repair, and the one from the parts mobo).

Reference sources say I should have 7.15909Mhz at the crystal, and 3.579Mhz on pin 6 of J4.

Result: Both did not give the proper reading and were in fact fluctuating or blanking out. Tested at the crystal.

 

Currently waiting on an order of replacement Crystals. Will update as progress is made.

 

Notes:

Both revisions of the colecovision rf unit have a coil where a trim pot would be.

That would be normally recommended to adjust for rf adjustment. As suggested by console5

 

Inferring some of my information based on info derived from http://www.kernelcrash.com/blog/recreating-the-colecovision/2016/01/27/

and https://console5.com/wiki/Colecovision#Capacitor_List

[fiddlepaddle]

Based on all the work you've done, this might seem like a stupid thing to suggest, but are you using a CRT? and have you eliminated any issues with the tuner? and the connector (ie, better not to use a switch box)?

[Gaetznes]

I'm using a Commodore 1702 monitor. Which I use when I do these types of composite mods. Fits nice on the test bench. So that rules out a potential tuner issues, I assume.

 

Since posting, I convinced myself to open my personal colecovision. Which I composited modded the same way I did this one.

 

I only swapped out the crystal. Sadly this did not change anything in regards to the white and black screen.

I tested my machine before and after removing and re-soldering the crystal. It has color on my commodore 1702.

 

I avoid switch boxes as much as I can, and haven't been used at any point during the testing.

 

I'm at the point, where I'm not sure what else to check. Taking into consideration the TMS9928a may still be the issue.

Additionally, I have not looked into the rf circuit, and what those two ic's do if anything, in regards to color.

 

I have already talked to the owner about the issue and he has already picked up another donor system. So I hope to get back to this as soon as I have it.

That or i'm going to get bored and risk my personal coleco and remove the TMS9928a.

Then I'll be able to eliminate or determine if the TMS9928a is in fact the issue.

 

Console5 has a note on the page I referenced in post 1. That I missed earlier. Its not TI, but I do have spare 74ls04 logic laying around. It wont hurt to try.

Wrong frequency 3.579 Mhz clock is the wrong frequency. If U22 is a Texas Instruments IC, replace it. Edited January 29, 2018 by Gaetznes

[Gaetznes]

So I caved and de-soldered the TMS9928a out of my personal console. As well, I tested the 74ls04 logic ic.

 

Current developments

 

1.Tested the 74ls04.

Result: Ic tested working

 

2.Tested my now pulled TMS9928a in the working console.

Result: Happily it still works as photographed below. Though there is graphic glitches in the b/w console, not pictured.

I'll have to clean the pins, to see if I can resolve that. If not I may need to check the continuity of each pin in the socket to its destination. On the b/w console.

 

3. Second TMS9928a

Results in working personal console: Pictured below. Blue in color, a negative image cross hatch?. I'll have to investigate this.

Results in B/W console. I still have clean Colecovision Logo, game play, etc.

 

 

4. Third TMS9928a no longer will work in the b/w console nor did it ever work in my personal.

 

 

To summarize.

I'm going to order a replacement tms9928a, as i'm not content on having this is the only one on hand.

I need to investigate the negative blue hatch pattern. See if anyone else has had this problem.

A moments ago, I found this http://www.colecovisionzone.com/page/coleco%20industries/tech.html , it's a service manual. With troubleshooting. Though I don't own an oscilloscope, sigh.

Even with out a scope. I hope with that and a set of schematics, to narrow it down. Either to another ic, transistor or what ever in the rf unit. Even if it means I start swapping more parts.

Extremely rough way of doing it, but hey, at least i'm learning.

[Zak]

hello Gaetznes, what happened in the end with your CV? were you able to find the component that was causing the color to fail? I am in the same problem. Cheers

[INTVCruise]

On 8/7/2022 at 3:32 PM, Zak said:

hello Gaetznes, what happened in the end with your CV? were you able to find the component that was causing the color to fail? I am in the same problem. Cheers

I'm pretty sure I had the color go out on one of mine that I was repairing. From what I can remember I just retouched broken solder joints inside of the RF box. I'm sure if you look inside there will be a couple that are obvious and reflowing should be enough. 

[Zak]

42 minutes ago, INTVCruise said:

I'm pretty sure I had the color go out on one of mine that I was repairing. From what I can remember I just retouched broken solder joints inside of the RF box. I'm sure if you look inside there will be a couple that are obvious and reflowing should be enough. 

Hello, thanks for the answer. In my case, I replaced the entire modulator and the problem continued, so I rule out that it is some fault in the modulator. Cheers

[Madmanguruman]

I had a similar issue on one of my three NTSC CVs. I also found that the primary clock was not stable at / around 3.579545MHz. Perfect black and white picture for both CV and EM#1.

 

First, I was able to improve the jitter by replacing C1 and C70 - C1=2.2pF, C70=100pF . The 'center' frequency was still running a bit on the low side though. (I have a digital scope with statistical functions, which aided this immensely).

 

I then incrementally reduced the value of C70 to nudge the center frequency higher; 47pF, then 33pF, 22pF and finally 10pF. With 10pF, color was restored.

 

I suggest playing with the value of C70.

[Zak]

I was finally able to solve the problem, I leave the thread here for you to see in detail, the cause of the black and white image was U30. (CV PAL). https://forums.atariage.com/topic/339036-colecovision-pal-fix-no-colors/