Jump to content

Photo

Post-"Ben Heck Mod" Issues

ColecoVision Ben Heck mod modification composite AV picture delay stray sprites

53 replies to this topic

#26 mobiusstriptech OFFLINE  

mobiusstriptech

    Chopper Commander

  • 192 posts
  • Location:Ohio

Posted Sat Jun 9, 2018 8:34 AM

I agree that to start you should remove the mod board from the equation. These consoles have so many potential issues, it's better to start with the basics.

Also I haven't seen yurkies board personally so I can't comment on it. I have 2 versions of my board on OSHPark both are designed to be extremely small and use SMD components with tight tolerances.

#27 Bmack36 OFFLINE  

Bmack36

    Dragonstomper

  • 610 posts

Posted Sat Jun 9, 2018 8:46 AM

I'll post the bom  to go with this later,  if anyone wants it. 

 

https://oshpark.com/profiles/Bmack36

 

 

https://colecovision...tegory=20695415


Edited by Bmack36, Sat Jun 9, 2018 8:51 AM.


#28 mobiusstriptech OFFLINE  

mobiusstriptech

    Chopper Commander

  • 192 posts
  • Location:Ohio

Posted Sat Jun 9, 2018 5:38 PM

You should put a cap on your audio lines. It will help clean up any noise that gets picked up in the wires or just from the poor filtering in the Colecovision.

#29 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jun 10, 2018 11:44 AM

@NIAD,

How would I go about testing the VDP's voltage? If I'm reading the pinout PDF correctly, pin 33 is +5v in, and pin 36 goes out to the RF? Where would be the best location to ground the negative probe when testing? Either I'm not testing from the correct ground, or I'm getting nothing for a reason. (Yes, I'm now testing sans-mod.) Note that I am still getting good readings on the RF board, including +5 at the original RCA output port.

I'm about to test Donkey Kong again in a few minutes with the original RF hookup (RCA to coax adapter); I'll let you know how that goes.

#30 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jun 10, 2018 12:14 PM

Okay, with the original RF out, I have a BIOS, a working menu, and a working game (I was easily able to complete the first girders stage. However, there was heavy screen noise/"snow", the picture was B&W unless I physically touched the RF shielding (the nose went down about 5% too), and there was only static audio, no game noise (when touching the main console shielding, the static noise went away but still no game audio.

I hope this means we're getting closer to a solution! Staying positive. Emotionally, this is an important piece to my collection. CV was the first console my family owned, when I was about 2 or 3.

#31 mobiusstriptech OFFLINE  

mobiusstriptech

    Chopper Commander

  • 192 posts
  • Location:Ohio

Posted Sun Jun 10, 2018 1:09 PM

So that tells us that the console is working. Which means that either the retrofixes board is messed up or there was an issue in your install of it.

Heavy snow in the RF is pretty normal. These consoles have some of the worst RF output you will find.

#32 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jun 10, 2018 3:08 PM

@mobius: Yes, it's almost definitely looking like a bad RF at this pointboth with audio and video. Here's what's I've gathered so far (and these are not yet conclusions, by any means):

1. I have a very suspect RF modulator, though it is getting the needed voltage all the way to the RCA-out port.

2. Because the RF is the final step in the sequence before going out to TV, I see this as a positive sign regarding the overall health of the rest of the console.

3. At the very, very least, I should try reballing and circuit-cleaning the whole RF board. Physical skin contact with the shielding results in a color picture and static noise, which tells me something's gone cold, dirty, or both.

4. It appears I have a faulty LM318 chip on my composite mod. Wesley from Retrofixes is currently working to remedy that issue, and I'm very, very grateful for that. Consider this an unpaid endorsement of his store (he sells much more than mod's). Wesley, if you're reading this: thank you.

5. The audio, when tapped from either C88 or the RF board, is crisp and clear, no static. This suggests interference is occurring at some point before the signal leaves the original RCA on the RF.

6. Up to this point, the Ben Heck mod gave me the best picture (with a long delay before fading in the screen from black), and it did reveal a possible need for new VRAM (I read up on replacing it with 4516/4164 chips and the other necessary steps), if you recall the OP with those few garbage pixels on the screen.

7. It looks like inserting that loose cartridge board into the slot in reverse two years ago didn't spell the end for my console, after all. Just don't do it again, Me.

#33 NIAD OFFLINE  

NIAD

    Quadrunner

  • 6,165 posts
  • Location:Chicago Suburb

Posted Sun Jun 10, 2018 9:07 PM

BTW... are you using an original Coleco RF cable. If you are using a newer RF cable, you will not get good results.



#34 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jun 10, 2018 9:14 PM

@NIAD: No; when I purchased it (an "as is" listing on eBay), the RF cable it came with was frayed and destroyed. I believe I got the one I'm using from a Radio Shack (shortly before every brick & mortar location vanished from the Earth). Are there any suitable replacement RCA cables I can buy to use for RF purposes? I suppose I won't need it, eventually, as I do plan on going full-time composite, but having the proper cable as backup wouldn't hurt.

#35 NIAD OFFLINE  

NIAD

    Quadrunner

  • 6,165 posts
  • Location:Chicago Suburb

Posted Sun Jun 10, 2018 9:25 PM

There should be plenty of sources for a Coleco RF cable such on eBat such as the Atari260.com store or send ADAMCON a PM. or any RF cable from that early 80s era.



#36 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jun 10, 2018 9:28 PM

Okay, I'll order one soon-ish and let you know how it works out. In the meantime, I'll have to wait for the new, working LM318 mod to come in.

#37 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Mon Jun 11, 2018 8:17 PM

So sometimes, my impatience pays off.

I did some research on what makes an old-school RF what it is, and also why automatic switch boxes for the NES and later generations won't work for earlier systems (not enough voltage to trigger the switch). I then came across this thread-- http://atariage.com/...m-system-to-tv/-- and had a mischievous little thought in my head.

I went rummaging for whatever RF cables I had in my little "junk tub," and I found a Sega Auto RF switch that I'm never going to use. Perfect. The question in my head was: what if I just bypassed the switch?

I took the case apart and, lo and behold, the RF out from console is indeed a shielded coax (with the negative conveniently held back by a tiny clamp. The pictures tell most of the story, but where I was most fortunate was when I grabbed a hacked-apart set of composite cables and noticed the video cable... was a shielded one! This made using the RCA connector that much easier.

I said "easier"; I didn't say "pretty." You can see for yourself: in the words of Jim Ross, it's "bowling shoe ugly."

But the darndest thing...I hooked it up after spraying the RF board with contact cleaner, adjusted the pots in the RF, and there it was: a vibrantly colorful picture with clear sound. My chop shop job worked. It's not something I would use full-time, but I damn-sure now have a capable RF cord for testing early-gen consoles.

Two things:
1. Those stray square sprites still appear, in the same spots. I'm thinking it's a VRAM issue, so I'll likely be doing the chip replacement in the near future.
2. I'm going to try 5-11Under's version of the Ben Heck mod. At the very least, it'll hold me over until I can get the lm318 replacement.

Anyway, thank you ALL so much for your help up to this point. I've learned a ton this past week.

Attached Thumbnails

  • 2A9B755B-56F2-4758-90FF-278D618FA78B.jpeg
  • 5924E44D-5F29-4198-9A9B-CDB85C607EF7.jpeg
  • 3A505B5C-01BC-4DB7-B262-6B7FF4F4CB05.jpeg
  • 5ABF1D00-C6DF-4E62-B3CE-CAAFFF058E60.jpeg


#38 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Thu Jun 28, 2018 8:03 PM

Update:

The new LM318 came in this week (Retrofixes was kind enough to send a whole new PCB and even put the IC into a socket for my convenience), as did my VRAM replacement kit from Console5.

Thankfully, I was right about the LM318 being faulty, as this new PCB worked wonderfully. Crisp, clear picture and sound.

...For a minute or less. Two new problems have arisen:

1. It looks like I lifted a soldering pad or trace on one of the VRAM vias, because I'm getting scrambled graphics (worse than before). By looking at the repair manual, it appears to be U12. I have yet to open everything back up and inspect/repair, and also I'm out of soldering braid for a few days.

2. This is the one I'm gonna need your help with. Shortly after my previous post, I was playing some of my games when, after a few minutes, faint diagonal lines appeared and the screen began to slightly distort, as if an invisible hand was pulling the picture from off-screen. Gradually but steadily, the distortion got more intense; after about 8-10 seconds, the picture was unrecognizable and without any color. A soft reset did nothing; a hard reset started clear for about one second then began distorting again. To repeat: this happened BEFORE AND AFTER the composite mod was completed; it's happening with both RF and A/V. Has anyone had this problem before? What could it be?

#39 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Tue Jul 10, 2018 9:05 PM

...Pray for me.

Attached Thumbnails

  • 760B94EA-60EF-406A-AAC1-22AA51DFDACE.jpeg


#40 Bmack36 OFFLINE  

Bmack36

    Dragonstomper

  • 610 posts

Posted Tue Jul 10, 2018 9:57 PM

...Pray for me.

hate to say it, but there are a lot of lifted pads there. I imagine it is going to take a lot of work getting that going.



#41 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Wed Jul 11, 2018 6:43 AM

hate to say it, but there are a lot of lifted pads there. I imagine it is going to take a lot of work getting that going.


Thus my post. This CV WILL be rescued.

#42 Bmack36 OFFLINE  

Bmack36

    Dragonstomper

  • 610 posts

Posted Wed Jul 11, 2018 7:06 AM

Worst comes to worst you could install F18a which would bypass the ram.



#43 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jul 15, 2018 10:49 AM

I've done quite a bit of work on the CV since my last post, including hunting for cold joints, jumping those destroyed traces, re-socketing the VRAM vias, and even replacing all of the 0.1uF caps surrounding the VRAM (they were pretty beat up, and it was a "just in case" measure). I'll post more about that later; the visual is hilarious to me.

Good news: the gradual video degradation is fixed. No more picture getting worse and worse.

Bad news: VRAM is still a problem area. Oddly enough, every screen is blue. BIOS, select screen, and actual game (runs fine, sounds great, but the image is all wrong. It's not pixels in a blender, per se, but I can barely make out what things are supposed to be. And everything's so BLUE.

Now I have some questions:

- What might it mean if I'm getting almost no voltage (about 0.2V) to any of the sockets? I tested each one at pin 8 & 9 with the negative lead on pin 16. Power is fine elsewhere. Yes, I've removed L2 and L3 and soldered the left L3 pad to C42.

EDIT: it means I'm a dummy and forgot to clean the flux off of the joints. I'm now getting the right voltage readings on each socket. Clean your flux, kids.

- Has anyone here ever experienced a blue screen from bad VRAM? If so, did you happen to know which "U" (10-17) was the problem?

Attached Thumbnails

  • 7B55AFCC-CAB2-4B67-AA4D-D5D3ECD766E5.jpeg

Edited by HeadcolorsTV, Sun Jul 15, 2018 11:27 AM.


#44 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jul 15, 2018 12:06 PM

Okay, after cleaning up the flux residue and testing correct voltages on each socket, I re-inserted the ICs, hooked up and powered on again. This time I took pictures.

Slight improvement. The BIOS screen is still blue and scrambled, but the game actually looks about 10% better. It's still garbage, however, and it's unplayable because of sprite drawing; I uploaded a quick video to show you what I mean:



I see that Console5 actually provides a link to a repair wiki that includes specific continuity tests between the VRAM and VDP, so I think I'll try that next.

About the ICs, by the way: according to the wiki, I should be getting +5v on pin 8, 0 or +5v on pin 9, and 0 on pin 1. But I'm getting about 1.2v on pin 1, on all eight ICS. Is that bad, or is that OK, seeing as how pin 1 should be unused after the mod?

Attached Thumbnails

  • 5F99AB78-1B80-417F-8D1F-186435E36340.jpeg
  • 92DE02D2-E55C-486D-8206-C850A685A372.jpeg
  • 9695B685-18F5-406C-9E81-7980954B5399.jpeg


#45 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Sun Jul 15, 2018 3:53 PM

Some more improvement. After more thorough continuity testing between like pins (pin 3 to pin 3, 4 to 4, etc.), I saw that I missed a couple of traces to jump. The blue screen and vertical lines are gone! But the sprites, text, and colors are still mostly wrong (pics attached). Interestingly enough, when I make Mario/Jumpman jump, you can see his full, intact sprite.

So no red or purple letters in the BIOS; the girders are green. No DK sprite (mostly red), no barrel sprites, no oil can sprite, no Pauline sprite, and everything is misaligned (in DK, Mario appears on/in the second girder up; in Gorf, your ship appears in the upper area of the force field).

Doing the continuity tests in the Console5 wiki, from VRAM to VDP and back (all the buses, in other words), every single test was good. Bottom line: every continuity check now tests OK.

So now the question is: Bad VDP, or Bad VRAM chips (maybe they were damaged from previous unsuccessful attempts)?

Attached Thumbnails

  • CAA19790-4823-486B-95F6-DA57C24112A4.jpeg
  • 5D278F57-063A-4465-87C1-175500356590.jpeg
  • C1A8DA4E-1748-4982-971A-AB0B53B82375.jpeg
  • 18C12711-58F5-40C9-BF37-5738877E7BF4.jpeg


#46 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Mon Jul 16, 2018 5:22 PM

I'm just gonna leave this here for now. Good news is on the way.

Attached Thumbnails

  • 2C9B66F7-C019-401E-80CD-85382BFD33F1.jpeg


#47 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Tue Jul 17, 2018 5:13 PM

Okay, so this will probably be the last "update" post in this thread, basically becauseI've successfully fixed the CV! It's a happy day in my game room. So I'll run down the final build and the steps I took. Once again: there are pics to go with the story.

To recap (heh, no pun intended): I had a scrambled picture (I figured it was the VRAM) and a gradual RF signal loss. For the latter, I thought it might be because of a gradual power loss. After testing the power switch I already cleaned and regreased, and SCOURING the internet for someone with a similar issue, I finally found it on a very old FAQ, where a tech expert mentions cold joints somewhere in the grounding (I figured it would either be that or the +5v). Looking at the ground traces, I saw that the pegs of the RF shielding box weren't grounded. I wondered if that would matter, and...well...yes it does. The composite mod bypasses the RF out, but not before entering the box and stealing its 12v and ground from the daughter board. On my CV, it appears that a reinforced grounding was already performed in a previous repair; notice the insulated grounding strap connecting the ground trace to the outside of the RF box.

Another thing I wanted to do was replace the 0.1uF 100v caps surrounding the VRAM, JUST in case. Turns out, these are very difficult to find replacements for. Literally the only thing that i could find with the same capacitance and tolerance was a 500v rated one on mouser.com. At $1.22 apiece, they're not a cheap replacement...nor are they a small one. LOOK AT THE SIZE OF THOSE. I was worried they wouldn't clear the main shielding! But thankfully, they did...barely. Just look at the pics of them installed for a good laugh.

On to the VRAM. When all was said and done, I had to solder THIRTEEN trace jumps. And I was STILL getting the picture you saw a couple of posts above, though I was making some clear progress. Thinking I had somehow damaged one of the VRAM chips in previous attempts, I rotated the chips one spot to the left on the sockets. Identical picture. Okay, so the ICs are fine. I mentioned before that I had tested continuity from pin 1 to pin 1, etc., for each of the 16 pins on each RAM chip. But you're supposed to skip pins 2 and 14, as they go directly to the VDP and don't connect with each other. So when I saw continuity between U10:14 and U12:14, I knew I'd found something. They each go out to a testing pad before following their respective trace to the VDP. The trace coming from U10 comes awfully close to the testing pad and trace from U12. Sure enough, there was the tiniest amount of stray solder between the two. A gentle little flick of the tweezers was the last remaining fix this CV needed.

When I hooked up and powered on, I was ELATED to see a perfectly running system. No stray or garbage pixels; no slow fade-in when powering on. No more worrying about RF or bad VRAM. And no more gradual loss of video signal after a short amount of time. This is it. I've done it. And I have Console5, Wesley at RetroFixes, and you fine people to thank. We've rescued a CV. Now I can enjoy it with my children.

Attached Thumbnails

  • E482B1B8-8489-47F9-979E-88CFE5A6F4D6.jpeg
  • 391EF0A6-9745-4E9F-972E-2EBFFFC20008.jpeg
  • 9B883587-77B9-486E-82CF-FF91BF4AE3EF.jpeg
  • EB2BB178-C6CF-48D6-AEC6-E042DEC257EE.jpeg
  • 5A1BBB87-AB85-43A7-8766-3B7627F8301C.jpeg


#48 -^CrožBow^- OFFLINE  

-^CrožBow^-

    Quadrunner

  • 6,750 posts
  • Collector of Fine Atari and Sega Antiquities!
  • Location:Ivory Tower, Fantasia (Oklahoma)

Posted Tue Jul 17, 2018 8:27 PM

You have the same revision I have with the +5 RAM on it...careful if you plan to purchase Risky Rick...it doesn't work properly on my +5 RAM modded H2 revision board but does on my older F1 revision board.

 

Still, that is a ton of work and glad you got it up and going. I likely could have gotten those caps here though locally. There is a shop here that had tons of stuff like that. They don't always have the best price which is why I order my resistors and other stuff from Ebay, mouser...etc. But for older stuff that is hard to find, they will usually have it on hand and not for too much money.



#49 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Wed Jul 18, 2018 2:04 PM

*sigh*.

I hate speaking too soon. I was still having signal loss issues, and wiggling the power switch a tiny bit seemed to affect it, better or worse. So I went back in, cleaned the pins of the switch (I had forgotten to do so before, when I cleaned and greased the contacts), reballed some of the ground leads, and cleaned any residue from the grounding straps.

Plugged it back in, hooked it up...BLACK SCREEN, NO AUDIO.

So I tested voltages, and found the proper numbers everywhere (power supply, switch, ~9v audio out from RF, 12v out from RF, +5v video out from RF, 5v at pin 6 of the LM318, ~9v at the audio jacks) until I got to video out from the mod. Big fat 0.

I'm going to (ugh... AGAIN...) uninstall the mod and try RF output. If that works, I'll know it was, unfortunately, another faulty mod board.

#50 HeadcolorsTV OFFLINE  

HeadcolorsTV

    Space Invader

  • Topic Starter
  • 41 posts
  • Cannot be classified as a planet
  • Location:Buffalo, NY

Posted Wed Jul 18, 2018 3:02 PM

Nothing from RF either; testing 0 volts at the RF jack.

How was I testing the correct voltage going to the mod but nothing going to the RF output with the mod removed??? And why would there have been no audio to begin with?

In any case, I could have shorted a trace when reballing. I'll have to closely examine everything later tonight.

I will not be denied my victory lap.





Also tagged with one or more of these keywords: ColecoVision, Ben Heck, mod, modification, composite, AV, picture delay, stray sprites

0 user(s) are browsing this forum

0 members, 0 guests, 0 anonymous users