Jump to content
Falonn

TMS-RGB: An RGB Mod for 2020 and Beyond

Recommended Posts

Anyone waiting to order from me should order from Mobius. I have decided to not produce the V2 board and instead focus on the V3 design. The V2 board is absolutely spot on, and I'm sure the Mobius parts are of very high quality.
 

Share this post


Link to post
Share on other sites

List updated.  (You'll need to hard-refresh if you've looked at it in the last 24 hours.)

 

Otherwise, please give your offshoot project a different name to avoid confusion.  "v3" would mean something specific for this GitHub repo's release list.  Maybe TMS-Pix instead? :D

Share this post


Link to post
Share on other sites

I installed this mod tonight. It's great to have my Colecovision working again (after a previous RGB mod that died on me).

 

I am having one issue. It looks like there is too much red in my video, causing blues to look purplish, etc. I noticed this both going through my OSSC as well as playing on my CRT (though it doesn't seem as bad on my CRT).

 

 

Smurf Rescue.jpg

Share this post


Link to post
Share on other sites

How curious!  Was this a board from Mobius or did you build it from scratch?  The first thing to double-check is that the resistors in this section of the board all have the values they're supposed to:

 

color-resistors.thumb.jpg.65db5f7fcf03a9d6745175652196281e.jpg

 

If any of them got swapped around, it could cause different level colors like that.

Share this post


Link to post
Share on other sites
20 hours ago, Falonn said:

How curious!  Was this a board from Mobius or did you build it from scratch?  The first thing to double-check is that the resistors in this section of the board all have the values they're supposed to:

 

color-resistors.thumb.jpg.65db5f7fcf03a9d6745175652196281e.jpg

 

If any of them got swapped around, it could cause different level colors like that.

My board is from Mobius. I wonder if the issue has anything to do with the previous RGB mod.

 

Do you think that adding a resistor on the red output might help? I know almost nothing about electronics, so forgive me if this is a not-so-smart question. :)

 

EDIT:

 

Okay, I played my Colecovisions again tonight, and I was having the same issue with there being too much red in the image. And then, somehow, that went away, and now all of the colors look correct.

 

Is there any way this could be power supply or capacitor related? I reiterate, I know almost nothing about electronics, but I do know that this system has never been recapped, and I'm still using the original power supply.

Share this post


Link to post
Share on other sites
22 hours ago, chicgamer said:

My board is from Mobius. I wonder if the issue has anything to do with the previous RGB mod.

 

Do you think that adding a resistor on the red output might help? I know almost nothing about electronics, so forgive me if this is a not-so-smart question. :)

 

EDIT:

 

Okay, I played my Colecovisions again tonight, and I was having the same issue with there being too much red in the image. And then, somehow, that went away, and now all of the colors look correct.

 

Is there any way this could be power supply or capacitor related? I reiterate, I know almost nothing about electronics, but I do know that this system has never been recapped, and I'm still using the original power supply.

Hi there.  If the colors fluctuate between slightly off and then normal again then it's some sort of power issues from three possible issues.  Dirty power switch, loose power connection at the rear of the console (you can test this by wiggling the power plug and see if colors or graphics change) or you need a full recap.  Recapping a coleco is fairly simple.  There's three 10uf caps inside the RF module, 3 10uf caps on the board and either a 10 uf or 1uf on the reset line.  If you have a 10uf on the reset line and you're not using an F18a mod then make sure to swap it to a 1uf.  You can purchase a Coleco recap kit on console5.com.

  • Like 1

Share this post


Link to post
Share on other sites
25 minutes ago, Ruggers Customs said:

If you have a 10uf on the reset line and you're not using an F18a mod then make sure to swap it to a 1uf.

What's the reason for this?  I can't recall having seen this recommended before, i.e. I can only recall having seen the recommendation to swap the original 1uF to a 10uF in all cases.  Even console5.com states that later CV revs may have 10uF factory installed.  I'm just curious as I have a few CVs where I've changed from 1uF to 10uF (no F18A) and have seen no issues.

Share this post


Link to post
Share on other sites
3 hours ago, Ikrananka said:

What's the reason for this?  I can't recall having seen this recommended before, i.e. I can only recall having seen the recommendation to swap the original 1uF to a 10uF in all cases.  Even console5.com states that later CV revs may have 10uF factory installed.  I'm just curious as I have a few CVs where I've changed from 1uF to 10uF (no F18A) and have seen no issues.

Well I mean it's not going to hurt anything.  It's more of an annoyance really.  When you put a 10uf on the reset line it can cause a boot delay with console sounds which to me is just irritating.  Putting in a 1uf eliminates this delay.  The reason they put them in some later revisions is because that's probably what they had laying around in the shop.  After a very detailed video showing C5 why you don't want a 10uf on the reset line and what it does, he revised some of his info.  The 10uf delay is useful in helping the F18 with booting issues which is why it's recommended in that setup.

 

Share this post


Link to post
Share on other sites
2 hours ago, Ruggers Customs said:

Well I mean it's not going to hurt anything.  It's more of an annoyance really.  When you put a 10uf on the reset line it can cause a boot delay with console sounds which to me is just irritating.  Putting in a 1uf eliminates this delay.  The reason they put them in some later revisions is because that's probably what they had laying around in the shop.  After a very detailed video showing C5 why you don't want a 10uf on the reset line and what it does, he revised some of his info.  The 10uf delay is useful in helping the F18 with booting issues which is why it's recommended in that setup.

 

This is great to know - thanks for the clarification.  Is the video available to watch somewhere?  I'd really like to understand more about the details.  I'm trying to make a concerted effort to learn and understand more about the CV hardware so this would really help.

Edited by Ikrananka

Share this post


Link to post
Share on other sites

Got mine installed last night into my CV. I'm quite pleased with the results! I might have to add it to my list of services to offer after all.

 

 

CV_mainboard_Overview.png

 

CV_RGB_Detail.png

 

CV_RGB_Out_Detail1.png

 

CV_RGB_Sid2.png

 

Fathom_title_RGB.png

Edited by -^CrossBow^-
Spelling errors
  • Like 2

Share this post


Link to post
Share on other sites

Hey guys, I’ve gotten the RGB board installed on my TI 99 after replacing the 9918 chip with 9928 chip, but I have this stupid B&W screen. :(

 

B2009144-104E-45DA-B7A8-2A60F5804C97.jpeg

Edited by Retro_Game_Lover96

Share this post


Link to post
Share on other sites
7 hours ago, aleomark said:

Its possible to use it on a sega SG-1000?

 

the original SG-1000 uses a TMS9918ANL, I would like to add rgb to my system if possible.

 


tms-rgb web shows:

This small adapter board can unlock the potential of many 8-bit systems from the 1980's including the Colecovision, MSX, SG-1000...”

 

You can if you swap the TMS9918 with a TMS9928 and add a crystal oscillator circuit for the system clock signal. You absolutely need that last part since the TMS9918 generates the clock signal for the entire system and without it, it will be totally non-functional.

Share this post


Link to post
Share on other sites
15 hours ago, aleomark said:

tms-rgb web shows:

This small adapter board can unlock the potential of many 8-bit systems from the 1980's including the Colecovision, MSX, SG-1000...”

I tried my best to find a list of systems that use the 9928A, but there is a lot of unreliable/conflicting information out there!  I've already had to amend that list once.  Does this mean I got the SG-1000 wrong, too, and that it should be removed from the list?

 

7 hours ago, ApolloBoy said:

You can if you swap the TMS9918 with a TMS9928 and add a crystal oscillator circuit...

For composite (or RF) out, you'd need an extra one, but just for the VDP's general operation, I think the datasheet shows that they both use the same 10.738635 MHz clock.  The only differences I see between the two are on three pins:

 

pin 35: EXTVDP on the 9918A / B-Y on the 9928A.

pin 36: composite video output COMVID on the 9918A / Y on the 9928A.

pin 38: the "typically not used" CPUCLK on the 9918A / R-Y on the 9928A.

 

It would be easy to just leave those three 9928A pins disconnected from the rest of the original circuit board by leaving them bent up or by cutting a trace on the board or something like that, except there is one key requirement laid out in the datasheet: they require a small "load circuit".

 

For pin 36, this should already be present, because it's required for both chips:

 

1865579076_9918Aloadcircuit.thumb.png.e72d3a6940266f28a6eeb7ceb424ee5c.png

 

For pin 35 and 38... who knows what circuitry is already there.  But there needs to be a resistor to ground on each:

 

1122297018_9928Aloadcircuit.png.71f36cc42052d12f1c08f8c6569303d8.png

 

If there is much else going to/from those pins, I'd worry that whatever the TI-99 board has in place for EXTVDP and CPUCLK, it might be causing the trouble you're seeing on your TV screen there.

Share this post


Link to post
Share on other sites
2 hours ago, Falonn said:

It would be easy to just leave those three 9928A pins disconnected from the rest of the original circuit board by leaving them bent up or by cutting a trace on the board or something like that, except there is one key requirement laid out in the datasheet: they require a small "load circuit".

Are you sure it will work?

Share this post


Link to post
Share on other sites

No, not at all. :D

 

I was able to find a 99/4A schematic though (on page 20 of this PDF) and it looks promising that you might not have to modify the board beyond adding/removing a couple jumpers.  It looks like it might have been designed flexibly to accept either VDP.  The J201 header is even labeled with B-Y and R-Y.  The main trouble I see is that the CPUCLK output line (which isn't a clock at all on the 9928A) appears to actually be used somewhere else on the board for something.

 

It might be worth trying to find an account from someone that has already done the 9918-->9928 conversion (with instructions for changing the jumpers) before trying to RGB mod it.

Share this post


Link to post
Share on other sites

In case anyone was concerned about stock, I still have a pretty healthy amount left from the first batch and have a second batch in production already. I expect to still have boards from batch 1 by the time batch 2 arrives.

  • Like 1

Share this post


Link to post
Share on other sites
8 hours ago, Falonn said:

For composite (or RF) out, you'd need an extra one, but just for the VDP's general operation, I think the datasheet shows that they both use the same 10.738635 MHz clock.  The only differences I see between the two are on three pins:

 

pin 35: EXTVDP on the 9918A / B-Y on the 9928A.

pin 36: composite video output COMVID on the 9918A / Y on the 9928A.

pin 38: the "typically not used" CPUCLK on the 9918A / R-Y on the 9928A.

The "typically not used" pin is the one that Apollo Boy is talking about.  I guess the SG-1000 uses it, so you need an alternate clock source after destroying the trace from pin 38 to the clock circuit.

  • Like 1

Share this post


Link to post
Share on other sites
6 hours ago, Falonn said:

No, not at all. :D

 

I was able to find a 99/4A schematic though (on page 20 of this PDF) and it looks promising that you might not have to modify the board beyond adding/removing a couple jumpers.  It looks like it might have been designed flexibly to accept either VDP.  The J201 header is even labeled with B-Y and R-Y.  The main trouble I see is that the CPUCLK output line (which isn't a clock at all on the 9928A) appears to actually be used somewhere else on the board for something.

 

It might be worth trying to find an account from someone that has already done the 9918-->9928 conversion (with instructions for changing the jumpers) before trying to RGB mod it.

Wow, I bow to your Google Fu.

 

According to the schematic, just make sure W3, W5, and W6 jumpers are not there.  The notes say that pins 2 and 5 of the video plug are jumpered to ground for US operation, but that has to be a misprint.  Pin 5 is composite out.  Pins 2 and 4, however, are R-Y and B-Y, so that's probably what they mean.  The TMS side really shouldn't be connected to anything for US (9918) use, but maybe someone jumped those pins to ground too?

Where do you see the CPUCLK going, other than to video pin 2?

Share this post


Link to post
Share on other sites
6 minutes ago, ChildOfCv said:

Where do you see the CPUCLK going, other than to video pin 2?

There are two pin 38's coming out of the chip.  On the left side, pin 38 goes to the SGCCK signal (through grounded shield W104).  On page 23, there is an "SGC CLK" at the far right that also passes through a grounded shield named W104, so I'm guessing they just changed signal names (accidentally?) between pages.  (All of these 1980's TI datasheet are riddled with errors, so it wouldn't surprise me.)  From there it appears to go to the audio chip and (through another jumper) to some "GROM" stuff.

Share this post


Link to post
Share on other sites
42 minutes ago, Falonn said:

There are two pin 38's coming out of the chip.  On the left side, pin 38 goes to the SGCCK signal (through grounded shield W104).  On page 23, there is an "SGC CLK" at the far right that also passes through a grounded shield named W104, so I'm guessing they just changed signal names (accidentally?) between pages.  (All of these 1980's TI datasheet are riddled with errors, so it wouldn't surprise me.)  From there it appears to go to the audio chip and (through another jumper) to some "GROM" stuff.

Oh, I see.  I was looking at page 28.  But apparently that's for a TI99/4QI model that I've never even heard of.

 

So W201 will be blank, W202 will be jumpered, W203 blank, W204 jumpered, W206 blank.

 

Since they were kind enough to provide source and destination tags on each line that enters or leaves a page, it's confirmed that the CPUCLK output goes to the 9919 audio chip.  So if you replace the 9918 with a 9928 and don't provide a replacement clock, you lose audio.  You'd have to get that other audio chip mentioned and then disconnect the coax and connect W104 instead.

So, you probably need to play with the jumpers too, after replacing the 9918 chip.  Move W202 to W201.  Move W204 to W203.  Add W206.  This provides the required resistor loading for the 9928 chip, which the TMS-RGB expects to already be there.

Edited by ChildOfCv

Share this post


Link to post
Share on other sites

Oh, I hadn't even gotten that far in the document.  Interesting.

 

I also just spotted the footnote on page 23 that confirms my hypothesis: "SN76489 uses 3.58 MHz clock supplied thru W104 coax".  So the TI-99/4A's sound chip definitely uses the VDP's clock out signal.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...