Jump to content
Falonn

TMS-RGB: An RGB Mod for 2020 and Beyond

Recommended Posts

8 minutes ago, Falonn said:

Having never seen this particular problem before, I wouldn't know where to begin, sorry.

 

If my hypothesis is true that it's more to do with a fussy, old chip than it is with the mod, you've always got the nuclear option of swapping the entire VDP out with something like the F18A.  (The last news we heard about the MK2 wasn't exactly what everyone was hoping for, but it's been a month or so since then, so hope begins to return.)

I've seen something similar.  Some consoles had sprite flicker and other sporadic graphic issues that only showed up in RGB but not composite or RF.  Replacing the VDP with another corrected the issues.  However, you won't know you have an issue with your vdp until you install the rgb board because like I said it won't show up in composite or RF.

Share this post


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

I've seen something similar.  Some consoles had sprite flicker and other sporadic graphic issues that only showed up in RGB but not composite or RF.  Replacing the VDP with another corrected the issues.  However, you won't know you have an issue with your vdp until you install the rgb board because like I said it won't show up in composite or RF.

Interesting. Given how delicate the traces are on the Colecovision, I'm a bit hesitant to desolder both the RGB mod PCB and then the VDP, but it's nice that there may be this option available. Is the VDP a standard part that I could buy from, say, Mouser, or would I have to pilfer one from another CV? Thanks.

 

Edit: How about this here?: https://www.ebay.com/itm/TI-TMS9918ANL-USA-ship/192932232032?epid=1818450792&hash=item2ceba83b60:g:wXUAAOSwXeJYECgS

 

Also, is there any reason why I couldn't install a 40 pin socket, then attach the RGB mod to its pins underneath? Just in case I have a future need to replace the TMS IC again.

Edited by Ianr757

Share this post


Link to post
Share on other sites

They haven't made new '9928A chips in decades, but they seem to be pretty readily available on eBay.  (Your link is for the 9918A which is the composite-output version of the chip instead of the YUV output, so don't buy that one.) :D

 

Otherwise: a 40-pin socket is a great idea.  I don't think the extra height would make the clearance on the heatsink much worse, so it should be safe.

Share this post


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

They haven't made new '9928A chips in decades, but they seem to be pretty readily available on eBay.  (Your link is for the 9918A which is the composite-output version of the chip instead of the YUV output, so don't buy that one.) :D

 

Otherwise: a 40-pin socket is a great idea.  I don't think the extra height would make the clearance on the heatsink much worse, so it should be safe.

Thanks for the catch!! How about this guy?

https://www.ebay.com/itm/TMS9928AN-Video-Display-Processor/222387146999?hash=item33c74e90f7:g:plUAAMXQ97RRHEk9

 

Also, is there any way to know before buying if I will have the same issue or not? Thanks.

Share this post


Link to post
Share on other sites

It's an easy mistake to make.  Most Wikipedia-type listings for these old systems mention them as having the '9918 (probably?) because the whole family of chips was lumped under the "TMS9918" banner.  You usually have to do more digging for schematics (or peek under the heat sink) to find out for sure.

 

That eBay listing looks good.  Coincidentally I purchased a set of three of the same item from the same seller about six months ago.  Mine even had the same little red paint dot added to the corners to show pin 1 for whatever reason.  As far as I could tell, they were OEM parts and not any sort of sketchy replica.

 

Otherwise, I don't think there would be any way to tell which might be free of the graphical problem.  Luck of the draw, I suspect.  (More reason for adding the socket!)

Share this post


Link to post
Share on other sites
1 hour ago, Falonn said:

It's an easy mistake to make.  Most Wikipedia-type listings for these old systems mention them as having the '9918 (probably?) because the whole family of chips was lumped under the "TMS9918" banner.  You usually have to do more digging for schematics (or peek under the heat sink) to find out for sure.

 

That eBay listing looks good.  Coincidentally I purchased a set of three of the same item from the same seller about six months ago.  Mine even had the same little red paint dot added to the corners to show pin 1 for whatever reason.  As far as I could tell, they were OEM parts and not any sort of sketchy replica.

 

Otherwise, I don't think there would be any way to tell which might be free of the graphical problem.  Luck of the draw, I suspect.  (More reason for adding the socket!)

Awesome, thanks! Any tips for safely removing the chip? I have a Hakko 301 and solder wick.

Share this post


Link to post
Share on other sites
5 hours ago, Ianr757 said:

Also, does anyone know what the heat sink is connected with? Glue, or something more nefarious?

The heatsink is attached using a thermal conductive epoxy. I have a small tube of the stuff myself that I use for attaching sinks to chips that didn't have them originally or I've had to do it with some early revisions CVs sent to me for service work that didn't have sinks on their VDPs.

 

Share this post


Link to post
Share on other sites
9 hours ago, -^CrossBow^- said:

The heatsink is attached using a thermal conductive epoxy. I have a small tube of the stuff myself that I use for attaching sinks to chips that didn't have them originally or I've had to do it with some early revisions CVs sent to me for service work that didn't have sinks on their VDPs.

 

Thanks. Could I use a hot air gun to remove it then?

Share this post


Link to post
Share on other sites

You likely could but why are wanting to remove the heatsink? So that you can apply it to a new chip? I think it would be easier and better to get a replacement heatsink for the replacement chip. I get my heatsinks for the VDPs from Console5 along with thermal epoxy they also sell in a small tube. It works really well too.

 

Heatsink:

https://console5.com/store/dip40-super-heat-sink-glue-on-thermal-epoxy-style-dip-40.html

 

Thermal epoxy:

https://console5.com/store/thermally-conductive-heat-sink-adhesive-glue-gd9980-10-grams-10g.html

 

 

Share this post


Link to post
Share on other sites
4 minutes ago, -^CrossBow^- said:

You likely could but why are wanting to remove the heatsink? So that you can apply it to a new chip? I think it would be easier and better to get a replacement heatsink for the replacement chip. I get my heatsinks for the VDPs on CVs that come for service without them from Console5 along with thermal epoxy they also sell in a small tube. It works really well too.

 

Heatsink:

https://console5.com/store/dip40-super-heat-sink-glue-on-thermal-epoxy-style-dip-40.html

 

Thermal epoxy:

https://console5.com/store/thermally-conductive-heat-sink-adhesive-glue-gd9980-10-grams-10g.html

 

 

I coincidentally just bought these heat sinks from this site in between my last post and yours, haha. Thanks for the heads up though. I basically came to the same conclusion once I saw that Console5 stocked heat sinks.

Share this post


Link to post
Share on other sites
On 1/5/2021 at 8:59 PM, Falonn said:

It's an easy mistake to make.  Most Wikipedia-type listings for these old systems mention them as having the '9918 (probably?) because the whole family of chips was lumped under the "TMS9918" banner.  You usually have to do more digging for schematics (or peek under the heat sink) to find out for sure.

 

That eBay listing looks good.  Coincidentally I purchased a set of three of the same item from the same seller about six months ago.  Mine even had the same little red paint dot added to the corners to show pin 1 for whatever reason.  As far as I could tell, they were OEM parts and not any sort of sketchy replica.

 

Otherwise, I don't think there would be any way to tell which might be free of the graphical problem.  Luck of the draw, I suspect.  (More reason for adding the socket!)

 

On 1/5/2021 at 6:59 AM, Ruggers Customs said:

I've seen something similar.  Some consoles had sprite flicker and other sporadic graphic issues that only showed up in RGB but not composite or RF.  Replacing the VDP with another corrected the issues.  However, you won't know you have an issue with your vdp until you install the rgb board because like I said it won't show up in composite or RF.

All right, I have installed a socket and received three extra TMS VDP chips. So far, all three of the new ICs exhibit the same exact flickering. I feel like it is very unlikely that the issue is, as such, a result of the VDP. Is there anything else I could try? Thanks.

Share this post


Link to post
Share on other sites
On 8/27/2020 at 1:00 AM, Ikrananka said:

You may want to hang fire on that.  The TMS-RGB actually generates component signals before it then converts them to RGB.  Falonn has kindly agreed to help guide me with cutting his TMS-RGB schematic down to a component only output design which I then plan to create a PCB design for.  Only downside is that this may be a few months away as we're working on another project at the moment which needs to be completed first.

 

Hi

 

Any progress with this ?

 

I especially like the idea of using larger components so we can hand solder.

 

 

 

Share this post


Link to post
Share on other sites

Hoping to get some educated opinions on how to fix the display from a Sord M5.

 

I have both a PAL and NTSC unit.  The PAL unit uses a TMS 9929A VDP.  I installed the TMS-RGB adapter on the underside.

INSTALL.thumb.jpg.d07da99d0681c876f7c50fe8a6c0379b.jpg

 

But the display from my Commodore 1084 is pretty awful.   The display is compressed at the top, and the left hand side of the display is underscanned, so some of the text is missing.  The colors are also off as this is supposed to be green, but I figured I'd worry about that later.

 

SPAD.thumb.jpg.bc7802bbbd4c3775ab6aefba23453bf9.jpg

 

This the same unit using composite output (the monitor is NTSC/RGB so I lose the colors).

 

2127035701_SordM5PAL.thumb.jpg.ffd654b088d6abb882fe43fa43621193.jpg

 

And this is what we are supposed to see (NTSC unit output via Composite Video)

 

698574639_SordM5NTSC.thumb.jpg.aec42e72f1d2b5b2307f6eddb1b66531.jpg

 

I figured the CSYNC wasn't working, so I tried both the TTL and even just feeding the whole composite signal to the CSYNC line, which corrected the underscan, but did not improve the top of the picture where everything is compressed and unreadable.

 

The Sord M5 is similar to an MSX or Colecovision as in that it uses a Z80 coupled to either a TMS9918A (NTSC) or TMS9929A (PAL).  I've seen the install guide at tms-rgb which suggests removing the L9 Inductor, but I don't know if the Sord has this, and this is a PAL unit anyway.

 

Hoping someone might have some suggestions on how what is causing the garbled display.  Is it the CSYNC or something else?   Any pointers on things to try?/

 

Thanks!

 

 

 

 

CSYNC.jpg

Share this post


Link to post
Share on other sites

What's the model number of your monitor?  Looks like there a number of them.

 

368806 - Luma/Chroma in separate RCA connectors, or analog RGB with composite sync in a DE9 connector

319650 - Luma/Chroma in separate RCA connectors, or analog OR digital RGBI with separate sync in a DE9 connector, with a switch to decide on digital or analog

319651 - Luma/Chroma in separate RCA connectors, analog RGB in a DIN-6, digital RGB in a DIN-8 (both RGB horiz/vertical sync separated)

363299 - Luma/Chroma in separate RCA connectors, analog or digital RGB in a 23-pin D-sub, both hsync/vsync and composite sync pins (but it appears to be a 23-pin D-sub on the monitor with a DE9 cable)

Share this post


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

What's the model number of your monitor?  Looks like there a number of them.

 

368806 - Luma/Chroma in separate RCA connectors, or analog RGB with composite sync in a DE9 connector

319650 - Luma/Chroma in separate RCA connectors, or analog OR digital RGBI with separate sync in a DE9 connector, with a switch to decide on digital or analog

319651 - Luma/Chroma in separate RCA connectors, analog RGB in a DIN-6, digital RGB in a DIN-8 (both RGB horiz/vertical sync separated)

363299 - Luma/Chroma in separate RCA connectors, analog or digital RGB in a 23-pin D-sub, both hsync/vsync and composite sync pins (but it appears to be a 23-pin D-sub on the monitor with a DE9 cable)

 

It's this one: 319650 - Luma/Chroma in separate RCA connectors, or analog OR digital RGBI with separate sync in a DE9 connector, with a switch to decide on digital or analog

 

The monitor has a DE9 for the RGB and a switch on the back to select between RGB or RGBI.   There are also RCA jacks for Chroma/Luma, and another switch to use just Chroma as Composite.

 

I just tried using RGBI mode but that looks even worse.

 

 

Share this post


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

 

It's this one: 319650 - Luma/Chroma in separate RCA connectors, or analog OR digital RGBI with separate sync in a DE9 connector, with a switch to decide on digital or analog

 

The monitor has a DE9 for the RGB and a switch on the back to select between RGB or RGBI.   There are also RCA jacks for Chroma/Luma, and another switch to use just Chroma as Composite.

 

I just tried using RGBI mode but that looks even worse.

 

 

That may be the problem then.  If it expects separate horizontal and vertical sync, the composite sync probably isn't adequate.  There is a vertical sync output on U2 pin 3, but the TMS-RGB wasn't designed to output that one.  Pin 3 of U2 needs to go through a buffer and then it can output through a 75-ohm resistor to the vertical sync input.

Share this post


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

That may be the problem then.  If it expects separate horizontal and vertical sync, the composite sync probably isn't adequate.  There is a vertical sync output on U2 pin 3, but the TMS-RGB wasn't designed to output that one.  Pin 3 of U2 needs to go through a buffer and then it can output through a 75-ohm resistor to the vertical sync input.

 

Okay, but the monitor does work with RGB/CSync inputs from my Camputers Lynx and Acorn Electron, and even the Amiga which did not use the HSYNC/VSYNC option on this display.

 

What does "go through a buffer' look like in terms of components?   Is that an easy mod?

 

 Thanks

 

Share this post


Link to post
Share on other sites
12 hours ago, oracle_jedi said:

 

Okay, but the monitor does work with RGB/CSync inputs from my Camputers Lynx and Acorn Electron, and even the Amiga which did not use the HSYNC/VSYNC option on this display.

 

What does "go through a buffer' look like in terms of components?   Is that an easy mod?

 

 Thanks

 

I don't have an answer to the previous question, but I would recommend redoing your solder work. A lot of the solder joints look partially empty, and the leads on the pads look a little loose. Not sure if this would correct your issue, but it would be a good start to rule out anything simple.

Share this post


Link to post
Share on other sites
12 hours ago, oracle_jedi said:

 

Okay, but the monitor does work with RGB/CSync inputs from my Camputers Lynx and Acorn Electron, and even the Amiga which did not use the HSYNC/VSYNC option on this display.

 

What does "go through a buffer' look like in terms of components?   Is that an easy mod?

Hmmm.  Well if it syncs properly with just hsync, then that's an undocumented feature.  The manual only says that they must be separate.

 

The circuit itself is fairly simple, but assembling it into a space that is usable is less so.  Especially since you have to solder a wire to a small IC on the TMS-RGB.  Just a small doodle, but this is the general idea (you'd need to add a separate board, or better yet if you have one, use a breadboard to prototype it).  Adjust the potentiometer until VSYNC sees the proper output swing, about 0.7V AC.
1622297992_ScreenShot2021-01-14at9_37_32PM.thumb.png.430769fc301cd635d642c4e9848d14bb.png

Edited by ChildOfCv
Third time's the charm?

Share this post


Link to post
Share on other sites
18 hours ago, Ianr757 said:

I don't have an answer to the previous question, but I would recommend redoing your solder work. A lot of the solder joints look partially empty, and the leads on the pads look a little loose. Not sure if this would correct your issue, but it would be a good start to rule out anything simple.

 

Roger that.   I'll try to check all the connections again this weekend and hope it makes it look better.  Otherwise I may have to try to add in the mod @ChildOfCv suggested, which looks challenging!

 

Share this post


Link to post
Share on other sites

I swapped the board for a second one, resoldered everything, and switched to a second cable.  

 

This is the output on a Sony PVM 14N6U.

 

Something is seriously messed up with the way the Sord does things.

20210116_193913.jpg

Share this post


Link to post
Share on other sites
1 hour ago, oracle_jedi said:

I swapped the board for a second one, resoldered everything, and switched to a second cable.  

 

This is the output on a Sony PVM 14N6U.

 

Something is seriously messed up with the way the Sord does things.

20210116_193913.jpg

That looks like a lack of horizontal sync.  Make sure your sync line is connected to the horizontal sync pin (8 on the DE9).  Also, it should not be TTL.

Share this post


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

That looks like a lack of horizontal sync.  Make sure your sync line is connected to the horizontal sync pin (8 on the DE9).  Also, it should not be TTL.

 

Hmm, okay.  The PVM has RGB/Sync through four BNC jacks.  And the monitor will report "NO SYNC" if it doesn't find a sync source, which isn't the case here.  I think I may try replacing the patch lines from the RGBS pads on the board and see if that helps. 

 

I don't know if this is relevant but the PAL Sord units have a pretty dreadful composite video quality at the best of times with a lot of noise in the image.  I can only assume the RF output hid a lot of that.   I have two PAL Sord units and they both exhibit the same behavior.

 

Thanks again for the insights.  

 

Share this post


Link to post
Share on other sites
9 hours ago, oracle_jedi said:

 

Hmm, okay.  The PVM has RGB/Sync through four BNC jacks.  And the monitor will report "NO SYNC" if it doesn't find a sync source, which isn't the case here.  I think I may try replacing the patch lines from the RGBS pads on the board and see if that helps. 

 

I don't know if this is relevant but the PAL Sord units have a pretty dreadful composite video quality at the best of times with a lot of noise in the image.  I can only assume the RF output hid a lot of that.   I have two PAL Sord units and they both exhibit the same behavior.

 

Thanks again for the insights.  

 

Oops, didn't notice about using the PVM this time.

According to the manual, external sync is expected to be 4Vp-p.  That's closer to TTL than 0.7, but TTL is 5V not 4V.

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.
Note: Your post will require moderator approval before it will be visible.

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...