TMS-RGB-V2 board with PAL TI99/4A (TMS9929A) – wrong colors, lacking blue

[Dr.OG]

Hi everyone!

Matt Buxton advised me to ask help on this forum. I bought a TMS-RGB-V2 board recently from VideoGamePerfection.com and built it into my Texas Instruments TI99/4A. It’s a PAL version with TMS9929A IC.
I think the orientation is correct, and the soldering should be OK as well.
Previously I owned a Samsung LE40B650 LCD TV with component input, which had an acceptable picture with the TI, blue and violet colors were distinguishable. Unfortunately the Samy TV “died” last year, and my new 4K LED TV (a 50″ TCL) does have only HDMI and composite inputs, so I decided to buy a Medusa retro scandoubler. In theory it should accept component input as well, but in practice it doesn’t.
I hoped that I could get a colorful picture with the TMS-RGB-V2 using the SCART input of the Medusa. Unfortunately the picture lacks almos all blue shades.
Blue and violet colors are undistinguishable, background is gray instead of cyan/light blue. I tried with a noname (Vortex) LCD TV directly via SCART as well, but the same occurs.
I checked all the solderings with a multimeter, and everything seems to be OK.
Undusted my cheap chinese RGB/YUV to VGA converter (a Gonbes GBS-8200), which converts component input to VGA. This gives an acceptable result (distinguishable blue and violet), but the picture is washed a bit, and has wrong white balance.
Attach the TI to the Gonbes via TMS-RGB-V2’s RGB output, the picture is a mess, no matter what output resolution I choose.

The TI motherboard has a number on the top, near the keyboard connector (ET 03163)
and an another one on the bottom, near the DB-9 connector (1039026-5J).
What should I do? Is it a defective unit, is the TMS-RGB-V2 incompatible with my motherboard, or are there any tricks (like in case of NTSC Colecovision bypassing the L9) to improve color fidelity? Otherwise picture quality (sharpness and stability) is flawless.
There’s a trimmer next to the TMS9929A, I assume it adjusts the clock frequency.
Perhaps bypassing the 75Ohm resistor on the blue output line would help, but this would affect the white balance as well.

Thanks your help in advance: Gábor

PS.: I made a lot of photos, but the forum doesn’t allow me to link them (***Forbidden. message seems to be spam.***)

[BuckoA51]

You can upload photos directly to the forum, or try linking only one or two in your message.

[Dr.OG]

Thanks for your suggestions!
Here is the first pic, which shows the orientation of the daughterboard (TMS-RGB-V2) soldered to the TI mainboard’s bottom:

[Dr.OG]

The 2nd shows the screenshot taken from my previous TV’s framebuffer, TI attached directly via component cable. White balance is not perfect, but blue and purple colors are different.

[Dr.OG]

The third screenshot: TMS-RGB-V2’s RGB output via SCART. Medusa and Vortex shows the same: gray background instead of cyan / light blue, and no real blue shades (seems to be violet).

[Dr.OG]

TI’s direct component output attached to Gonbes GBS-8200, VGA scaler set to 800×600. Bad white balance, but blue colors seem to be OK on TV.

[Dr.OG]

Connecting TMS-RGB-V2’s RGB output to Gonbes GBS-8200’s RGB input the picture is a mess, no matter which resolution I choose.

[Dr.OG]

The last picture shows the mobo of my TI99/4A, with the TMS 9929A chip.

[BuckoA51]

I’ve no experience with Medusa but have you tried it with a different source to verify it’s working?

[Dr.OG]

Of course medusa is working great with other RGB sources. Tried with various retro devices, and MiST FPGA as well. Attachment shows NeoGeo core, colors are OK.

[Dr.OG]

Besides my only SCART-capable flat panel TV shows the same with TI99/4a+TMS-RGB-V2 (see the 5th post in this topic): grey background and purple squares instead of blue. Medusa shows the same via HDMI (sorry for the stripes, I didn’ adjust pixel clock properly):TMS-RGB-V2)” width=”2080″ height=”1560″ />

[TheCorfiot]

At this point you could do with the use of a scope to confirm your connections to the tms pin b-y is good and signal appears.

You can also check the B output of the board with the scope.

 

But basically you are missing Blue

[Dr.OG]

Unfortunately I don’t have an oscilloscope.
Checked the connections with a multimeter, solders seem to be fine. Actually true blue is missing, but I have i.e. magenta, which contains some blue.
I feel I’m stuck with the problem. I can’t investigate it further without special equipment (i.e. oscilloscope) which I don’t have. Besides I’m not good in SMD soldering, and I don’t want to make things worse during experimenting…
MB will send me another board, as they consider the possibility of a faulty board at this point.
If that won’t work either, I’ll send my TI99/4A to to BetaGamma in Greece for further inspection.

[Dr.OG]

Hello! The store sent me a new board that arrived on Friday. The colors are fine with the new unit, blue is no longer missing! Image quality is brilliant, the first board was faulty. So it’s all good now, thanks for the help!

[Author]

Posts