Electronics > Repair
Vintage chip Programmer : " Micropross ROM 3000U "
<< < (15/38) > >>
Robert763:
Hi Vince,
Try running the GBS8200 from a 5V DC supply, even if the board says 9 or 12V. I have a feeling there were boards with PSU issues. This may be the reliability issue.
If you have a smart phone or video recorder try recording the screen and playing it back with freeze frame.

Robert.
m k:

--- Quote from: Vince on December 27, 2022, 05:16:55 pm ---From my google, all sites say CGA has separate H and V sync signals, tehy aren't not combined.

--- End quote ---

Don't hang to that, it's just a name and were those board makers even born when that standard was on stage.
More commonly the meaning was just pixels, and not even including colors of them.

Your earlier picture with JAMMA connector has RGBC connection, so it is more likely a working one.
CGA/EGA is the name of one connector yes, but the connector has only 5 pins and one is named S, how you connect CGA or EGA with that?

14kHz H-sync is so low that your signal most likely fits.
But if you connected through VGA connector and the board expects something else then you obviously have a scrambled picture.

Occasionally not working buttons can be so that when the board starts its conversion some buttons are disabled.
m k:
Robert linked a manual,
there page 6 has frequencies.
Only B has lower frequencies, C and D are the highest only.
The board has also 3 connectors for those.
If all connectors are equal then why three sets.

Page 3 has features, there are also proportional YUV resolutions missing from page 6.
Applications have also some oddities, like own lines for plain CGA and EGA.

Included booklet is v3.0.
It has RGBS for VGA connector input, not very standard.
(later pinout can have CSync through HSync, Green can also include SoG)
8-pin connector is RGBS only.
5-pin connector has EGA and CGA inputs, not very standard again.
Specifications then has lower H-freqs through 5-pin P3 only and RGBS is totally missing.
Very clear.

It's possible the board is not supporting real CGA or EGA at all.
Though it is supporting CGA and EGA resolutions but only with always same polarity composite sync.
8-pin connector has markings for separated H and V syncs but specifications is not supporting lower frequencies through it.

Earlier pictured JAMMA connector is connected to 8-pin P10 and it uses four rightmost pins from the booklet picture.
(pins are GND GND VS HS S ? G ?)
If JAMMA signal is not hi-res then booklet's specifications is inaccurate and P10 will accept lower frequencies, but how those VS and HS are treated is still unknown.
Though booklet specifically advises how separated JAMMA syncs should be connected.

Before the big chip there are so many passives that all signals can be separated.
Are VGA HS and 8-pin HS the same?
5-pin S and 8-pin S also?
Vince:
@mk : thanks for that PDF, it's yet one more notch better than the previous one. In this one we have a nice table at the bottom of page 3 which explains better what each connector can do or not do.

Thanks for trying to help with this board, I appreciate... but honestly the more you post about it, the more confused I become. It's a huge mess.
BUT..... I thought I would search Youtube Videos to see if I could find interesting about this board and sure enough I did.

It's now much clearer.  So to summarize all my findings and conclusions on this whole video situation, he goes

1) Yes sorry mk, CGA is not composite sync, just is not.

2) CGA is only 50 or 60 Hz, not 40+ Hz like my programmer is using.
Other than that the sync signals are of the same type : TTL positive pulses. Pulse width does not conform to CGA standard though, but is in the same ballpark  (single digit micro seconds)

2) That video board is NOT designed for CGA monitors ! :palm:  It's designed to work from the CGA otput of arcade games... which apparently is not real CGA. It's got composite sync signal. The board is meant to work with composite signals.

3) This board is old, well known, it's the cheap route for CGA to VGA. However what's also known is how crappy and unreliable these boards are. One guy even ordered 3 of these boards in a row and all had some kind of problem... so mine is no different.


Now for the "hope" part : the guys who like me tried to get that board to accept REAL CGA signals.. eventually managed to do it. They merged the H and V signals to make a composite sync signal (which they fed to the appropriate pin on the 5 pin connector on the edge of the board).

Thanks goodness they even said what they did to create this Csync signal : nothing more than a  XOR logic gate ! That works apparently...it gives you a picture. However guy said he then inverted the Csync signal . changed polarity, and it looked even better.  Honestly from his video, it looked just fine without inversion but hey...

So... knownig all that...  I can add the XOR gate knowing it should wrok IF I had a CGA signal. But I don't. However the main differences seems to be the slightly different Hsync frequency which the board should be able to cope with, but most importantly the Vsync frequency which is quite lower than CGA, so not sure the board can handle that. However I guess it should probably give us a better picture than what we have right now, so it's still worth a try.

@ Robert :

I adjusted the supply to 5V but still no joy, it crashed just as happily. Was worth a try...

Freeze frame : my 10 year old smart phone is not the best tool for that.. it's buggy as hell, I don't even know if it can do video (it's already useless for pictures...), and if it did I have zero storage space left to record a video.  BUT... well I have already made plenty of video clips with my still camera so I used that. I recorded 10 minutes of the board cycling, so I could get a handful of "events", hoping some of them would be exploitable.
I uploaded the raw footage on YT :



I then played it back at the slowest speed, x0.25 and even then it was difficult to freeze frame at the correct moment, because the events last only a split second and within that split the second the content is not steady, but rather changing rapidly as you see the video board trying stuff out in real time before yours eyes...

Example of the crap you see flying by :



Still, you brought me luck, it was a good idea, and I managed some cool shots, see below.
It's perfectly consistent with the pictures I posted earlier of people's working 5000 models.
What struck me though is that the text is in English rather then French for all the other people that have these programmers... so maybe it was sold to UK company back then, then found its way back to France years later. Or maybe it's just a dead backup battery so it uses default valuesd and it so happens to be English rather than French. Who knows...

1) Right at power, when the CPU is booting from its ROM I would say, it displays the Micropross logo, then some text to tell you to insert the boot floppy, then a few  error codes.



What I find odd though, is that if you pay attention, and even though the pic is blurry.... the text appears to mention "ROM5000" but I have a 3000 model !
So could it be that even the firmware in the 3000 is the same as that of its big brother the 5000 ? So that means that maybe I could get it to boot a 5000 system disk and have more features to play with... It's very odd and intriguing at the least.

2) Then once it's booted, it displays always the same image, again perfectly consistent with other's pictures : a row of five "buttons" / rectangles.
Here however the text and options displayed clearly show that it's a "headless" unit not a fancy 5000 : look, the menu bar has zero option, only the last / fifth button  is labelled, and it just says "END". And the text below below the buttons says : " REMOTE CONTROL " (as I thought it did)




... because well, this unit is only meant to be operated via the serial port with an MS-DOS S/W. You are not supposed to have a K/B to "press" buttons on the menu bar, never mind have a screen to see what the programmer displays..... so it makes sense not to offer the user any options that he could not see nor select anyway...

Yeah it all adds up, I am happy with that !  :D


OK so now I will add that XOR gate to the breadboard to create a Csync signal and see if that helps....

Robert763:
You can custom program the scaler chip on the GBS82x0 see
https://www.vandoeselaar.com/tinkering/modding-a-gbs-8200/
Robert.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod