| Electronics > Repair |
| Vintage chip Programmer : " Micropross ROM 3000U " |
| << < (13/38) > >> |
| Vince:
--- Quote from: Robert763 on December 26, 2022, 10:17:27 pm ---Hi Vince, see this manual too https://docplayer.net/42901042-Cga-ega-yuv-to-vga-converter-gbs-8200-ver-3-4-manual.html You have ver 4. Just connect ONE of the video pins of the progrmmer to the converter board, I suggest to the Green (G) input. Each input has a 75R load so 3 in parallel is likely to overload the single output driver in the programmer. Connect the 15kHz signal to Horizontal sync (HS) and the 50Hz one to Vertical sync (VS). Trouble is you have an unknown specification signal so hard to tell what is wrong. B.T.W when looking at the serial ports the data should be standard pin-out e.g. on pins 2 & 3. Robert. --- End quote --- Oops I missed your first post sorry. Thanks for the link. OK will try that. Yeah I noticed the Serial port and also the I2C headers next to the MCU. Maybe the serial port output useful data that could help me understand what the board is unhappy with exactly...but I don't have a computer in the lab just yet, so can't hook a terminal to that board. Well, OK I could if I hooked my vintage computer to it, with Windows 2000 and 95, but that would require an enormous logistical effort... just no space to put it anywhere sadly. An old Win95 laptop would be handy here for sure.... |
| m k:
You should be able to see OSD menu. CGA and EGA modes are sync polarity selected. SW seems to be swapping between modes and not between connectors. It says the left 5-pin connector is for JAMMA, no idea what kind of signal it has. Included pinout picture has csync, so H+V. I remember somebody had problems with one of these converters. Can't remember how it went. One other recollection is that CGA/EGA connector on the left uses different ASIC input pins than VGA. If so then SW can pick connectors. Your earlier picture with edge connector has no yellow, only RGBC and GND. Maybe the converter needs that csync, one inverting gate chip can do that. If you have a VCR you can try its YUV and see that the converter really does its down conversion. (there sync goes with Y) |
| Vince:
--- Quote from: m k on December 27, 2022, 09:05:46 am ---You should be able to see OSD menu. --- End quote --- Not sure what you mean, I clearly said that the menu works just fine ?! :-// --- Quote from: m k on December 27, 2022, 09:05:46 am ---It says the left 5-pin connector is for JAMMA, no idea what kind of signal it has. --- End quote --- "JAMMA" from what I understand refers to something specific to the arcade game world. It requires a special connector to interface the game with that board. Earlier as you saw, I ordered this JAMMA connector by accident, instead of ordering the actual board... oops. --- Quote from: m k on December 27, 2022, 09:05:46 am ---Included pinout picture has csync, so H+V. --- End quote --- " CSYNC " , is that a thing ? I didn't know about that... I knew only about RGB + H + V sync signals, or composite which merges all the signals... but I didn't there existed a middle ground where the two Sync signals were merged and the 3 colours signals were still separate. Interesting... --- Quote from: m k on December 27, 2022, 09:05:46 am ---Your earlier picture with edge connector has no yellow, only RGBC and GND. --- End quote --- I don't understand what you are referring to. --- Quote from: m k on December 27, 2022, 09:05:46 am ---Maybe the converter needs that csync, one inverting gate chip can do that. --- End quote --- Not sure what you mean. The connector I used, as per the manual, has already H and V pins on it, it's meant to be directly interfaced to the CGA signals. --- Quote from: m k on December 27, 2022, 09:05:46 am ---If you have a VCR you can try its YUV and see that the converter really does its down conversion. (there sync goes with Y) --- End quote --- I did have a VCR but I tossed to make space. Used to have a CCTV camera with a composite output, but I too tossed it to make space. My digital camera maybe has a composite output built into its USB looking I/O port, just maybe, but I would need to find the manual and the adapter cable and I think they must be long lost.... I checked all my old hollow-state Tek scopes and not one of them as a video output ! I am very surprised and disappointed.... think I will return them to Tektronix. OK maybe we are lacking gear here at Vince's lab to be able to do basic stuff like this... but that's nothing new, lab is a work in progress, as in.. slow progress... @Robert : thanks for the manual, it's 90% identical to the one on the site I saw, but the 10% difference was interesting. It states the range of horizontal frequencies that it can accept : from 14+ to 16+ kHz, so that means anything close enough to 15kHz would work just fine, so I am good here. However that manual too insists on calling the 15kHz clock.... VERTICAL. It drives me nuts. |
| Vince:
--- Quote from: Robert763 on December 26, 2022, 10:17:27 pm ---Just connect ONE of the video pins of the programmer to the converter board, I suggest to the Green (G) input. Each input has a 75R load so 3 in parallel is likely to overload the single output driver in the programmer. --- End quote --- Here is another day, troubleshooting continues.... OK so I disconnected red and blue. Amplitude increased to 2.6V on the green pin now, that's better...though I guess it's irrelevant, as any voltage is good enough for the colour channel,, I should still be able to see "something", and at any rate it's not going to cause the board to fail to output a VGA signal... so I guess all that matters is the amplitude of the H/V signals. These have a higher amplitude : 3.6V or so, one volt more than the colour channel. Next step is... I could add a TTL chip, a buffer, on my breadboard. I measured the input impedance for the H/V input. It's not 75R it's a high(ish impedance, so a regular TTL buffer chip should have no problem giving proper amplitude output there. I measured the inputs at 1.8K and 10K+ . OK let me do that and report... One odd thing I noticed on that board : I measured the 75R RGB input resistors (directly across their terminals), they all measure 75R, fine. However if I measure the impedance between the RGB inputs pins and the round pin on the connector... I get 75R, 82R and 88R ?! :wtf: Still, not too concerned about that, as only the H/V signals are of importance in how the board reacts to my video input. Well, from my limited understanding I mean... |
| Vince:
Making progress here. Did as I just said... pulled my bread board and added buffers to the H/V signals. At first with a TTL chip.... got only 4.4V at the output, bummer.... then light bulb.. ah yeah I am rusty, TTL outputs have never been able of 5V to begin with.. there is a diode to Vcc internally, the best you can hope for is 4.4V or something, and that's what I get. So instead I used a 74HCT chip, a 244 8 bit bus driver, because that's what I had.... not I get 5V at the output. Well, the first chip I used was defective, was doing strange things... the second chip was the winner. Plugged the wiring harness to the video board. Amplitude drops a bit but not much. Most notably the leading edge of the 15kHz signal is now quite rounded, see pic. The rising edge looks good for the most part, but the very top of it, between 4 and 5V, is quite rounded. That's when I was scoping on the breadboard. I then scoped straight on the video board connector in case all that spaghetti wiring were degrading the signals badly... but no. Signals look exactly the same no matter where you look. I got curious and searched which one of the millions SMD passives the H/V signals were connected to, on the board. See pic. There is a long row of passives between the connector and the video chip. I scoped these passives and quickly found where the H/V signals were routed to. They go each to a 1K resistor. My signals get to one side of the resistors just fine, 5Volts, however on the other side, where it buggers off to the chip, the amplitude drops to 4V. Hope that's OK for the chip. Anyway, signals had at last a proper amplitude, so I fired up everything... ... and we have improvement ! Now the board can detect something ! It can't make sense of it though, so it keeps cycling endlessly, displaying various kinds of garbage, but GREEN garbage which is promising since I used the green colour channel... so clearly it's seeing my signal now, no doubt about that. It cycles between 3 states mostly, from what I can see : 1) A solid green screen filling the entire screen. 2) long pause with a black screen.. probably it's thinking hard on what to do next 3) A "half solid" green screen... looks like a fine mesh, a woven fabric, something like that... 4) long pause with a black screen again 5) Now the interesting bit. On this part of its cycle, it displays a black screen with some green garbage flying by, that looks like it contains valid data, but garbled in the way you get when there is a timing issue. Whenever it displays this, it lasts for a split second, half a second if even that, so my brain doesn't even have time to absorb what it's seeing... but after repeating the cycle many times (you never get the exact same garbage twice of course, depends on the pot luck of the timing at that particular time), I am 100% certain it displays intelligible ASCII text, and some geometrical shapes. It displays a row of horizontal rectangles, just the outline of them, like you would expect from a status bar displaying place holders for menu options, and you would press function keys on the K/B to activate this or that menu option. That pleases me, it adds up. All these rectangles are empty but sometimes I see one that has text in it (no time to read what it says though), in reverse video. So that's perfect, I like that a lot. Also, on the next line, right below this line of rectangles, it displays some more text, a couple of words : " REMOTE "... followed by another word but I am not 100% sure what it is. Maybe " REMOTE CONTROL ". Again I like that, because we know that the big brother of this programmer, once it's booted, gives you the choice of being used either remotely via a desktop computer, or locally using the K/B attached to one of its serial ports at the back. So it all adds up nicely. So the board clearly can make "some" sense of my video signals to display something interesting.... but it's not clever enough to get the timing 100% right. Two possibilities here I guess : - The programmer indeed uses non standard signals, some custom variation of CGA, but not compatible with off the shelf CGA monitors... just so that they could force you into buying their own, expensive of course, monitor... - The board is defective. Indeed, it just keeps crashing and crashing and crashing. It's a disaster. Every time I power it up, I do get the chinese "splash scree" while its booting, then a black screen, and from there it's pot luck : half of the time the buttons won't be responsive and I can't enter the on-screen menu. Other times menu will work, but then when I exit it and try to use the SW and AUTO buttons... they won't be responding.. so as a quick sanity check I press the MENU button to see if the board is still alive and... most of the time it's not, menu doesn't work any more. It's a disaster. So I guess I could try to get a new board under warranty, see if that's any better. I guess it's also now time to read the 5 pages of messages on that French forum I was pointed to at the beginning of this saga, where a few people discuss their programmer, the big brother of mine. Maybe they talk about the display... and if not, I could post a message about it. https://forum.system-cfg.com/viewtopic.php?t=5743 So I am making progress eh, what do you say ?! ;D |
| Navigation |
| Message Index |
| Next page |
| Previous page |