Space firebird boot blank screen

[Lurch666]

Got a nice firebird boot from jonhughes and have been diving into it for a while now but seem to have hit a bit of an impasse.

Got it to the point of running code but still have a blank red screen. Even though it's a boot it 99% follows the original firebird schematics. It just takes it's rgb and sync from the part of the circuit before the transistors (the positive sides of C2,C3 and C4) that I assume invert the signal on an original (I think the original boards require an extra converter for the video signal).

Anyway what is happening is there is no output from 74LS174 chips at 5H,5I and 6I because the clock on pin 9 is out of sync with the signals going in so no signals go out.

I have found that putting a small capacitor between GND and pin 9 of one of the 174s can modify the timing enough to allow the signals out but then after going through the graphic roms at 5K and 6K the signal is now stopping at the pair of SN74198 at 5M and 6M.

This appears to be because of the state of pins 1 and 23. pin 23 is always low but if I bring it high I get something on the screen (a mess of moving lines) but I don't understand how it's supposed to work properly.

So if anyone has a firebird could they test a couple of things for me?

Does the board run properly if you don't have the sound board plugged in?

I know some graphics will be missing but do you get anything on the screen without the sound board?

Do you get anything on the screen of the board you remove the 1st CPU rom at 5E so the program isn't running.

And when the game is running normally do you get activity on pin 20 of the CPU or activity on pins 1 and 23 of the SN74198 chips at 5M and 6M.

This will tell me if the board has to be running properly for me to get graphics and if those pins on the SN74198 chips are significant.

 

[Ronnie Dent]

It will run without sound pcb but the starfield will be missing as it's generated by the sound board

 

[Lurch666]

Thanks for verifying that dude.

I knew some graphics are generated on the soundboard but just wanted to verify the other graphics should still appear without the soundboard being present.

Didn't want to try and fix missing graphics if it was down to a problem with the sound board.

 

[Lurch666]

I just found out that a none running (garbage screen) boardset should still have a star field if the soundboard is attached.

I don't get any star field so something is very wrong with the video circuit.

Will have a look at 1A on the video board as it's where the star field data comes off the soundboard.

EDIT:OK I had checked the chip at 1A on the CPU board before since this generates some of the timing signals and it was ok but now I have rechecked it and found something I had missed before.

The clock signal comes from a 20 mhz crystal but according to slice it's only getting a 16mhz signal so maybe the crystal is bad.

Have ordered a new 20mhz crystal and hopefully if it's the crystal at fault it will explain the timing problems I am having in the video circuit.

Lurch6662021-01-12 16:53:17

 

[patzik]

I have 3 running boards, I just finished repairing then and am waiting for some ram to arrived to complete them as they had more than 12 bad rams between them.
If you need me to do some measurements just let me know.

 

[Lurch666]

What effect did the bad rams have?

These boards have a lot of different ram and I'm not sure all my rams are good so did you get a blank screen with your ram faults?

 

[patzik]

To be honest, I can't remember but I still have the faulty ram so I will put it in and see what happens. When I get the new ram I will grab the board and do it (should have arrived today as tracking code indicates delivered today, but nothing here... hope it didn't get lost).
patzik2021-01-13 15:35:48

 

[Lurch666]

New crystal fitted but still no difference.

I have gained a bit of experience with this part of the circuit that converts the bytes into a stream of bits while trying to fix my space guerilla boardset (there's a thread in the fix logs) so now I have an idea of what SHOULD be happening.

The SN74198 chips at 5M and 6M have two inputs that dictate what should be happening.

These are pin 1 (S0) and pin 23 (S1).

If S0 is high and S1 is high the data from the input pins (3,5,7,9,15,17 19 and 21) are loaded onto the output pins (4,6,8,10,14,16,18 and 20).

If S1 then goes low for each clock pulse the bits on the output pins are moved onto the next pin so the data on pin 4 gets moved onto pin 6,the data on pin 6 moves to pin 8 and so on (it could be in the reverse direction but you get the idea).

Pin 20 of each chip is connected to a 157 which directs the outgoing bit stream to the colour prom.

So what is happening is the eight bits presented on the input pins are presented to pin 20 one at a time for each clock cycle so each byte of graphic data is turned into a stream of single bits.

On my firebird pin 23 (S1) is not going high so the graphic data isn't being loaded onto the outputs to be turned into a stream of bits that are required to display anything.

That pin is driven by two inputs.

Pin 15 of 1E which is the carry bit but it's never active. Pin 1 (clear) is toggling and I can't find out from the schematic where the signal to pin 1 is coming from.

The other signal is RV from the CPU board that comes from the LS174 at 3K pin 2.

The input pin is active and the clear is high but the clock on that chip is missing.

The clock comes from pin 12 of 4I and that chip is working (as are the rest when tested with my slice) and it seems to be getting data properly but I can't see how it's all supposed to work together.

So on a working firebird is there a clock signal on pin 9 of 3K on the CPU board and can anyone spot on the schematics where the input for the clear of 1E comes from since I can't seem to find it.

It's marked 'J' but can find the point it references.

 

[Lurch666]

Progress at last!

Now I have a blank black screen instead of a blank red screen.

BUT:If I plug in the sound board I'm now getting stars.

The problem was I had missed a ram chip.

3I on the video board is a 27LS00 (it was marked with other numbers so I had to refer to the schematic) and I didn't realise this was ram and had ignored it but on finally noticing it and pulling it from it's socket I noticed several legs were short and it was very tarnished.

Removing it made no difference so I found I had some 93411 ram which is the same and plugging that in gave me a blank screen and with the sound board I get stars so now I know at least the video output is working as it should so if I do anything to get the last part of the video generation working I should see something on the screen.

Also have found that the 20mhz signal that was only at 16 mhz or so was down to voltage.

I was running the board at around 4.5V as that caused activity on the CPU board but I have now found it was also causing the crystal to run slow so now I'm running it a 5V and the output is now 20 mhz.

I'm hoping that burning new CPU roms will get it working at 5V and if not I'll replace the 2114 ram in the CPU board.

Still got to figure out the other problems but it's nice to see something on the screen at last even if it's just the stars from the soundboard.

 

[Lurch666]

I have found I can make some graphics appear by doing certain things to this PCB.

Normally it just displays a blank screen but it's running code if I keep the voltage low.

Here I have grounded pin 8 of 2B (74LS02) that causes the game to appear albeit garbed.

The game is running as can be seen in the video but the PCB only runs if the voltage if kept below 4.5V which I'm assuming is throwing the timing off on the 20mhz crystal.

I've traced all the signals leading 2B and can't find any faults.

The pins on 2B are active so the signal timing must be incorrect.

One thing that's really puzzling me is the CPU roms.

There are eight 2732 roms and using slice I have checked them to see if they are operating correctly and they all are except one that's showing errors on the output.

I've changed the rom and replaced the socket but still am getting errors and I think it might be this chip stopping the PCB running at higher voltages.

I just don't understand how it can still be showing errors on it's outputs as it can't be anything else on the bus or all the roms would be showing errors on their outputs.

 

[Lurch666]

So I figured that since I can make the graphics appear by grounding pin 8 of 2B on the video board what I'm now doing is checking other components to see if there's any change when I ground the pin and the graphics appear.

Then I can narrow down what changes and that should lead me to what is going wrong.

Pin 23 of the two SN74198 chips start toggling when I turn the graphics on where before it was stuck low and I knew that had something to do with me getting a blank screen.

Now I will trace backwards to find out why pin 23 becomes active when I ground pin 8 of 2B

 

[patzik]

Today I checked my working board with the faulty ram on the video board, and I get a similar output as what you are getting. The game runs, but there is nothing on screen except the stars.
Did you check the outputs of the 4 SRAM 256 x 4 chips on the video board?

 

[patzik]

The 'new' ram I bought came in yesterday and unfortunately it is not working. The board has AM93422 in it and I replaced it with a different brand, and I guess these are too slow, because I get corrupt graphics.
The ram works fine in my centipede board which only needs slow ram.

Anyone have an affordable source for these AM93422 ic's?

 

[Lurch666]

Did you check the outputs of the 4 SRAM 256 x 4 chips on the video board?

Which ram are these?

There are eight rams of 8029 (I used 2101A) at 5B,5C,5D,5E,6B,6C,6D and 6E,two MB7063 (I used 82S09) at 6G and 6H and a single 27LS00 (I used a 93411) at 3I.

 

[Ronnie Dent]

I know 5101 ram is compatible for some of the rams used?

 

[patzik]

The ram I used was 5101, but I think it's not fast enough.

 

[patzik]

The video ram are the 8 2101's. They have 4 input bits and 4 output bits. The output bits on my bad ram are all floating. Game runs, star pattern is there, sound is there, but no graphics.

 

[Ronnie Dent]

There was a lot of fake philips 5101 going about but some of them worked in my boards

 

[Lurch666]

Just checked mine and even though I'm not seeing graphics I'm still getting outputs on those eight chips so I'm assuming you got duff ram sent to you.

 

[patzik]

Actually the 5101 rams are fine, they work in a centipede board I have. They are just not fast enough for this board. Anyhow, back to your board, if the rams have activity on the output bits, the fault will likely be somewhere after this. On the video board, there are some testpoint, TP1 though 4, are RGB and sync.

Can you measure if the is a signal on these points?