Amstrad CPC 464 blue screen and Dandanator problems

[CPC78]

Hi all.

First poster here.

I have this CPC 464 I got in a mixed lot. I cleaned it up and it did work for some time. But now it just gives blue screen, but still manage to make sound and some image related things with Dandanator Mini.

I made a youtube video where I try some different things. But besides that I don't know where to go from here.

I checked some random stuff with my multimeter, but I don't really know what to check. It seems all print lanes between rams are ok and I did not find any other faulty print lanes, but again I mostly just checked on random.

[Rabs]

Hi, sounds like the CPC is working and looks like the monitor is ok. Have you checked the monitor socket on the back of CPC for dry joints. Try reflowing the connections.

[CPC78]

Well, that is one thing I have not tried for some reason.

I will let you know if that works.

[GUNHED]

Connection to monitor seems to be good. Check the expansion port connector. Maybe needs cleaning?

[CPC78]

I cleaned up the monitor socket, checked the soldering joints too.

I also took out all socketed chips and cleaned the legs and connections.

You cannot view this attachment.

This is how it looks when I boot without the Dandanator.

You cannot view this attachment.

This is the screen it gives when I boot with the Dandanator.

The picture is actually better and connection seems less glitchy, but there is still no actual text on start up or anything.

It did work at one point (before I had the Dandanator) and I have managed to play several different games on cassette.

[Fessor]

Does it beep when you press Delete?

I would think that one of the multiplexers is causing problems and the CRTC/GA cannot access the memory.
If the CPC responds to keystrokes, blindly type poke &4000,255 and see what happens.

[CPC78]

It does beep when I press Delete and when pressing the arrows. It also has sound when I am using the Dandanator and responding to keystrokes when I use the Dandanator. 

I tried that poke just now, but I could not get it to do anything as a result.

[GUNHED]

Could be a ROM problem. Somewhere in this forum there is a test ROM, maybe this can be put into Dandanator. Does somebody know?

[Rabs]

It does beep when I press Delete and when pressing the arrows. It also has sound when I am using the Dandanator and responding to keystrokes when I use the Dandanator.

I tried that poke just now, but I could not get it to do anything as a result.

Confusing. I guess because the CPC is beeping when you press delete that this implies that the AY and PIO are scanning the keyboard, which implies the Z80 is running the lower ROM and I guess implies that RAM is good (maybe). This then points to a display issue but the display with the Dandanator contradicts this and looks like it is hanging.

My go to diagnostics tool is the Diagnostics ROM and I use a ROM board with the Diagnostics ROM acting as the lower ROM. I guess you can get the Dandanator to do the same.

I have had problems with edge connectors and even after countless cleaning attempts, sometimes I still have to reinsert connectors to get a clean contact. I clean the edge connectors with an old ink eraser and I never connect anything while the CPC is on. 

Did the Dandanator ever work? Sounds like no. So maybe two separate issues.

I would clean and clean again the edge connector and get the Dandanator to boot with the Diagnostics ROM and go from there.

Hope this helps.

[eto]

Trying to make sense of this:

- there is no proper display - but at least with the Dandanator we see some lines
- Dandanator works - games will play as we hear the sound
- CPC also boots without Dandanator as the "DEL-beep"-trick works

To my knowledge, we can rule out RAM and ROM:

- ROM fault is very unlikely. It is no short as then the CPC would not boot at all. A bit flip could happen, but then it would work with the Dandanator, as the Dandanator takes over full control (afaik) anyway 
- RAM fault is also very unlikely as Dandanator copies over a snapshot to RAM and then only having a display problem but the rest of the game/system works wouldn't make sense

Since there is a somehow proper display then the Dandanator is connected, my next best guess would be, that something is wrong with the CRTC or GateArray.

[Bryce]

My bet is that one of the 74LS153's has failed.

Bryce.

[CPC78]

My bet is that one of the 74LS153's has failed.

Bryce.

OK.

I will see if I get time today or during the weekend and I will test to see if there is any print lines to/from the 74LS153's that are broken.

Is there any way I can test the actual 74LS153's? Or will I have to find some other ones to swap with?

I don't know where I can find new ones or if there is alternative chips I can use.

Thanks for all the help so far. I am very new to working with Amstrad CPC and even new to working with the electronic side of old tech in general. 

[Rabs]

I defer to @Bryce and @eto as they have way more knowledge and experience than me but agree it looks like the CPC is working but the display is suspect. But cannot understand why the Dandanator has a display (of sorts). I did have what appeared to be a 74LS153 failure, showed as a total RAM failure (grey screen), so different. But this worked then failed. In the end the power was not stable and I had a bad joint on the internal power connector. Reflowed the joints and this fixed it. Just a thought.

[CPC78]

I defer to @Bryce and @eto as they have way more knowledge and experience than me but agree it looks like the CPC is working but the display is suspect. But cannot understand why the Dandanator has a display (of sorts). I did have what appeared to be a 74LS153 failure, showed as a total RAM failure (grey screen), so different. But this worked then failed. In the end the power was not stable and I had a bad joint on the internal power connector. Reflowed the joints and this fixed it. Just a thought.

Actually I previously completely replaced the power connector in the CPC as I thought it might be a problem, but that did not change anything. So I don't think that is the problem, but my solder skills are not the best, so I might check out again if I made a mistake there.

To my knowledge a lot of old machines often has problems in the power connectors since for many years people have pulled them out and plugged in over and over. This has been the case on a couple of Commodore machines I worked on in the past.

[Rabs]

I defer to @Bryce and @eto as they have way more knowledge and experience than me but agree it looks like the CPC is working but the display is suspect. But cannot understand why the Dandanator has a display (of sorts). I did have what appeared to be a 74LS153 failure, showed as a total RAM failure (grey screen), so different. But this worked then failed. In the end the power was not stable and I had a bad joint on the internal power connector. Reflowed the joints and this fixed it. Just a thought.

Actually I previously completely replaced the power connector in the CPC as I thought it might be a problem, but that did not change anything. So I don't think that is the problem, but my solder skills are not the best, so I might check out again if I made a mistake there.

To my knowledge a lot of old machines often has problems in the power connectors since for many years people have pulled them out and plugged in over and over. This has been the case on a couple of Commodore machines I worked on in the past.

Oh well just a thought, My problem was on the internal jumper pins.

[Fessor]

But cannot understand why the Dandanator has a display (of sorts).

Because that are Rasterbars that don't change anything in Memory. They are only perfectly timed ink-changes.

[CPC78]

On another note. I don't know if it is of any importance.

My board is the "COPYRIGH 1983" "MC0001A" with the slashed zeros and "PT NO 270100" version.

It also seems like all the 74LS153 chips are "Toshiba Japan TC74HC153P 8134H" chips

[Bryce]

On another note. I don't know if it is of any importance.

My board is the "COPYRIGH 1983" "MC0001A" with the slashed zeros and "PT NO 270100" version.

It also seems like all the 74LS153 chips are "Toshiba Japan TC74HC153P 8134H" chips

The 153's should be easy to find. Only the 74HC153 bit of the name is important, it can be from other manufacturers than Toshiba and the 8134H is not part of the name, it's the manufacturing date and packaging type. The reason why I suspect the 153's are the issue is because I think the CPC's ROM isn't able to set up the CRTC properly. Possibly just a single bit not arriving. The Dandator possibly uses different values where the bit is missing. You could try starting with a different ROM such as Firmware 3.1 which also uses different values if you have a method to do this.

Bryce.

[CPC78]

The 153's should be easy to find. Only the 74HC153 bit of the name is important, it can be from other manufacturers than Toshiba and the 8134H is not part of the name, it's the manufacturing date and packaging type. The reason why I suspect the 153's are the issue is because I think the CPC's ROM isn't able to set up the CRTC properly. Possibly just a single bit not arriving. The Dandator possibly uses different values where the bit is missing. You could try starting with a different ROM such as Firmware 3.1 which also uses different values if you have a method to do this.

Bryce.

I just included all info on the chip, sometimes there is known issues on specific batches or such.

But yes. I think I will see if I can fin dif there is any fault on any of the print lanes to or from the 153's and if I find no issue with print lanes I will try and order some new ones. I guess I can desolder and install sockets while I await new ones.

I have no means to change firmware at the moment. This is my only Amstrad computer and I have no Eprom Programmer or such yet.

[Rabs]

The 153's should be easy to find. Only the 74HC153 bit of the name is important, it can be from other manufacturers than Toshiba and the 8134H is not part of the name, it's the manufacturing date and packaging type. The reason why I suspect the 153's are the issue is because I think the CPC's ROM isn't able to set up the CRTC properly. Possibly just a single bit not arriving. The Dandator possibly uses different values where the bit is missing. You could try starting with a different ROM such as Firmware 3.1 which also uses different values if you have a method to do this.

Bryce.

I just included all info on the chip, sometimes there is known issues on specific batches or such.

But yes. I think I will see if I can fin dif there is any fault on any of the print lanes to or from the 153's and if I find no issue with print lanes I will try and order some new ones. I guess I can desolder and install sockets while I await new ones.

I have no means to change firmware at the moment. This is my only Amstrad computer and I have no Eprom Programmer or such yet.

In a gin fuelled ebay session, I bought 40 74LS153s. I only wanted 4 and did not realise they came in packs of 10 and in the end mine were not faulty. 

You could have some for free but they will take forever to get from the UK.

[Anthony Flack]

The computer itself seems to be working, the gate array seems to be working, the CRTC seems to be generating the sync signals correctly and the monitor is displaying colour splits but no pixel data...

I assume the Dandanator is trying to display this screen:




Like Bryce says, all it would take is one bit not arriving to the CRTC correctly for this to happen. Or could one of the counters inside the CRTC have failed internally? Are CRTCs known to fail? I assume you've checked all the traces.

Or perhaps the CRTC is fine and it just can't access the screen RAM to display the pixels, so it's just reading &00s?

[Bryce]

CRTC's do fail, but they usually fail completely and then you get no sync signals on pins 39 and/or 40.

Bryce.

[Rabs]

CRTC's do fail, but they usually fail completely and then you get no sync signals on pins 39 and/or 40.

Bryce.

@Bryce, as I have loads of 74LS153s  and some spare CRTCs, I am thinking of making a little electronics project mock-up to better understand their use. What is the best info on how they are used in the CPC?

[gerald]

If you get sound out of it, I assume the basic is running OK.
From the 2 screen you posted, one option is that the DISPEN output of the CRTC is not getting to the gate array such the gate array always display a border.
On simple thing to test is to change the border color with a

Code Select Expand

BORDER 0

[Bryce]

CRTC's do fail, but they usually fail completely and then you get no sync signals on pins 39 and/or 40.

Bryce.

@Bryce, as I have loads of 74LS153s  and some spare CRTCs, I am thinking of making a little electronics project mock-up to better understand their use. What is the best info on how they are used in the CPC?

The 153 is a dual 4 into 2 multiplexer. Each multiplexer has a 4bit input and 2bit output. Using pins 2 and 14 (S0 and S1) you can decide which two of the inputs are getting passed on to the two outputs. The chip has two of these multiplexers, so it has 8 inputs in total and 4 outputs. On the CPC, the GA controls S0 and S1 to control whether the RAM is getting address info from the CPU or the CRTC. This is how the CRTC accesses/reads the screen content from the main RAM.

You can experiment with this chip without having a CRTC or RAM, just using an Arduino and some LED's to see what's being outputted.

Bryce.