News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu
avatar_talrek

Keyboard or Mainboard problem

Started by talrek, 15:57, 21 October 14

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

arnoldemu

Just had a thought. If a test program was designed to be run from |cpm then you could force a link on the board to boot com when cpc starts, that may make it possible to run some keyboard tests.

A better test program would run from Rom, but that needs extra hardware to start it putting it out of the reach of many.
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

Quote from: arnoldemu on 11:28, 26 October 14
Bit 7 or D7 is stuck. All these keys are on the same bit but over multiple keyboard columns. Is there a short somewhere between keyboard and ay?

Yes, that's what I thought too, but I can't find the spot where it's happening.

Bryce.

arnoldemu

Quote from: Bryce on 12:55, 26 October 14
Yes, that's what I thought too, but I can't find the spot where it's happening.

Bryce.
I have attached a dsk I made.

If /exp is setup so disc rom is 0, the cpc will autoboot from this when started up. No idea if the plus will auto start I never tested it.


The display will show lines. Red blobs will appear when a key is pressed.

Without any keys pressed it should just be vertical lines. If keys are pressed, or held by the hardware red blobs will appear.

From left to right are the bits with a line, with the lines going down the screen.
Bit 7 on the left and bit 0 on the right.

From this it should be possible to work it out a bit more easily.

My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

talrek

Quote from: Bryce on 10:51, 26 October 14
Ok, so we have exactly the same issue, good to know. The plus doesn't have a PIO, so that's not going to be the issue. However, I have noticed that this PCB also has problems with the floppy drive, does yours too? Also one other thing that I noticed. Mine has a French keyboard and an English ROM, what about yours?

Hopefully I can check it some more today and maybe find the solution.

Bryce.


English keyboard and ROM for me.
No problem with External Floppy drive.


With the joystick i have multiples characters before the arrows :




Bryce

Thanks, Talrek can try this, but unfortunately the floppy drive is dead on my PCB too, so I'd need to fix that first. I'll continue the manual search first.

Bryce.

arnoldemu

Quote from: Bryce on 13:29, 26 October 14
Thanks, Talrek can try this, but unfortunately the floppy drive is dead on my PCB too, so I'd need to fix that first. I'll continue the manual search first.

Bryce.
If I create a cart version would that help?

Eventually I could extend it to detect and pin point the problem, and add other testing to diagnose other problems.
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

#31
It would be excellent, but no need anymore, I've found the problem. Here's what's normally on each scan pin (coming from the ASIC in my case):

[attach=2]

And here's what's on pin 1 of CP104 (ie: pin 80 of the ASIC):

[attach=3]

As you can see, the signal is inverted (or not being inverted to be exact), which is causing the mess. On Talreks CPC I suspect IC101 (74LS145) has failed, as it is responsible for these signals on a classic CPC. In my case the ASIC seems to have dropped an inversion stage :( All is not lost however, I'll cut the trace and add an external inverter and see if that can solve the internal failure.

Bryce.

Edit: Hmmm, looks like this is an ex-CPC. I added an inverter and it was still causing problems. Then I connected all pins up to my Logic analyser to see what's happening. Although the signal is now correctly inverted, the pulse is about 4 times longer than it should be and also not at the correct time. This isn't something that can be easily fixed externally, so we have a dead ASIC. Time to look at the second plus that Craig sent me. It has a questionable ABBA mod that needs to be removed and it currently boots to a blue border with yellow box, so RAM is the first suspect.

talrek

wooooowwww  :o :o :o
Ok if you say so.... i will try a good one so..


I keep in touch  :)

arnoldemu

Quote from: talrek on 15:47, 26 October 14
wooooowwww  :o :o :o
Ok if you say so.... i will try a good one so..


I keep in touch  :)
I would be interested to see the result from my programming because then i could detect this in code and tell somebody to try that chip first.
My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

Unfortunately my PCB wasn't in a state to read disks, but Talrek could still test it on his CPC. If not, it's relatively easy to simulate this failure, so I can test it later if Talrek doesn't.

Bryce.

CraigsBar

Quote from: Bryce on 14:35, 26 October 14
It would be excellent, but no need anymore, I've found the problem. Here's what's normally on each scan pin (coming from the ASIC in my case):

[attach=2]

And here's what's on pin 1 of CP104 (ie: pin 80 of the ASIC):

[attach=3]

As you can see, the signal is inverted (or not being inverted to be exact), which is causing the mess. On Talreks CPC I suspect IC101 (74LS145) has failed, as it is responsible for these signals on a classic CPC. In my case the ASIC seems to have dropped an inversion stage :( All is not lost however, I'll cut the trace and add an external inverter and see if that can solve the internal failure.

Bryce.

Edit: Hmmm, looks like this is an ex-CPC. I added an inverter and it was still causing problems. Then I connected all pins up to my Logic analyser to see what's happening. Although the signal is now correctly inverted, the pulse is about 4 times longer than it should be and also not at the correct time. This isn't something that can be easily fixed externally, so we have a dead ASIC. Time to look at the second plus that Craig sent me. It has a questionable ABBA mod that needs to be removed and it currently boots to a blue border with yellow box, so RAM is the first suspect.
Oh Shambles!!


Sorry :( Is the 2nd one the one that originally had an STS Copymod on it? I don't remember either of those actually having an ABBA switch?


Craig
IRC:  #Retro4All on Freenode

arnoldemu

Quote from: Bryce on 16:04, 26 October 14
Unfortunately my PCB wasn't in a state to read disks, but Talrek could still test it on his CPC. If not, it's relatively easy to simulate this failure, so I can test it later if Talrek doesn't.

Bryce.
I think I could simulate it too. I am guessing that you can press any key, and when you do the result is replicated across all keyboard lines. So that pressing a joystick direction will give a different set of keys than delete, but this assumes nothing shows until you press a key.

An alternative is a key always appears stuck because an input always stays low, so both could be tested for.

I think if the membrane failed you could see all keys in a column be unresponsive, or more than one if it is a partial failure.

My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Bryce

Quote from: CraigsBar on 16:04, 26 October 14
Oh Shambles!!


Sorry :( Is the 2nd one the one that originally had an STS Copymod on it? I don't remember either of those actually having an ABBA switch?


Craig

CPC Laptop.

:)

The socond one had a chip piggybacked onto the 74HC14, some tracks cut and wires hanging from the floppy port..... and now works :)

Bryce.

talrek

IT WORKS !!!!!!!!!!!!!!


8) 8) 8) 8) 8) 8) 8)


It was the problem... damn IC  :laugh:


So, i'm sorry i am not able to test your disk as the CPC works now   :-[

Bryce

Result! Congrats.

@TFM: Guinness or Reissdorf Kölsch please :D

Bryce.

talrek

Yes really thank you !!!
If you come by France near Bordeaux i would offer you a beer  :D

Gryzor


TFM

Quote from: Bryce on 17:36, 26 October 14
Result! Congrats.

@TFM: Guinness or Reissdorf Kölsch please :D

Bryce.


Ok, Guinnes ready for pick up. However it wasn't the BASIC ROM either ;)  Happy to see that the problem is solved.  :)
TFM of FutureSoft
Also visit the CPC and Plus users favorite OS: FutureOS - The Revolution on CPC6128 and 6128Plus

zhulien

hi All,


I bought a 6128+ just recently on ebay, and although it was advertised as working it has the exact same (observed) problem described here as per the image in reply#20.  I have 2 6128+ machines and I have swapped keyboards and they both show the same problem on the faulty 6128+ but  both work perfectly on the good 6128+.  On the faulty one however I found out that the guy who sold it to me modified the 6128+ to use a different PSU plug which is another annoying thing (that wasn't advertised).  Whether he did or not, one of the smaller chips has an extra wire wired to it on this 6128+ also.


Any queues as to what I should look at to isolate and fix this problem?


Thanks, Julian

Bryce

Unfortunately, this probably means that the ASIC is faulty and this isn't a part that you can get, other than removing one from another Plus. Which is kind of pointless. That's an ex-Plus you have there.

Sorry,

Bryce.

zhulien

I guess that sucks... could there be a firmware fix for it?  if the del key is held down, those same extra characters appear but it does register as a del key too.

Bryce

On the plus I tested there were several additional problems inside the ASIC, no amount of Firmware fiddling is going to fix this.

Bryce.

gladders

Hey all :) so I have a CPC 464 which boots up seemingly pretty well to a small HDTV I have (although a slight image wobble which I'll leave for now)


But the keyboard isn't entirely working.


The Caps lock, c, b, n, m, /, \, plus 2 and + on the numpad won't respond. According to the technical specs, this makes sense - one of the lines on the matrix of the membrane has expired.


I've opened up the 464 and given everything a good clean and looked at the membrane. It all looks intact, so I guess a new membrane is needed.


Does anyone know where I can get one? :)

dragon

#48
Quote from: Bryce on 22:36, 29 August 15
Unfortunately, this probably means that the ASIC is faulty and this isn't a part that you can get, other than removing one from another Plus. Which is kind of pointless. That's an ex-Plus you have there.

Sorry,

Bryce.

I know one shop selling asic chip but is no cheap :( .

Semiconductor: 40489 - AMSTRAD-IC GATE ARRAY

Amstrad made much more componets that plus computers  builded :).

arnoldemu

I had a 6128 in storage. I had cleaned the membrane carefully.
But some keys didn't work. I removed the membrane carefully cleaned the lines on the membrane where it goes into the connector. I used a bit of spit and cleaned it very carefully and dried it.

now all my keys work again!

So in my case, the membrane was good and I either needed to re-insert it again, or the end had become a bit dirty.

Two times I've done this and both times it worked great.

So try this if you've not done this already.

My games. My Games
My website with coding examples: Unofficial Amstrad WWW Resource

Powered by SMFPacks Menu Editor Mod