Hi guys,
Having extended problems with a 6128 that i've recently replaced some RAM on. However, only the first bank seems to be working.
Using this topic: http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/repairs-to-a-cpc6128/50/ (http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/repairs-to-a-cpc6128/50/)
I have come to the conclusion that one of the banks of ram isn't working. All RAM is good, so there must be a trace gone on one of tha banks.
As it stands, the machine boots but only in 64kb. If I try to load CP/M 3 or 128k games, the computer crashes, I can confirm that all traces on the left hand bank of ram are good, but have yet to check the right. Which is bank 0 and which is bank 1??
And, presumably Bank 0 is the initial working space ram when the computer is in AMSDOS?
Taking a look to the 6128 schematic:
- 1ST BANK (CAS0 signal) is IC 127 to 134 (all CPC)
- 2ND BANK (CAS1 signal) is IC 119 to 126 (6128 only)
It could also be a problem with the PAL. Try removing the PAL IC and add jumpers to swap the upper and lower banks. Then you'll know whether it's a PCB track that's bad.
Bryce.
Yes. It require to link LK5, LK6, LK8 and put CAS1 to VCC to got a nice CPC 6064! ;D
Okay cool.
I tried a bit of that last night.
At least I know that the bank on the RIGHT is bank 0.
I put links in LK5, lk6 and lk8 and removed the PAL chip. Seems the 'RAM Square' appears then. But I suppose at least I know which bank I am trying to put my focus to.
I'm a little confused. I got continuity between each ram chip. Perhaps the traces between the banks are interupted?
In order to test the PAL chip, I have a spare I can swap it with, but do the jumpers need to stay in place to test? The purpose of the jumpers is just to swap the primary ram banks no? The PAL can be removed regardless as to whether these jumpers are inserted or not?
Toto... CAS1? VCC?
Because, if you remove the PAL, the CAS input of the RAM bank previously connected to the CAS1 output from the PAL is now floating.
Thant means, the RAM should produce a random activity... (Vcc to the unused bank CAS input avoid that)
This configuration currently works. So we know that bank 0 is all fine... ? What I don't understand is..
Isn't putting the three jumpers + removing the hal chip supposed to use the LEFT bank instead?
So.. there must be a continuity issue between all those chips on the left, or between those and the hal. I've printed off the spec sheet, so we'll see.
If I put chips into bank 1, then the square come up. Does the bank swap only happen if both banks are occupied!?
When the PAL is installed, the /CAS signal is used to create /CAS0 (1st bank) and /CAS1 (2nd bank) signals.
If you remove the PAL and add the links, LK8 connect /CAS to /CAS0 to use this bank, so looking the schematic, there is no memory bank swap.
I'm not sure exactly what signals these are (no schematic here at the moment), but this is my bank swapper that I use to test 6128 RAM. This device definitely swaps the upper and lower banks.
Bryce.
Hehe nice! ;)
I does the same with "flying wires" on the PAL socket when I have done tests for the X-MEM engineering.
I see. So doing the jumpers plus removing the pal doesn't truly test bank1?
I spend a few hours last night following traces and I'm completely stumped. They all hav continuity between themselves and the other bank (where they should). I also checked top side too, just incase the chip legs themselves weren't making a connection.
If the main bank is working, just put it back with the PAL and run the memory test I've attached to this thread
http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/repairs-to-a-cpc6128/msg114838/#msg114838
Then post the result screen.
A working 6128 should have the 7FC4 line green after the test. Anything red is not a good sign
I have not read the first post link... I though that was already related to your RAM test. ;D
Quote from: TotO on 14:12, 22 January 19
Hehe nice! ;)
I does the same with "flying wires" on the PAL socket when I have done tests for the X-MEM engineering.
If you have a pile of 6128's all with RAM issues to fix, then this little jumper socket is a lot easier and faster than individual flyleads. It's just one of my "Custom CPC tools" that I use, along with things like a 3 space jumper to power 464's without the tapedeck attached.
Bryce.
Thanks for that Gerald.
Can that image be used to 'burn' to floppy discs?
Either it can't, or i'm seriously running out of good 5.25 discs to use to try and get this programme on the 6128, for that is the only way i can transfer data.. unless you could upload one in audio data format? :P :P
Spent a good while last night going through my floppies.. i think i must have come to 25 5.25" discs that are bad. Even when they've never been used. Pitty :(
Tape version here :
http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/x-mem-tools-disk-software-and-manual/msg150412/#msg150412
Thank you for that Gerald.. You are really very cool!
This is the output... It didn't go beyond this point.
ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg (ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg)
(Local file upload wasn't working)
(http://ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg)
Quote from: Blurredman on 18:27, 25 January 19
Thank you for that Gerald.. You are really very cool!
This is the output... It didn't go beyond this point.
ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg (ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg)
(Local file upload wasn't working)
(http://ftp://blurredmanswebsite.ddns.net/Pictures/Computers/Amstrads/memtest.jpg)
It clearly show that the second bank is not working. Unfortunately this test does not give any hint about which bit is not working.
It does appear that way, thanks.
I've checked all chips are good, and they have continuity between themselves..
What do I need to check from the second bank to the 'link' between the two?
Quote from: Blurredman on 19:08, 25 January 19
I've checked all chips are good, and they have continuity between themselves..
Did you check that :All ram devices are powered (ie connected to VCC / GND)
All ram devices (both bank) shares RAS (pin 4), WE (pin 3) and addresses (pin 5,6,7,9,10,11,12,13)All ram devices in the same bank share CAS (15) which come from the PAL (pin 6 and 7 though a 47 ohm resistor)All data bit are shared between banks, ie din (pin 2) and dout (pin 14) of ram device on the same bit lane are connected together respectively
Make sure you test the connection of second bank on the ram themselve and not on the socket solder point.
Also, how did you check that the RAM device are OK ?
Thank you for clarifying that.
I can confirm that:
- All RAM has power.
- All ram get RAS, WE and Addresses with eachother.
- Data bits are shared.
I've checked these top side using the ram leg themselves..
I have another CPC with sockets that I have tested. Plus, I have around 15 spare chips that I have all tested okay, and used them too, unsuccessfully.
One thing: What is CAS? (On the RAM chip?) I find no connection between either pin 6 or 7 of PAL with pins 15 on the ram chips.. Or did I mis-interpret that? As they also have no connection to the working ram.
I am an amateur, so I'm pretty new to IC and board diagnostics.
Quote from: Blurredman on 20:45, 25 January 19
One thing: What is CAS? (On the RAM chip?) I find no connection between either pin 6 or 7 of PAL with pins 15 on the ram chips.. Or did I mis-interpret that? As they also have no connection to the working ram.
These are connected through a 47 ohm resistor (R161 / R160), which may be a bit high for a continuity tester.
You may need to check that R160 is fine as well as connected to pal on one side, and to Ram on the other side
Okay. Thanks.
Have tested the resistence. 47ohms. good.
Checked link from ram pin 15 to resistor. good.
checked link from pin 6 (17?) of HAL to resistor. Good.
Checked directly from HAL to Chip. good...
Only one thing I notice, is that i'm not testing pin 6 and 7, but 16 and 17.. And they have conintuity to each consequitive pin on a ram chip. Am I correct here?
All ram chips on pin 15 connect to the appropriate pin on the HAL/PAL.
I'm lost :laugh: :laugh:
Quote from: Blurredman on 21:51, 25 January 19
Only one thing I notice, is that i'm not testing pin 6 and 7, but 16 and 17.. And they have conintuity to each consequitive pin on a ram chip. Am I correct here?
Yes, my eyes are getting old, or ... lazy ;D
Any ideas on next things to check? :)
Check the continuity between each data and signal pin to the next point they should be connected to.
Bryce.
I've done as much as I could do given my lack of expertise.
Bryce.. Is it alright if I send you a PM to post this board to you, as well as the other I was having an issue with on a different post?
PM on it's way to you...
Bryce.
Hi Blurredman,
had a few minutes free and took a quick look at the first CPC you sent me. The one with Bank 1 not working. Looks like the RAMs weren't all good, although you said you had somehow checked them. One was bad and after swapping in a good one Bank 1 now works. Should I leave the replacement (good) RAM in it, or do you have further spares yourself?
I still have 15 minutes before "Bedtime story" with my son, so I might take a look at the other board now... :)
Bryce.
Just took a look at CPC 2. The one that turns off after a few seconds. It's turning off because the PSU can't handle the load. Keeping it forced on with a constant current limited supply, the voltage drops to 2V and almost every RAM IC on bank 0 gets to a temperature that you could fry an egg on. Did this computer suffer an over-voltage? Or do you know its history?
Looks like the entire bank needs to be swapped out, but I don't have time tonight. I'll get back to it tomorrow if possible.
Bryce.
Bryce.
Thank you for looking at these.
IE board 1... Strange.. Yes.. I tested all my chips on a socketed CPC.. Just throw the bad chip away and leave the socket open. :-*
Board 2 - I have no history at all on this board.
If you want. You can just put open sockets on the first bank to see if it will boot up? Then simply test with chips you have available? If the problem is chips I have a large excess of RAM chips so don't need them. Obviously, replacing said chips with sockets and your own test chips may open up other problems however..
Ok. I'll isolate the known shorted RAMs and see if it boots to the grey screen with border, which would show me whether the rest of the board is good. If it passes that test I'll socket bank 0 and try it with some spare RAMs. After posting last night, I noticed that the main electrolytic cap is getting hot, so I'll swap that first and see what changes.
Strangely, one of the RAMs in Bank 0 stays completely cold which either means it's good or completely fried open circuit, but I can test that too before removing it.
Bryce.
So unfortunately we've come to the question I never like asking: How badly do you want this 6128 fixed? It's definitely had a bad over-voltage or reversed polarity. Almost every chip on the board is fried. Unless this is the board your favourite Grandfather gave to you for your birthday shortly before being run over by a bus and has serious sentimental value, then I'd leave it. Fixable yes, but at a crazy cost.
Bryce.
That's unfortunate..
:(
Give me a PM of payment details and for my address. I guess...., send them both back. I can at least make use of the board for scroungy bits.. :(