6128 Failed repair :(

[mrbee]

Ok, so what went from replacing one ram chip to then completely ruining my 6128!

I had bought all new sockets for all of the ram chips, there were about 7 traces on the edge of their life that got destroyed. I believe I have rewired them all correctly, they all test good continuity wise but Im looking for a guide on exactly what pins are supposed to pair up to what ones. I know some do not, they do pair to the chip to the left/right of one particular pin, but that pin does not connect in series then to the chips above it (if that makes sense). whilst other pins then testing the continuity even from the z80 itself and also the ic108 testing that the continuity is all good there.

I dont have a dandanator to perform a test.

I was going to perform the "simple" pal chip socket trick where you swap the pins with a wire setup like bryce has done, and then that swaps the ram banks so i can rule out if one is working or not. 
then the bottom half of the leg of the pal chip snapped off! So with that sorted then, i went to power on and...

whoops, i connected the 12v IN 2 pin connector to the 2 pin connector on the top left where the relay is located, powered on and... nothing.
I realised my mistake then, facepalmed really hard, and then connected the 12v in the correct pin.
Powered on and what I get is a black border with a solid colour screen, i believe its a pink colour (i am colourblind so it could be grey lol!).

My thinking is that ive missed a trace somewhere with the ram, but just looking for a good diagram of what is supposed to continuity test to where to best test this out. I am a real beginner at this

I have repaired a 464 and also a ctm 644 monitor, but this 6128 is really fragile! It did work before it was just that lone ram chip but since i decided to delve in.. ive made it worse and just am hating myself right now for attempting it.

Just looking for some advice on the next best thing to do

thank you

[Bryce]

I won't comment on your trace destroying activities, other than to say that it is wise to practice ALOT on scrap PCB's until you get to a level where you never damage traces. As for replacing the traces you don't seem to be referencing the schematics much (or at least not mentioning it). The schematic will tell you what connects to what and if you don't get continuity, then that trace is gone. You also need to make sure you haven't caused any shorts between pins or traces, because if you've been ripping copper off the PCB, then there's a lot of curly copper strips around the PCB that could be shorting things out.

Unfortunately, I think you are in the UK, otherwise you could send the PCB over to me for repair.

Bryce.

[mrbee]

Hey Bryce, thanks for replying. Yes am in the UK so woulda been great for you to have a look at it for me!
Comment all you like on the trace trashing, I deserve it. For some of them anyway, I've worked on quite a few boards of the years, replacing caps, replacing chips, fixing traces but the only two that have given me such an issue is an amiga 600, and this 6128.
The amiga 600 literally you could whisper and the pads fall of that board for me.

Anyways so with the schematics I am just confused looking at them to rule out the problem with the traces.

So for example, the top left pin (looking at the board on the chip side) of the top left ram chip in the first bank - I put my continuity tester on this pin, and then check the same top left pin on each chip going down the first bank - all connected to each other good. and then staying on this first top left chip and first pin top left chip in the first bank- i move to the second bank to test the top left pin and all is good - so both banks are connected together on every chip for this first pin.

The second pin across from the left then on the top row first bank, testing each chip this pin to each chip going down the bank is all good. But going across to the second bank then there is no connection. I'm thinking this is correct, as they aren't connected both banks together for this pin...
or is it?
So just getting confused with looking at the diagram here https://www.cpcwiki.eu/index.php/File:CPC6128_Schematic.png as I have a super basic understanding of this diagram.. I can't seem to see where the ram chips are here for me to check where each pin goes to. Is there a diagram for the ram?

Sorry if this has astounded you XD I am just learning all of this slowly, and just looking to rule out the traces of the ram first. I looked at noels video on traces, he ran in to the same sort of issue with me with the traces except the connection from the z80 and so on is all good to the ram

[McArti0]

Check your PAL conections. On picture you have number PAL pins on other ICs.

Look on IC115. D0-D7 (Pin 2,4,6,8,11,13,15,17) Check join to Pin2 RAMs (D0-Pin2 in IC119 , D1-Pin2 in IC120 etc.)

[Bryce]

The two big squares in the middle are banks 1 and 2 of the RAM. Each square represents 8 chips. As the note in the middle states: "All pins except data fully parallel". That's all you need to know.

Bryce.

[mrbee]

I will check this and have another go at this on thursday and report back  thank you for your help its very much appreciated!

[Rabs]

Hey Bryce, thanks for replying. Yes am in the UK so woulda been great for you to have a look at it for me!
Comment all you like on the trace trashing, I deserve it. For some of them anyway, I've worked on quite a few boards of the years, replacing caps, replacing chips, fixing traces but the only two that have given me such an issue is an amiga 600, and this 6128.
The amiga 600 literally you could whisper and the pads fall of that board for me.

Anyways so with the schematics I am just confused looking at them to rule out the problem with the traces.

So for example, the top left pin (looking at the board on the chip side) of the top left ram chip in the first bank - I put my continuity tester on this pin, and then check the same top left pin on each chip going down the first bank - all connected to each other good. and then staying on this first top left chip and first pin top left chip in the first bank- i move to the second bank to test the top left pin and all is good - so both banks are connected together on every chip for this first pin.

The second pin across from the left then on the top row first bank, testing each chip this pin to each chip going down the bank is all good. But going across to the second bank then there is no connection. I'm thinking this is correct, as they aren't connected both banks together for this pin...
or is it?
So just getting confused with looking at the diagram here https://www.cpcwiki.eu/index.php/File:CPC6128_Schematic.png as I have a super basic understanding of this diagram.. I can't seem to see where the ram chips are here for me to check where each pin goes to. Is there a diagram for the ram?

Sorry if this has astounded you XD I am just learning all of this slowly, and just looking to rule out the traces of the ram first. I looked at noels video on traces, he ran in to the same sort of issue with me with the traces except the connection from the z80 and so on is all good to the ram

Hi, I'm in the UK, Hampshire if that helps.

[SerErris]

The RAM chips are connected in very simple ways.

All the ram chips are connected with all PINs in a row- meaning that the same pin on the first chips is connected on all the other RAM chips with the same pin.

You cannot view this attachment.

There is three exceptions:

1. PIN 15. You have two banks and they are getting different CAS Signals. So each bank has the CAS line connected, but the two banks have it not connected in between the banks.

2. PIN 2 and PIN 14. Inside of each bank, the Data In and Data Out pins are not connected in between RAM chips. That is actually the only thing that makes chips individual. However the corresponding PINs of each chip from both banks are connected to each other, PIN2 of IC119 connects to PIN2 on IC127.

You cannot view this attachment.

[SerErris]

And you want to check the service manual - like this copy here: http://www.cpcmania.com/Docs/Manuals/CPC6128%20Service%20Manual.pdf

[mrbee]

Check your PAL conections. On picture you have number PAL pins on other ICs.

Look on IC115. D0-D7 (Pin 2,4,6,8,11,13,15,17) Check join to Pin2 RAMs (D0-Pin2 in IC119 , D1-Pin2 in IC120 etc.)

Ok so just checking through these now and all of the connections seem good but into the ic115 -

for ic115 the continuity results are -

pin 20 of ic115 connects to pin 8 on all ram chips
Pin 19 of ic115 there is no connection to any ram pin. 
Pin 18 of ic115 connects only to pin 14 of ic119 and 1c127
Pin 17 of ic115 connects only to pin 2 of the ic126 and ic134
pin 16 of ic115 connects only to pin 14 ic120 and ic128
pin 15 of ic115 connects only to pin 2 ic124 and ic133
pin 14 of ic115 connects only to pin 14 of ic121 and ic129
pin 13 of ic115 connects only to pin 2 of ic124 and ic132
pin 12 of ic115 connects only to pin 14 of ic122 and ic130
pin 11 of ic115 connects only to pin 2 of ic123 and ic131
Pin 10 of ic115 connects to every pin 16 of every ram chip, also (more on this below) i get a very short continuity beep on the multimeter not continuous) on pin 10 of ic115 to pin 8 of each ram chip
pin 1 of ic115 connects to no ram chip
pin 2 of ic115 only connects to pin 2 on ic119 and ic127
pin 3 of ic115 connects only to pin 14 of ic126 and ic134
pin 4 of ic115 connects only to pin 2 of ic120 and ic128
pin 5 of ic115 connects only to pin 14 of ic125 and ic133
pin 6 of ic115 connects only to pin 2 of ic121 and ic129
pin 7 of ic115 connects only to pin 14 of ic124 and ic132
pin 8 of ic115 connects only to pin 2 of ic122 and ic130
pin 9 of ic115 connects only to pin 14 of ic123 and ic131

Following your points on the board (coloured dots and circles)

the continuity error I get is that on your board picture there is the amstrad 40010 chip but mine is a 40007.

So 7) on the picture (the dark blue dot) the point above r161 connects to pin 7 of ic118 (pal) and to the 40010 chip but mine is a 40007. So for me the connection is from the point above r161 to pin 7 on pal is the same, but it is for pin 3 on the 40007.
for connection 11) the black circle, down to the 40007 chip it is pin 1 for me. so from ic118 pin 11 to 40007 is pin 1.

on 2) you have the red circle (in colourblind so i hope i got these right) on pin 2 on the ic115 and ic114 but then a red circle on the 400010 chip on pin 35. I tested the continuity from pin 2 of ic115 and ic114 and theres no connection to my 40007 chip. I am not sure if I got the colour circle wrong



two strange things going on with the z80 and the ram and the pal chip though..
pin 20 of ic118, the pal chip, i cant see the colour you used for this one anywhere else on the board. it looks like a dark red to me. but anyways, i get a very subtle short beep of continuity (not continuous like it should be) on pin 29 of the z80 cpu, but there is also a long beep of continuity on pin 11 on the z80 to pin one of the pal chip.
This short subtle beep is puzzling me, my multimeter only really gives a continuity beep of just two points connecting, im not sure what this very short subtle beep means..

Continuing on this, I tested the continuity of pin 16 of all ram chips to pin 29 on the z80 and continuity is all good.
This one is puzzling me though. the pal chip ic118 PIN 20....
With my tester always on pin 20 of the pal chip ic118 - I test pin 16 of each ram chip and i get this short beep on the multimeter, not continuous. still on pin 20 of ic118 when i test it to pin 8 of each ram chip i get a full beep of continuity. 
with the tester still on pin 20 of ic118 (pal) i also get a short beep of continuity going to pin 29 of the z80...

So i tried then to see if there are any others of these short beeps from ic118 pin 20 to anywhere else.

I get it on pin 10 of 1c106, 
pin 6 of ic102
pin 7 of ic107
pin 8 of ic 101
pin 7 of ic 110
pin 1 of ic108
pin 8 of ic109, ic104, ic113 and ic105
pin 16 of every ram chip
pin 29 of the z80
pin 10 of the same chip (the pal chip ic118)
pin 10 of ic115 and ic114
on ic 117 i get the short beep on pin 5 and pin 7
ic102 pin 7
on all ic209, 211, 212 and 210 pin
ic208 pin 20
ic203, ic206, ic208 pin 7
ic207 and ic112 pin 7

bearing in mind with this short beep of continuity on those pins, i also get a full continuity beep on the pins of every chip except the pin would be on the opposite corner.
This short beep could be from the multimeter "charging" the circuit or whatever and the beep is the last remnants of that charge coming to an end. But just wanted to be sure to report it incase there is a short. But I figured a long beep would be a short.


Anyways all other coloured dot points all test good for continuity.
It's taken me 2 hours to get through this and type it, as I counted the pins from the top left and across rather than how they are numbered on the board   so i had to recount and test everything   but i hope you can understand it

[McArti0]

Pin20 of PAL is +5V. PIN16 all of RAM ICs is 0V. Why do you tested capacitors on power supply?
This is not geometry, this is electronics. You must understand what you are doing.
On schematic You have Number of PIN and Name of SIGNAL LINE.
Which signal line has a break?
Which signal lines are shorted?

D0-D7 are good.

CAS0-CAS1 ? good?

RAS? etc.

[mrbee]

Pin20 of PAL is +5V. PIN16 all of RAM ICs is 0V. Why do you tested capacitors on power supply?
This is not geometry, this is electronics. You must understand what you are doing.
On schematic You have Number of PIN and Name of SIGNAL LINE.
Which signal line has a break?
Which signal lines are shorted?

D0-D7 are good.

CAS0-CAS1 ? good?

RAS? etc.

Im learning, and I am following the points to check, i will check capacitors next. But someone else said it could be the ram continuity, then you said to check these points, then someone else in a pm said the chips are fine but it could be the pal or connections to it. I know its not geometry. 

[SerErris]

He was just saying if you measure between pin20 of pal an pin16 of RAM ICs you actually measure the capacitors instead of measuring anything meaningful.

So you would need to measure the correct pin on the PAL. and one pin from the PAL goes to bank0 and one pin goes to bank1.

This is painted a little bit difficult to understand in the schmatics.

However pin16 of pal ~CAS0 is going to bank0 pin15. 
pin15 of PAL ~CAS1 is going to bank1 pin15.

So that are the only lines directly connected to the RAM.  connected from PAL to the RAMs. 

You can also check the muxer chips (IC109 etc) 

They need to connect the output lines to the RAM as well and connect them to all RAM chips.

For instance IC109 connects
Pin7 to all RAM pin 9
pin8 to all RAM pin 13

So just trace the schematic from the service manual and see if all the lines are getting connected correctly.

Hope that helps.

[McArti0]

mrbee. Can you identify the name of the signal lines of the tracks you've been repairing? How many are there and what names do they have?

[mrbee]

mrbee. Can you identify the name of the signal lines of the tracks you've been repairing? How many are there and what names do they have?

I had taken off the 40025 chip and put in a socket, this came out nice and clean and the traces seem all good. I will double check these as well to be sure. When I get time I will sit down with this for a day and go through every single trace and point on the board.
The traces that needed repairing were only on the ram, traces that just go vertically up to the next ram chip, there were about 7 traces that needed sorting.

I thought that you meant to do a continuity check as I was trying to do a continuity check to check that I made all of the traces correctly, and there was no shorting.
It's just time that is the thing at the moment, I am only able to look at this whilst repairing other things and serving customers (I usually repair controllers, charging ports, handheld consoles and things like that - retro console repairs mostly things like capacitors or changing out components. Recently recapped and fitted new flyback to a hantarex monitor, recapped and reflowed a ctm644 monitor, changed the tube and things like that. With chips I have removed roms from arcade boards such as turtles and neogeo and added sockets and things like that no problem.

This last week or so though I got hit with an Amiga 600 recap that went dreadful, the pads just fell off of each capacitor trace and so this one has gone into storage for another day! So then I got this Amstrad to do (this is for myself) and it had one single bad ram chip. Upon taking it off, someone had tried to repair a trace on it before, so I fixed the trace and all was well! it booted perfectly.
I bought a diagnostic rom chip, also another chip but with 40025 and 40015 roms on it, sockets for all of the ram wanting to do this to future proof it for myself.

At first i desoldered the 40025 chip thinking this was the chip to attach the diagnostic rom to but I think I was incorrect about this, so I just left it as a socketed chip and was going to add another socket at a later date to be able to change the chips if needed. Then I moved onto doing the ram chips, reflowed every solder joint of the ram pins, solder pumped the solder away and as I go to remove the chips two pads just fell away just like that, and then onto another chip and even though I was very careful, the trace wires of some of the ram chips either werent connecting, or you can see the trace wire ontop of the board all loose. It was like this for about 6 or 7 traces for the ram.

I added the sockets and went on to repairing the traces, using a picture of a motherboard as a guide I believe I got all of the traces sorted and began testing continuity but there were some things that confused me as some were connecting to the chips above, some pins connecting to only the chips in bank 0, some pins then connecting to both banks so I was looking then for help checking that so i can rule out the traces.

Then powering on again, because Im distracted at my shop a lot, I quickly went to power it on but I connected the 12v into a 2 pin connector at the top left of the board (for the led etc I believe) next to the relay. Nothing seemed to have happened, the same black border solid colour screen showed. Then I realised my mistake and turned it off, reconnected the 12v into the correct 2 pin connector and powered on - still black borders and solid colour screen.

So I came here then to ask if connecting this 12v to the wrong 2 pin connector would have damaged it, or is the issue still with the ram traces. Hope I have made sense.

I think my best option is to try and find a dandanator to buy and see if I can rule out things that way.

[Bryce]

Due to your latest unfortunate endeavours, may I suggest that you take some time out to read the following article on how to remove IC's without destroying PCB traces? https://www.cpcwiki.eu/index.php/IC_Repair

Bryce.

[mrbee]

Thank you, although I want to keep all of the legs intact as much as possible 
Theres a video that Noel posted on his channel about the issues he had with the traces even ones that he was very careful with that ended up being bad or lifted up. I had removed ram from a 464 previously as well (there was one that already had sockets on and one without) and those came out just fine, just this particular 6128 and amiga 600 I have.. sooo sensitive! 

[Bryce]

PCB traces are only glued down. The glue melts at around the same time as solder, so while you are messing about trying to remove 14 pins at the same time, the pads will start floating and the traces will break. You are trying to save the CPC, not a broken RAM chip!

Bryce.

[mrbee]

Update so far - I have bought and now connected a Dandanator cpc dual mini to my 6128, performed the diagnostic check and I get the red screen and then the ram check screen with the giant numbers 0 to 7, on each number every bar is red. As if it doesn't see the ram bank at all.

I am looking now into swapping the banks to see if I can get it to see that one bank is working.
I individually tested each ram chip on another amstrad and the ram chips are all good.
Im not sure which points to check now to see why its not seeing any of the ram.

[robcfg]

If the ram chips test good, then you should look at the logic chips that are in the way from the RAM to the CPU.

[mrbee]

I had just swapped the ram banks with the jumper on the hal chip and there is still no ram being seen (all red). What chips exactly would I check? im out of my depth here but am learning as I go to try and see what the next thing to check would be. thank you

[robcfg]

At a first glance over the schematic, I'd check the following ICs:

IC114 - 74LS373
IC115 - 74LS244
IC104
IC105
IC109
IC113 - 74LS153

[Bryce]

At a first glance over the schematic, I'd check the following ICs:

IC114 - 74LS373
IC115 - 74LS244
IC104
IC105
IC109
IC113 - 74LS153

There's no point looking at other IC's until all RAM connections have been confirmed. It's known that traces have been damaged.

Bryce.

[mrbee]

I'll re check all of the ram, so far the continuity is all good, voltage from the z80 is good etc I just need to check exactly where else they go to see why they arent being seen

[SerErris]

He was just saying if you measure between pin20 of pal an pin16 of RAM ICs you actually measure the capacitors instead of measuring anything meaningful.

So you would need to measure the correct pin on the PAL. and one pin from the PAL goes to bank0 and one pin goes to bank1.

This is painted a little bit difficult to understand in the schmatics.

However pin16 of pal ~CAS0 is going to bank0 pin15.
pin15 of PAL ~CAS1 is going to bank1 pin15.

So that are the only lines directly connected to the RAM.  connected from PAL to the RAMs.

You can also check the muxer chips (IC109 etc)

They need to connect the output lines to the RAM as well and connect them to all RAM chips.

For instance IC109 connects
Pin7 to all RAM pin 9
pin8 to all RAM pin 13

So just trace the schematic from the service manual and see if all the lines are getting connected correctly.

Hope that helps.

Did you do the continuity tests from this post above? And what was the result ?

/Edit:
Okay I do think you need to learn to read schematics. It is all in there.

I will write down how you verify conductivity of all RAM chips, so that you can follow:

For all measurements - turn power of the system off!

Start with the power lines (+5V and GND).
Put one of the probes to +5V and then check on every RAM chip (IC119-134) Pin 8 with the second probe. You should have connectivity to all of the chips.
Do the same with GND line and Pin 16 of every RAM chip.

Okay if that is settled, we know that we should have power everywhere.

Now test the data lines

DataIn:
Those come directly from the CPU and connect to every RAM chip to Pin2.

D0: Z80 Pin 14 to IC119 and IC127
D1: Z80 Pin 15 to IC120 and IC128
D2: Z80 Pin 12 to IC121 and IC129
D3: Z80 Pin   8 to IC122 and IC130

D4: Z80 Pin   7 to IC123 and IC131
D5: Z80 Pin   9 to IC124 and IC132
D6: Z80 Pin 10 to IC125 and IC133
D7: Z80 Pin 13 to IC126 and IC134

DataOut:
Those all go to IC115 and IC114 and the GateArray please check all.

Every RAM chip has DO at Pin14.

But the Pins on IC115 and IC114 connect to different RAM chips.

D0: IC114 Pin3 and IC115 Pin18 and GA Pin35 to IC119 and IC127 (PIN14)
D1: IC114 Pin4 and IC115 Pin16 and GA Pin37 to IC119 and IC127 (PIN14)
D2: IC114 Pin7 and IC115 Pin14 and GA Pin38 to IC119 and IC127 (PIN14)
D3: IC114 Pin8 and IC115 Pin12 and GA Pin39 to IC119 and IC127 (PIN14)

D4: IC114 Pin13 and IC115 Pin9 and GA Pin40 to IC119 and IC127 (PIN14)
D5: IC114 Pin14 and IC115 Pin7 and GA Pin1 to IC119 and IC127 (PIN14)
D6: IC114 Pin17 and IC115 Pin5 and GA Pin2 to IC119 and IC127 (PIN14)
D7: IC114 Pin18 and IC115 Pin3 and GA Pin3 to IC119 and IC127 (PIN14)

Address Lines:
The addressing is done via the 74LS153 MUX chips. (IC104,105,109,113)

From the MUX chips one pin connects to all RAM chips - to the same pin.

IC105 Pin9 to RAM Pin5
IC105 Pin7 to RAM Pin7

IC113 Pin9 to RAM Pin6
IC113 Pin7 to RAM Pin12

IC104 Pin9 to RAM Pin11
IC104 Pin7 to RAM Pin10

IC109 Pin9 to RAM Pin13
IC109 Pin7 to RAM Pin9

The last Pins are the control lines ... 

GA Pin33 to RAM Pin3 (Write Enable)
GA Pin34 to RAM Pin4 (RAS)

The last PIN connected is the CAS pin. THis is a little bit more difficult, as Bank1 and Bank0 are conneted differently.

The CAS signal is generated by the Gatearray and switched by the PAL IC118. So the actuall connection goes from IC118 to the RAM CHIPs and from GateArray to IC118.

IC118 PIN17 to IC127-134 Pin15 (CAS)
IC118 PIN16 to IC119-125 Pin15 (CAS)

You should also measure:
GA Pin16 to IC118 Pin7

If all of that has good connection, you then could also next measure for any shorts.

Just go over the RAM chips and check between any two neighbouring pins if the connect somehow. That also should not happen.

And if all of that measuring does not indicate any broken connection, we are then into a logic issue.

So if the RAM chips have been tested in another system without any problem, then there is something wrong with the logic.

That includes the following chips:

PAL IC118
74LS153 Muxers IC104,IC105,IC109,IC113
The RAM Read data latch IC114
The RAMD_DIS Bus gate IC112.

You can check if you have a voltage on in10 of IC112. That would disable RAM reads to the CPU completely. 

But to be honest, logic checking should be done only after all connectivities are ensured.