CPC 464 dead

tyranthr · 06:14, 24 September 24

Hi
I'm recently bought an untested CPC 464, and guess what? It's faulty!!

I was trying to find any tips for my problem on the internet but no luck. Anyway to the point:
I connected the 464 to a GT-65 monitor. When i power on the 464, LED light instantly turn on for half second (maybe 1/3 of a second) and then off. After that no power, no LED, nothing. So i used a multimeter and the result was:

When i connect the monitor with the 464 and power on the monitor only, BEFORE turning on the 464, my readings was 5,2V on 464 power connector and 5,2V also on motherboard's tape connector (pin 1 and pin 3).

When i power on the 464, LED instatly turn on and then off, as i mentioned. After that my multimeter readings on 464 power connector and Tape connector was zero! So i suspect some kind of short somewhere.

Because i don't want to desoldering and resoldering everything to find the suspect and i don't have any capacitors, transistor etc in stock, do you have any experience with this kind of problem?

ps: There is no the common Power switch problem, i bridge the pins 1-3 on Tape connector to power on with the same results. I' ve also changed 47 Uf capacitor with no luck either!

Thank you in advance!

McArti0 · 08:48, 24 September 24

Take out the chips you have in the sockets. Z80 and test the board, then GA 40007 or 40010 and test the board.

tyranthr · 09:31, 24 September 24

Thanks for your answer. I'll do that and i'll be back!

McArti0 · 09:37, 24 September 24

Also try disconnecting the tape recorder.

tyranthr · 09:52, 24 September 24

Quote from: McArti0 on 09:37, 24 September 24Also try disconnecting the tape recorder.

It's already disconnected. I try to boot it with the motherboard only shorting the pins 1 and 3!

retro space · 10:37, 24 September 24

First try a separate power supply. I would not expect the LED to turn off because of the system running into an error, but because the power supply is bad. Usually power LED's are put straight between the 5V and GND with a single resistor.

tyranthr · 10:55, 24 September 24

Quote from: retro space on 10:37, 24 September 24First try a separate power supply. I would not expect the LED to turn off because of the system running into an error, but because the power supply is bad. Usually power LED's are put straight between the 5V and GND with a single resistor.

I'll do it today. Unfortunately i have a 5V 1.2A power supply, but i think it will power the mainborad just fine considering that the tape recorder is disconnected to draw more current!

McArti0 · 11:11, 24 September 24

In the Amstrad monitor the power supply is short-circuit proof. It is unlikely to break down in this regard.

Use an ohmmeter to measure the resistance across the 464.

Bryce · 12:58, 24 September 24

Just confirming McArti0, the regulator in the monitor has a thermal shutdown, so if the CPC is pulling too much current it will switch itself off. However, it's a bit strange that the LED turns on at all, because a short like that would most likely pull so much current, that the LED would get no current. However, the delay might suggest that some output bit is shorted to ground and only when the CPC starts booting and the bit gets set to "1" does the short actually happen and bring the PSU down. So the failure might not occur when the CPU is removed, because the computer won't boot then and the chip with the short never sets the bit with the problem to high.
It could also be an issue with the monitor. If certain parts of the 5V regulator are damaged it manages to show 5.2V, but even a small amount of current draw overheats it and it shuts down even though the CPC wasn't drawing more than it should. Have you any way of testing the 5V output or powering the CPC from a different PSU? In idle it only pulls around 700mA as far as I can remember.

Bryce.

tyranthr · 17:44, 24 September 24

Ok I have some progress. I' ve managed to power on the computer using an 5V 1.5A external power supply. BUT I have no picture and some chips are burning a lot!! I post a photo with the hot chips.

tyranthr · 05:33, 25 September 24

Update:
I measured the voltage with a multimeter and i have 3.1V all over the mainboard including RAM and power plug after power on the machine. I think the first suspect is a bad RAM, so i've ordered 8 4164B to replace them and i cross my fingers to solve the problem. The only thing that troubles me is that the ROM is burning hot too and maybe the issue is not the RAM only. I forgot to mention that the other IC's are quite cool or slightly warm.

McArti0 · 06:54, 25 September 24

You can cut the power pin8 on all memories and see if the symptom goes away.
Rom pin28 +5v
P.S. Note that you skipped removing the z80 and GA from their sockets.

tyranthr · 07:35, 25 September 24

Quote from: McArti0 on 06:54, 25 September 24You can cut the power pin8 on all memories and see if the symptom goes away.
Rom pin28 +5v
P.S. Note that you skipped removing the z80 and GA from their sockets.

Oh sorry! I forgot to mention, i removed the Z80 and GA, but no change. Same symptoms!

Bryce · 08:32, 25 September 24

Sounds like someone may have powered the CPC from a 12V supply. You may have a lot of chip swapping ahead of you.

Bryce.

retro space · 10:45, 25 September 24

Quote from: Bryce on 08:32, 25 September 24Sounds like someone may have powered the CPC from a 12V supply. You may have a lot of chip swapping ahead of you.

Bryce.

Is there a simple way to make some sort of circuit that prevents killing the whole board if the connected voltage is too high?

eto · 11:58, 25 September 24

Quote from: retro space on 10:45, 25 September 24Is there a simple way to make some sort of circuit that prevents killing the whole board if the connected voltage is too high?

The channel has lots of videos around repairs of CPCs and is a great resource.

The "usual" ICs that get fried from overvoltage are the ROM and the RAM. They often fail first and then act like a fuse that protects the rest of the computer. You could start with the ROM and see if there is any change and if you at least get a square grey image on the monitor.

But as Bryce already said, it might require some IC swapping to find the culprit(s).

tyranthr · 15:57, 25 September 24

Yes Amstrad seems a little bit fried!
I made more progress. I desolder all the RAMs and ROM and Voltage is back to normal! Also when i connect the motherboard with the monitor, led light goes on and screen illuminate. No boot of course without RAM and ROM but i'm close i think!! Fingers crossed!
Just waiting for the RAMs to come and i'm searching a 40009 ROM also!!

eto · 16:58, 25 September 24

The ROM image can be burned to an EPROM. If you don't have an EPROM programmer I'm sure someone can help out. Due to shipping costs it probably makes most sense from the same country.

McArti0 · 17:15, 25 September 24

Now you can confirm that GA and CRTC are working.

retro space · 17:18, 25 September 24

Thanks. The diode lets 7V through. Is there a variation that works on a lower voltage, like 6V?

McArti0 · 17:42, 25 September 24

this is not a proposal. I'm boosting offtopic

https://www.pcbway.com/project/shareproject/W282713ASE22_eprom_usb_emulator_sch_V1_2_ef62fcb1.html

https://github.com/wickerwaka/PicoROM

https://sidecartridge.com/products/sidecartridge-rom-emulator/

tyranthr · 18:10, 25 September 24

For a little search i did, it's a little bit difficult to find a ready 464 EPROM. No to mention an "authentic" Amstrad 40009!! So any recommendatinos or web pages where i can buy one will be much appreciated!

Bryce · 19:54, 25 September 24

Quote from: retro space on 17:18, 25 September 24Thanks. The diode lets 7V through. Is there a variation that works on a lower voltage, like 6V?

What diode lets 7V through???

Edit: Ok, just watched the video. I'm not convinced of the method used. In practice that diode would burn through in seconds and before you realise your mistake your CPC would get fried anyway. Normally a Zener Diode is used in conjunction with a resistor to limit the current, but Noel didn't do that for some reason. There are circuits designed to do this type of protection, either an additional 5V regulator inside the CPC (which Amstrad should really have done in the first place, but didn't) or a "Crowbar" circuit.

Bryce.

retro space · 20:11, 25 September 24

Quote from: Bryce on 19:54, 25 September 24
Quote from: retro space on 17:18, 25 September 24Thanks. The diode lets 7V through. Is there a variation that works on a lower voltage, like 6V?

What diode lets 7V through???

Edit: Ok, just watched the video. I'm not convinced of the method used. In practice that diode would burn through in seconds and before you realise your mistake your CPC would get fried anyway. Normally a Zener Diode is used in conjunction with a resistor to limit the current, but Noel didn't do that for some reason. There are circuits designed to do this type of protection, either an additional 5V regulator inside the CPC (which Amstrad should really have done in the first place, but didn't) or a "Crowbar" circuit.

Bryce.

Do you have examples? The crowbar circuit? The zener+resistor?

Bryce · 08:18, 26 September 24

Quote from: retro space on 20:11, 25 September 24
Quote from: Bryce on 19:54, 25 September 24
Quote from: retro space on 17:18, 25 September 24Thanks. The diode lets 7V through. Is there a variation that works on a lower voltage, like 6V?

What diode lets 7V through???

Edit: Ok, just watched the video. I'm not convinced of the method used. In practice that diode would burn through in seconds and before you realise your mistake your CPC would get fried anyway. Normally a Zener Diode is used in conjunction with a resistor to limit the current, but Noel didn't do that for some reason. There are circuits designed to do this type of protection, either an additional 5V regulator inside the CPC (which Amstrad should really have done in the first place, but didn't) or a "Crowbar" circuit.

Bryce.
Do you have examples? The crowbar circuit? The zener+resistor?

https://en.wikipedia.org/wiki/Crowbar_(circuit)

https://www.electronics-tutorials.ws/diode/diode_7.html

There are also other methods to restrict voltage levels such as clamp circuits, but the Crowbar is the most practical and safe.

Bryce.

News:

CPC 464 dead