Amstrad pc 1640 black screen, any advice?

[VintageVic]

So, more dead stuff has arrived.

Had this untested Amstrad PC 1640 for a year or two and now had my hands on another untested Amstrad PC-ECD monitor that
should work with this PC. If the monitor is actually working. But as the topic says, result is black screen.

I do however have a working IBM 5154 EGA monitor that I managed to fix some years ago. I applied that and still black screen.
IBM monitor is actually giving bright gray screen when the Amstrad has no power, and when power is given to the Amstrad motherboard,
EGA monitor goes black. So there's that at least.

I tried ISA post analyzer. It's alternating between some codes, but quite quick, I can roughly assume that two codes are
0480
8400

But not 100% sure. mobile camera doesn't give perfect answer to this either.
I haven't done anything to the motherboard just yet, other than reseated EGA ROM and the other couple ROM IC's and the CPU in case
of bad contact to the IC sockets. No help there.

I did notice with the ISA post analyzer that 3.3V was not present at first, but came present after a bit of a delay, while I kept power on.
Might suggest dried up electrolytic capacitors, no ? There is a bunch of small caps near RAM IC's.

Have any of you guys had a black screen Amstrad 1640 that you managed to revive? Where would you start?
My guess is first on the list are those old el.caps.

One more thing; I kept power on for a couple of minutes to see if any IC gets warm. The CPU picks up heat pretty fast. I hope
it's not fried - I still think not, because the post codes are changing. I figure, it might be stuck on some loop which might be the
reason for it getting hot.

 

[VintageVic]

all right,
I had my dip switches se backwards. On was Off and viceversa.
Motherboard seems to be alive! No reason to not replace those aging elcaps anyways.

But, the Amstrad ECD display remains black screen! That is a bit of a bummer.

 

[modem7]

I did notice with the ISA post analyzer that 3.3V was not present at first, but came present after a bit of a delay, while I kept power on.

For later use of that particular model of card, note that per the 'Example of a modern POST card' section of [here], when the card is in an ISA slot, the 3.3V LED is monitoring the -5V rail.

 

[VintageVic]

Some progress, I managed to bring the ECD monitor back to life. Wrote about it here:

V

Thread 'Amstrad PC ECD display "Black screen" - advices welcome.'

Jun 18, 2026

And so, another challenge. I'd really like to revive this black screen monitor for the Amstrad PC1640 - that used to be ECD, Enhanced Color Display.

Here's what I know so far.
Powers on in such a manner that provides needed power to the PC1640 - which then does POST (checked and tested with another working EGA monitor).
Screen remains black at all times, even if waited for several minutes.

Touching the screen with hand will result static electricity chatter, what you'd expect.
When powered off - briefly blinks.

So. How to go on from here.
Recapping the whole thing? Maybe starting...

• VintageVic
• Replies: 4
• Forum: Vintage Computer Hardware

Next step would be to boot the unit, but I'm stuck at the imfamous 'check keyboard and mouse'
-error. I have both plugged in.

Pressing caps lock or scroll lock at keyboard briefly lids the leds on those keys. So there is some life at least.
I know I'm not the first to experience this issue. Any suggestions for this?

 

[VintageVic]

Great,

I passed the keyboard issue just by opening the keyboard and giving it a thorough cleaning.
It had really a lot of dust and grime inside. blowed the dust away and wiped the plastic contact... 'thing'
or whatever it is, clean with wet table rug. and it did the trick!

Next, it would be nice to get this thing to boot to ms-dos. Of course, not so nicely, hard drive
and floppy drive has been removed before the unit was sold to me. I've tried with a couple of 360kb
drives and tried with several dos floppies, but no luck booting so far.

Then I realized that IBM pc type floppy drives may not be directly compatible with the Amstrad floppy pinout.
I tried to set one floppy drive to DS0 from DS1 so that has been tried. There's something else. spindle is spinnig
and the drive led is working, but the drive head is not moving. I figure, that drive head should do some back and forth
movement every time the pc is started and it's not doing that.

So, I guess it's time to study amstrad floppy drive pinout and compare it to the ibm pc one.

 

[VintageVic]

Oh yeah.

Indeed, it appears Amstrad 360kt floppy cable pinout differs from ibm pc.
I have to draw necessary modification on the paper and make adaption to the amstrad pc 1640 cable
for it to support ibm style floppy drives.

 

[VintageVic]

Made a plan to adapt Amstrad cable and then implemented it.
But a bit to my surprise, nothing changed.

If there is a disk in the floppy drive, spindle keeps spinning. But no head movement at any
time during the boot.

Screen says "Insert a SYSTEM disk into drive A, then press any key".
Amstrad is not stuck, pressing a key results a short delay and then this is asked again.

I'm a bit puzzled now. Maybe there is faulty IC at the mainboard. But before I go replacing
any of those, I better check that these floppy drives do work and boot on ibm pc.

 

[daver2]

What do you observe on the /drive select A, /TRACK 0, and /INDEX signals when the drive is running?

Dave

 

[VintageVic]

Dave,

this was a great suggestion! Because it lead me to think where do I get correct measurements
for each signal. After a few checks and double checks I realized my floppy cabel is upside-down.
Doh!

And when I installed the cable in correct orientation, my result is "Error: Faulty floppy disc controller
or disk drive". I believe I saw this earlier and switched the cable upside down - and as it seemed to
improve things, I wrongfully believed the cable being set up correctly.

Okay, I know there is another long thread about this same error. I figured, I might read through that
topic if it helps. But there were a few checks I could make first. I tested with two of my 5.25" floppy drives
and got same result. I put the third drive, it also failed, but then I switched that drive to DS0 and
voila! It booted!!

This is really great. Next for the project is to build some kind of braket for the floppy drive.
And then I will go on trying if I have any ide controller that might work with 8bit isa bus. I got CF
card ide adapter already and some 256 meg cf card, from which I might then try to partition a 30meg
drive...

But the main thing is, this thing boots! Even without a hard drive, it can do fun stuff from here!

 

[daver2]

Yes, you either need DS0 for drive A and DS1 for drive B (if the cable doesn't perform the drive select swapping) or DS0 for both drives (if the cable contains a twist for the drive select swappping).

Dave

 

[orac81]

I had an Amstrad pc1512sd that had a 720k dd floppy as a B drive working fine for ages. I thought the amiga drives would also work. Yes you need the cable with the twist for the disk select.

If swapped it would also boot as A, either a 720k disc or an image of a 360k on the first 40 tracks. Be careful using 1.4m hd disks, I think the 720k has a lower erase capacity and may not be able to wipe the old 1.4m recording properly.

 

[VintageVic]

I could add to the above, that the Amstrad pc 1640 has straight cable, which has no twist.
Also, Amstrad pinout is Shugart interface standard, not the standard IBM PC pinout, so standard twisted PC cables will not work.
Without modification that is, I modified my cable as I drew picture a few posts back.
Also, Amstrad cable is fixed, soldered on the motherboard. Thus one cannot just easily change the cable.

 

[VintageVic]

Assembled the unit and tested some media I found on some old untested 360k floppies
I had lying around. Colors and graphics work great with the monitor, I'm very pleased with it.

All internal brackets were gone, I had to improvise something to hold the floppy drive in place.
Also missing back / top lid above the extension cards.

Added a reset button (+5V -> cpu pin 21), but I think I will add 10k resistor to that before using it much.
Also put 4x AA batteries, but date and time settings will still reset every time the Amstrad reboots?
heh, Amstrad refuses year 26 anyway.

Hard drive solution is still not implemented, but something will have to be done about it.
It's very inconvenient to use like this. And super slow, for example that Operation wolf I tried,
it's in two (or three) floppies, and asked to change floppy 4-5 times before the game could start .

Good old times. Lovely machine!

 

[orac81]

An interesting short term solution for an HDD for XTs is to bitbang an SPI interface with the parallel port to connect an SD card:

MicroProject - MicroSD to LPT Printer Port \ VOGONS

www.vogons.org

Device driver:

https://www.vogons.org/download/file.php?id=53412

Its only a bit faster than floppy, but it can be easily plugged into different PCs, useful for experimenting.

As for cables, now you say it, you maybe right I don't remember the exact setup on my pc1512, I might have set a link on the 3.5 drive for A/B on it. I think it had a little 3.5/5.25 adapter dongle. On later 386s I did use the twisted cable, with 4 connectors, 2 for 5.25, 2 for 3.5. It was a long time ago!

 

[VintageVic]

Curious little thing, Orac81.

But I still prefer a hard drive solution that can be booted from. I have a number of possible solutions, but
first, I needed to implement XT-IDE rom.

I have had my challenges with this and so I did again. But some notes for someone else (or myself later when I have forgotten solutions).
- I used 3Com509B - C network adapter for a ROM socket.
- As for the ROM IC, Atmel at28C64 - 8kb eeprom.

I programmed the eeprom and put it into 3com socket.

First try, the PC does not recognize any additional ROMs while booting.
Problem was, 3com adapter does not handle AT28C64 |WE pin (pin number 27) correctly.
Pin must be set to +5V or eeprom write enable is enabled and thus the IC will not be in read mode.

Correction for the issue: I shorted 3Com rom socket pins 27+28 at the back of the network adapter card. This
will give VCC to pin 27, WE. There was no need for any other changes or socket adapter.

Now, the ROM was recognized, but during the boot I got error message:
"External ROM checksum incorrect"
and the XT-IDE ROM was not loaded.

I tried a number of troubleshooting issues with this one, rewrote the ROM and double checked that all seemed to be ok.
But I only got it solved, when I used another AT28C64 ROM IC !

I am still not 100% sure why the first IC did not work. These are cheap chinese and maybe/probably used ICs.
Maybe the first ROM has failed byte somewhere in the 8kb range - but it is not in the beginning or at the end
of the ROM data, I checked both ends. Or maybe, because I used software write protect enabled in the first IC
but not the second. I will try if disabling this will make the first IC usable. If not, then I suspect the first AT28C64
is failed chip at some memory address.

But with the second eeprom IC, I got nice XT-IDE menu displayed!
Haven't implemented any hard drive solutions, but I will try those next.