CPC6128 - Boot pb , wierd caracters

[Philabs]

Hi All,
Thanks for this great website and all Amstrad fan's helping vintage gamers as us.
In advance , please excuse my poor english ,

So This weekend i tried to start my old Amstrad again, i read many topics , saw long videos ( Oscilloscope memory test from Noel's RL ) etc without success.
I need some help , if possible.

First Visual PB: A Ceramic capacitor 104 , on bank 2 last one down was half broken , i changed it.


So what i did :
- 104 CHANGED
- 5volts TESTED
- New Power Supply (external) TESTED
- CTM644 TESTED (got an another mint CPC6128 working fine on it (i don't want to open it for now)
- Power Switch TESTED
- Z80 TESTED and also SWITCHED with a part one
- I Re-soldered all the board
- Caps kit made ( except the Bi polar one , and also i don't have, and , two 1Uf 50v changed by 1Uf 250v)
- R160 - R161 tested and some others not all the board , but the oscilloscope says activity.
- Continuity between RAM themselves and also PAL junction etc TESTED
- RAM Activity checked with Logic Probe and also with my Little China Oscilloscope SEEMS OK (same waves as on the Noel's RL Videos)- PAL Trick trying to start on second RAM bank TESTED (same glitch matters appears)
Things wierd:
I dont have basic prompt or only 1 time on 50 tests.Sometimes the" white box Relay" Ry301 seems to panic and i got blank screen

In my opinion it should be RAM or BASIC chip , what's your 's ?Thanks in advance
Phil

[Bryce]

One of the ROMs is failing.

Bryce.

[Philabs]

Thanks for your reply Bryce,

So two more questions :
The million dollars one ; which CHIP ?  I saw on the forum a kind of External diagnostic board , is it for sale somewhere ?The 50Cents one : I'll soon receive an external floppy emulator from Zaxxon , is there a way to automaticaly start a kind of test ?
ThanksPhil
Oups, ROM , ( i read RAM ) , so the Million Dollars Question changing , which one ? 40010 ? 40015 ? 40051 ?
Thanks again

[Bryce]

It's difficult to say which one has failed because they are (almost) completely wired in parallel. The 40051 is a 27C256 containing the Firmware and BASIC. 40015 is a 27C128 and contains AMSDOS.

The 40010 is a gate array. If this fails the computer rarely boots as far as yours is.

Bryce.

[Philabs]

Ok thanks
I'll may order dip sockets ,  they are 28 Wide (not narrow) ? right ?will i be able to find both ROM chips on ebay ?

Also , i have a strange value on the Resistance network NR101 , the third pin value is 1.12 and not 2.2Kvalue.
All pins are 2.2 except the 3Th one see picture ... normal ?

[Bryce]

The sockets are 15.24 wide sockets. The empty EPROMs (27C128 and 27C256) can be bought in many places and also on ebay, but you'll need to program them yourself. The files are available on the Wiki. I wouldn't worry about NR101. These are mainly pull-ups for the option settings and have no further effect once the machine has booted (other than one that handles the Printer Busy signal and one that handles the (unused) /EXP signal. A variance in value is caused by an option jumper being in place and even at 1.12K it would still work correctly.

Bryce.

[llopis]

I agree it's probably a ROM, but it could be anything that is intermittently screwing up the data/address bus.


If it's a ROM, the easiest way to test it would be by putting some kind of external ROM board (X-Mem, Dandanator, etc).

Alternatively, if you have an EPROM writer, you could take the ROM out, read it, and compare the binary to what it's supposed to be (they're all in the ROM page in the wiki).

[Philabs]

The sockets are 15.24 wide sockets. The empty EPROMs (27C128 and 27C256) can be bought in many places and also on ebay, but you'll need to program them yourself. The files are available on the Wiki. I wouldn't worry about NR101. These are mainly pull-ups for the option settings and have no further effect once the machine has booted (other than one that handles the Printer Busy signal and one that handles the (unused) /EXP signal. A variance in value is caused by an option jumper being in place and even at 1.12K it would still work correctly.

Bryce.

Thanks again
I put two tulip 28pin socket in place of both ROM and find an old programer 32pin with a 27c256 , no 128 in stock.Will try to program one for test.

[Philabs]

I agree it's probably a ROM, but it could be anything that is intermittently screwing up the data/address bus.


If it's a ROM, the easiest way to test it would be by putting some kind of external ROM board (X-Mem, Dandanator, etc).

Alternatively, if you have an EPROM writer, you could take the ROM out, read it, and compare the binary to what it's supposed to be (they're all in the ROM page in the wiki).

Thanks Llopis

It's a good idea !  , i'll try to do it with my "ETETRON xprog USB V2 ET-A015" , need to learn and find how to use it before 

[Philabs]

So so so , i dumped both chips to check , i think both are correct , could someone confirm pls , you'll find the 2 dumps here, it is the french version azerty cpc 6128. 1985

The bad news: chips arent the pb ...
Any idea ? i saw on some topics the use of an external diag cartridge ....?
Thanks
Edit:  seems an error with the 40051 dump , could someone confirm, the adress 00000400/04 , thanks

[Bryce]

So you've tried a new EPROM, or are you just deciding it's not the problem because the data looks good? I have had readable EPROMs that still didn't work in a computer because they couldn't drive enough current on the data pins.

Bryce.

[Philabs]

So you've tried a new EPROM, or are you just deciding it's not the problem because the data looks good? I have had readable EPROMs that still didn't work in a computer because they couldn't drive enough current on the data pins.

Bryce.

So except the fact i found an little error in 40051 dump.


I have to test new eproms , i tried to erase eprom with my wife nails machine but it does not works and no blanks eprom here so i ordered an eprom eraser from Ali and by the way new Ram chips if the pb move to.


I can have the right text with the prompt on screen 1/20 tests , that could confirm chips with low current on data pins.


Today i'll try a 10A external Power supply to check , maybe the screen and my little power supply don't have enough Amp for my old Amstrad ( while my mint Amstrad seems works on , i dont have any disc for real try ) and i'll check all Resistors and diode with care.


I also ear sometime strange "clicks" on the white box before a blank screen.


Is there a particuliar component on the board it could reduce power Amp / voltage like a filter or a transistor ?

[Bryce]

Any part with a short will pull the 5V rail down, but also a bad solder joint could cause it too. It's highly unlikely that it's a RAM problem as RAM is an "all or nothing" failure: If one lower RAM chip fails, the entire 64K bank has failed. Have you also cleaned and reseated the 40010 and other socketed chips?

Bryce.

[Philabs]

Any part with a short will pull the 5V rail down, but also a bad solder joint could cause it too. It's highly unlikely that it's a RAM problem as RAM is an "all or nothing" failure: If one lower RAM chip fails, the entire 64K bank has failed. Have you also cleaned and reseated the 40010 and other socketed chips?

Bryce.

Here a little video

[Bryce]

Wow, that's random. You don't see that often. My guess would be something like a failing 74LS153 or CRTC.

Bryce.

[Philabs]

Wow, that's random. You don't see that often. My guess would be something like a failing 74LS153 or CRTC.

Bryce.

Grrrr hard to fix , thanks again for your help Bryce.
Today , i'll try to unsolder the 6845 and put a dip socket.
Is this chip will be ok to remplace it : ?https://fr.aliexpress.com/item/33005420859.html?spm=a2g0s.9042311.0.0.7d266c37wfyYyg
Also , is there a way to test some 74ls153 (from an old Arcade pcb board out of service) without desoldering old ones ?
Thanks again
Phil

[Bryce]

Yes, that's the right chip, but they are usually quite easy to find in Europe for a few Euros more. The only way to properly test the 153 is to monitor the inputs and outputs with a logic analyser or scope.

Bryce.

[Philabs]

Yes, that's the right chip, but they are usually quite easy to find in Europe for a few Euros more. The only way to properly test the 153 is to monitor the inputs and outputs with a logic analyser or scope.

Bryce.

Bad newsI was Looking for a pb with bus signal and found a strange thing on the 74HCU04P ( ic117 ) and the clock.
On my working CPC ( i finaly opened it ) , i found different value:
1) The Left pin of the clock is 14.9mhz on the bad cpc and 8mhz on the good one
2) On the 74HCU04P : 

On the Working CPC : Vcc   6A              6Y          5A         5Y               4A         4y
                                   5.5v  Nothing     8Mhz      8Mhz    14.9Mhz      4Mhz     4Mhz


On the Bad CPC :       5.5v Nothing      14.9Mhz 14.9Mhz 14.9Mhz    4Mhz     4Mhz       


--> Pb appeared  on Pins 6Y and 5A with 14.9Mhz in place of the 8Mhz

I unsoldered the 74HCU and looked for a new one , and finaly found a HD74LS04P on an old Arcade PCB

http://www.datasheetcatalog.com/datasheets_pdf/H/D/7/4/HD74LS04.shtml
I placed it , BUT BLANK SCREEN , and not any clock frequency .... BAD...
I deceided to place again the original one , and NO MORE PICTURE , STILL BLANK SCREEN...

New Diag :
40010 Ok
Z80 NO ACTIVITY at all , receive VCC  , Replaced , but no change
All IC Have VCC74HCU have sometime Mhz , sometimes not , sometime right values sometime nothing
Now i'm so afraid to use good CPC parts for test...

                 

[Bryce]

Two importrant things:


1 - Unless you have an expensive oscilloscope, you can't accurately measure the clock signal. You were most likely measuring a harmonic or an aliased signal.

2 - A 74LS04 is NOT a 74HCU04. It must be a HCU part.

The original HCU was most likely ok. Unfortunately, you've now started doing more harm than good.

Bryce.

[Philabs]

Two importrant things:


1 - Unless you have an expensive oscilloscope, you can't accurately measure the clock signal. You were most likely measuring a harmonic or an aliased signal.

2 - A 74LS04 is NOT a 74HCU04. It must be a HCU part.

The original HCU was most likely ok. Unfortunately, you've now started doing more harm than good.

Bryce.

Yes i know...I'll wait parts from Ali express like Dip sockets and IC, and then if i don't find the matter will probaly look for somebody to repear it.
Thanks for your helpI'll keep you informed
Phil

[Bryce]

I can repair it if you like, but please don't replace any further parts, you may be just making the repair more expensive.

Bryce.

[Philabs]

I can repair it if you like, but please don't replace any further parts, you may be just making the repair more expensive.

Bryce.

Nice , from where are you ?
I can see a French flag but ?

[Bryce]

The flag is just me messing about. I'm actually in Germany.

Bryce.

[Philabs]

The flag is just me messing about. I'm actually in Germany.

Bryce.

Ok Bryce, I'll wait parts from Ali express like Dip sockets and IC, and then i'll contact you.I'll may send the both Amstrad , the mint one may have a matter with the floppy drive , because the light still on. (have no floppy for test) but saw this kind of pb on forum.
CU

[Bryce]

Just send me the bits that don't work, ie: the PCB and/or floppy drive to save on postage. Send me a PM for details.

Bryce.