Solved - A4000cr Rev d - green screen

[screemo]

Hello,

I need some pointers fixing a board I got as a defect.

It displays green, blinking and then green again. Sometimes it flashes yellow also, but goes back to green.

Capslock blinks twice if I do ctrl+amiga+amiga, but caps lock keeps working if I press it 20 or so times.

I have tested connections on the ram okay now (some previous repair attempt was probably done with a blowtorch..)

Chips marked with red in the attached picture was replaced with new ones.

I would be really grateful if someone might have pointers where to go next?

 

[vibros]

Hello,

i do not know this revision of board; i can only suggest a very close inspection of the entire motherboard. Alice is related to the chip ram.

Regards
Stefano

 

[screemo]

Thanks, I will look closer at the signals on Alice.

 

[vibros]

Hello,

you can look for ALICE and LISA if clock signals (with datasheets in your hand) and all the RAS and CAS lines are there. If you don't own a scope, you can try to measure with a multimeter with, on an Amiga, a comparison against this value:

- a stable value of about 5 volt, means HIGH
- a stable value of about 0 volt means LOW
- a mid value of about 1.5-2 volt means or HIGH IMPEDENCE (look at the datasheet if there is this possibility) or some activity on the line; the multimeter is to slow to show the real value of the line instantly and it show a mid value; this is typical of a clock signal.

This technique can help in some situations.

Regards
Stefano

 

[screemo]

Very nice detailed description, I will measure tonight to see if I can spot anything wrong in the signals.

Thanks for the pointers [emoji106][emoji16]

 

[tbtorro]

Hello,

you can look for ALICE and LISA if clock signals (with datasheets in your hand) and all the RAS and CAS lines are there. If you don't own a scope, you can try to measure with a multimeter with, on an Amiga, a comparison against this value:

- a stable value of about 5 volt, means HIGH
- a stable value of about 0 volt means LOW
- a mid value of about 1.5-2 volt means or HIGH IMPEDENCE (look at the datasheet if there is this possibility) or some activity on the line; the multimeter is to slow to show the real value of the line instantly and it show a mid value; this is typical of a clock signal.

This technique can help in some situations.

Regards
Stefano

Nobody has proper data sheets for neither ALICE nor LISA. CAS and RAS lines strobe too fast to be meaningfully measured with a multimeter, and after displaying the color code the system will be in a halt state, so there will be activity there only for a short while (if at all). A scope is a must (and minimum IMHO) for diagnosing these kind of faults.

 

[vibros]

Hello,

i mean schematics of Amiga 4000D, with the pinout of ALICE and LISA; then the user can argues where are the clock signals and the RAS and CAS lines for the chip ram (related to the green screen showed by the faulty Amiga).
What you have wrote is clearly right!

Regards
Stefano

 

[screemo]

The board works now [emoji23]

Luckily I had access to very competent individual here locally (you know who you are [emoji16]).

For helping others, it was a broken cia chip (U350) and also a 74hct174 (U354) near the zorro slot.

There's no guarantee that your board has the same faults as these green screen errors is caused by so many things.

Only real way is to diagnose and not swap parts blindly [emoji6]

 

[tbtorro]

Hello,

i mean schematics of Amiga 4000D, with the pinout of ALICE and LISA; then the user can argues where are the clock signals and the RAS and CAS lines for the chip ram (related to the green screen showed by the faulty Amiga).
What you have wrote is clearly right!

Regards
Stefano

It's absolutely NOT guaranteed that green screen is related to actual chip RAM components themselves. If the ROM can't read back the correct pointer after it's (very basic) chip RAM test, it will put up the color code, but any part of the system that's hooked on to the address/data bus can cause this not only the RAM chips themselves, could be RTC, Paula, Bridgette the list goes on, or in this case a faulty latch and CIA.

(Btw, you can go ahead screemo and say it that I fixed your board - it's not a secret

 

[screemo]

Hello,

i mean schematics of Amiga 4000D, with the pinout of ALICE and LISA; then the user can argues where are the clock signals and the RAS and CAS lines for the chip ram (related to the green screen showed by the faulty Amiga).
What you have wrote is clearly right!

Regards
Stefano

It's absolutely NOT guaranteed that green screen is related to actual chip RAM components themselves. If the ROM can't read back the correct pointer after it's (very basic) chip RAM test, it will put up the color code, but any part of the system that's hooked on to the address/data bus can cause this not only the RAM chips themselves, could be RTC, Paula, Bridgette the list goes on, or in this case a faulty latch and CIA.

(Btw, you can go ahead screemo and say it that I fixed your board - it's not a secret

I will remember that, thanks again [emoji106][emoji6]

 

[vibros]

Hello,

sorry, but a green screen is, from the Time Amiga appears on the market, caused by failure on the chip ram; to be more precious, the lowest 256 kbytes are the cause of this color on screen (green!). This is not necessarily caused by Agnus/Alice, but it is a faulty on chip ram...and i know only Amiga that do so! I have wrote about Alice and RAS/CAS signals because was the first thing that came in my mind...

Cheers
Stefano

 

[tbtorro]

Hello,

sorry, but a green screen is, from the Time Amiga appears on the market, caused by failure on the chip ram; to be more precious, the lowest 256 kbytes are the cause of this color on screen (green!). This is not necessarily caused by Agnus/Alice, but it is a faulty on chip ram...and i know only Amiga that do so! I have wrote about Alice and RAS/CAS signals because was the first thing that came in my mind...

Cheers
Stefano

Sorry but no, green screen is NOT a 100% indication of fault of the RAM chips themselves. Rather ANYTHING that prevents the ROM of correctly accessing the low part of chip RAM, which can be many things that operates on the BUS, as said CIA, Paula, BRIDGETTE, programmable logic (U213), RTC etc.. All can cause green screen. On top of this if there was battery leak then even open traces can cause it too, even if the RAM chips are OK.

In the case of THIS board, the previous owner thought the same, green screen = MUST be chip RAM components. WRONG! So he replaced all the RAM chips and in the process burned the board because didn't have the skills. It was NOT caused by RAM chips but faulty CIA and latch that operates on the same bus. So bottom line, people who don't know how to do diagnose signal level, should not try to repair boards causing more damage.

To put it simpler so you can understand:

Green screen = chip RAM component fault: WRONG! Only sometimes.
Green screen = chip RAM access fault: CORRECT! Always.
Caused by which componnent: Always RAM chips WRONG! (only sometimes)
Caused by which components: Any component that operates on the data/address bus
Or: Any traces open/short on the data/address bus (even if ALL components are OK)

I hope it's clear now.