Commodore 128 - Very hard to repair!

[demonlg]

Hi to all,

2 year ago i have turned on my C128 and after 3 hour of gaming in C64 mode i have turned off it...Again, in the same day, i have returned it on but without use it, only stay on in C128 mode. About after 10 minute the screen going to black, in both mode C64 or C128. This with a normal LUMA CHROMA output, but same with RBG.

I have checked power supply and replaced with another one, but all are ok and all respective point on the board have a correct power value.
I have fast checked it, replaced socketed rom but nothing and i have decided to store it. One Week ago i have decided to try rapair the board....................but.....................it's impossible..........Why?
I have replaced the RF modulator but nothing.
I have replaced and socketed all RAM and tested on C64 and all are ok but replaced with another 16 new tested chip, i have replaced all ROM and reburned new in a 27c128 EPROM but nothing (tested originale and burned in other 128 and all ok)
I have replaced and socketed Zilog Z80 CPU and tested original in a different computer (KAYPRO II) and it's working perfectly.
I have replaced and socketed MOS 8502 CPU with four new one tested in another C128 and all working perfect.
I have replaced and socketed video RAM and tested it in another computer and its working perfect.
I have replaced the MOS 8563 and MOS 8566 video chip but nothing, tested in other C128 and working.
I have replaced the MOS 8721 PAL and MOS 8722 MMU but nothing, tested in other C128 and working.
I have checked clock signals, present but i have replaced both oscillator nothing changed, replaced regulator capacitor on video section, nothing changed.

At this point i have decided to replace the 74LS06 and check on a reset circuit but all is ok, Then i have...........................socketed and replaced ALL, very very ALL logic chip on the board, but nothing same problem.
With the Diagnostic Cartdidge setted for C64 mode i have a correct tools border and background colors but garbage and colored character, then in C128 mode settings i have same black screen.
With the Jupiter Lander cartridge it's try to start bud i have a garbage screen, smilar at C64 with a broken kernel rom.

I have printed schematics on paper and i have starded a trace check................................................i have checked all trace, dataline, address line, power, logic, very very all trace, section per section.....all trace are ok!!
Then i have passed all data line on CPUs, MMU, PLA, SID, RAM and CIA chips with scope...........and:

When i insert Jupiter Lander cartridge i have pulse in all dataline of 8502 SPU and in all Address Line of it.
When i isert a Diagnostic Cartridge and system stays in C64 mode with previouse mentioned garbage screen i don't have any pulse in Address line, or little oscillation at 5 volt, and nothing in dataline.
Then i have tried to turn on it or reset system frequently and i have noted the addressline, in any case pulse, and in other case pulse one second and then stuck.

Now i have lost all ideas!
I have repaired other C128 o C128D or C64 and more and more different computer, but this is impossible!

Can anyone help me?

Thanks

Emanuel

 

[ClassicHasClass]

With everything you've replaced now I wonder if there's a bad trace on the logic board. That would be very hard to smoke out.

 

[demonlg]

Anyone have a image of scope waveform signal for the dataline and address line? I dont'know if is normal value or not, and i don't know if the signal have a 5 volt peak in all address/data lines, because in any case i have pulse but with max scale to 3,3 volt.

The trace on board are checked but i try to recheck again! In any case the address/data line change waveform or going out when i reset machine.

 

[Ruud]

With everything you've replaced now I wonder if there's a bad trace on the logic board. That would be very hard to smoke out.

IMHO a bad trace can not be counted for working for ten minutes and then quit. But, you never know.

@demonlg: you mentioned using another C128 for testing the ICs. Did you place the ICs to be tested back after the test or did leave them in the second C128? In the first case: exchange the parts and leave them. And turn on both C128s. The idea is that at one moment the problem should move to the other C128. That is: if it is a faulty IC. If the problem persists, you have to check all non-digital parts like resistors etc.

 

[dabone]

I had a bad 128 that had a intermittent failure. I don't remember the actual error, but I replaced most of the logic like you have.
I eventually gave up and used it to build a Neo128. It's not the cheapest option, but it is the best (IMHO). The latest version has much improved video quality.

Link to Neo128 on Tindie.

 

[demonlg]

Hi ruud! Do you remember me? Year year ago i have contacted you for my pet 8032 and you helped me to repair it! I'm Emanuele from Italy..

Yes i have moved all big chip from faulty 128 to the good one, and all big chip from good 128 to faulty one, but the problem is the same on faulty and the good one works.

i don't have moved logic chip because the good 128 don't have socket for it, only for big chip.

Today i have changed socket for the two video chip and two CIA chip, very poor, but nothing changed.

Yes trace check is a long work, i have tested with schematics in one month!!

But, i don'know why i have a garbage graphics screen with c64 game cart and garbage character screen with dead test cart!
30 minute ago i have removed the eprom on socket U32, the C64 kernal and basic rom, then turned on the 128 with dead test cart and i have garbage screen but its started with the audio test (the last test of cart before restart loop test)!!!
I jave burned new eprom chip for the four rom but nothing!

I have an old scope with any problem and i don't say if the waveform i see is correct or not, i have checked the 74ls245 from pin 1 to 10, and i have a scope line on 5v with little positive movement, but from pins 11 to 15 i have activity. Replaced 74ls245 but nothing!

Then i have noted, on SID chip on pin 23 a similar sinuisoidal upper to 5v waveform, and on pin 24 same but at 2 volt peak max. But, on pin 23 when i turn on 128 i have waveform for one second then stop but if i press reset the waveform stay up, this with dead test cart inserted configured in c64 mode and garbage screen, with the dead test in 128 mode i have black screen and waveform do not stay up!

Im very ungry!

Emanuele

 

[demonlg]

Question, i have a dubt on a pair of resistor pullup pack....but i have only new 10x10kohm, on 128 board are present 3 of this and i have replaced all of it, but other pullup (ex. RP8) is not a 10k resistor in pack, if i replace it with 10k pack is a big problem?

I have suspect for two resistor pack on the two ram bank, but tested with multimeter have a correct value , it is a two pair of 33ohm, i have this in a faulty and damaged C64C Board, and i suspect for RP2 and RP5, but i don't have another one!
then i have replaced all electrolitic capacitor.

And, please what is the value of C66 position on the board? This holes ar in bottom right of ram bank, aproximately under RAM chip U45, i have missed it and now i testing board without. Schematics and manual says a capacitor but in my two board i have a resistor in position C66, and in my working board the value of this is not readable!

The next step is to check Diodes and resistors, sure the 1n4148 on reset circuit is new and replaced, then i don't know value of CR8 zener.

I think to make a photo of my scope signals and post here!

Thanks

 

[demonlg]

Nothing, i have turned on the 128 without pullup resistor RP10 and RP5 but same garbage screen then replaced and same issue!

RP5 have the last point connected to pin 12 of PLA and PIN 46 of MMU (EXTROM), if i measure resistenze from +5v and this point i have a little resistenza value, not 3,3 k same of other RP5 pins, but i think is normal.

Now i desolder one point of all resistor on board and test all one to one, then all transistor C1815 and then i don't know!

 

[demonlg]

Nothing, i have used the troubleshoting guide for check c128 and with a new digital scope i have checked all clock signals, I have all 1mhz line, 2 mhz line, 4 mhz line 8 mhz line and 17 mhz line(not 14 because is PAL and not NTSC), then i have.turned on other working c128 and checked signals with scope and two input, one on broken board and one on working board. All signals are the same, only the address line and data lines are different because when turn on or reset the broken board the line pulse for one second then stop, in a working continue pulsing.

I have rechecked trace and found a lost pin 2 from U32 and U35 and fixed it but, not resolve.

Finally i have desoldered PLA and cpu 8502 from working C128 and socketed it, and i have exchanged it with broken board, but nothing, all my 5 PLA chip and five 8502 cpu are perfectly working, then I have piggy backed al logic chip from broken to working and board start up correctly, and replaced Q6..nothing!

At this point remain to check other Qx transistors and resistors...i think the filter 5v caps for all chip are ok, not shorted.

It's a very hard board repairing i have made!

 

[demonlg]

News, i have restared the trace check. When i have tested pin 47 of MMU 8722 i have noted the level stay to high with a C64 cartridge inserted or diagnostic setted in C64 mode, this is not normal and i have found any pin hole under logic chip in U60 U37 U8 U56 and U31 dsconnected from the trace, then another lost trace for SPI line on U1 6526 connetced ad pin 9 ov U8 fixed. But nothing changed, Now the system going in C64 mode but screen is with garbage again, only color of board and screen are correct for the diagnostic cartridge.

Now i have checked the DE and CE line on ROMS........and.......i have noted this:
When i insert a C64 Cartridge or Diagnostic, or if i press C= Keys at boot the DE of U32 (C64 KERNEL and BASIC) stay high and i have a 1 Mhz clock, if i don't insert cartridge and start system in 128 mode i have DE on U32 low but i have high DE on U33 U34 and u35 (128 BASIC ROMS and KERNEL) but the clock is not 1 Mhz, is at very low freq.
On the AddressLine of roms i dont' have activity, only one second when reset or power on then hangs and not pulse, i have checked with working C128 and its pulse continuosly.

Then, and this is a strange fact, i have removed all RAM CHIP and inverded from bank0 to bank1, now the screen is with garbage but with different pattern...and same, with JUPITER LANDER cartridge, now i have a different garbage screen with any particular corrupted graphic of the game...

This is a probably simpthoms of faulty RAM, but all chip are tested in a C64 BOARD and it's boot up with 38911 bytes free and working with game. The trace under RAM are OK,all checked and all socket replaced with new one, the CAS and RAS line are present.

I have checked all RP resistor package and replaced the variable capacitor on VICII position, then checked all clock frequency again and it's ok..

In Last point, i have checked pin 3 of CPU 8502 (IRQ) ........in not working board i don't have activity, in working one yes ! But trace of all IRQ Line are ok.

Any Idea?

Emanuel

 

[demonlg]

This is the screen i have with dagnostic cartridge in C64 mode, in C128 mode the screen stay black, if i turn off and on again the pattern of garbage change, is not the same pattern on new boot or reset, and if i reset C128 the position of letter A rotate for one second all character then stop.

Then other photo are with Jupiter Lander, Blue Print and Kickman cartridge respectively inserted, same result if i reset the pattern change and with kickman, in any case with more reset, i have a hang of sound, continuosly, seems to be try to start and then hang.

 

[demonlg]

Anyone have an idea? I have retested all RAM chip on two different C64 board and with ciclic diagnostic test, i have 24 good chips...In the next days i continue to retest all traces! The correct simpthoms are a CPU hanging about 2 second after startup.

 

[rittwage]

"Usually" the 128 has bad ROMs or ROM sockets.

But, on my last C128 repair I had bad VRAM on the 8563 (that caused bizarre problems you wouldn't think are related) and a bad 8502 CPU. The CPU also caused odd problems - it would work in 128 mode, but not C64 mode. Once I swapped it, the opposite was true until I replaced the VRAM.

It truly is a bizarre array of custom chips that is hard to diagnose.

 

[demonlg]

At this moment the socket of ROM are new, only the option ROM U36 (not used) is the original, and have a oxided pins, i don't know if this cause problem when is not used. The VDC RAM is possible, this RAM are original and not replaced, but i have socketed it, it is the only two chips not replaced. Al other one you mentioned are replaced and tested in another board and working, CPUs tested in working C128, C64 and CPM mode for Z80 and working.
But, the system boot up in the same issue with or without 8563 VDC chip installed.........mmm.... i have replaced also the 75ls244 in the video section upper to the video RAM, i can check again this area for eventually bad soldering or trace hole broken same of logic chip founded, but at the moment i don't have this ram chip for replace, i have an old expansion board removed from a 128D with enhanced video grafic, this board have two chip ram installed but with model code erased.

 

[demonlg]

I have terminated all wires check on the board, all connection now are ok, but the problem persist. Now, i have only one difference from good board and a fault board, at the U9 the good board have a 74F32 chip but on faulty board i have replaced with 74LS32. This chip is used for RAM CAS operativity bus, it's possible this chip cause a faulty on RAM on computer boot startup?

 

[dabone]

U9 NEEDS a F series, LS is too slow.

 

[demonlg]

Mmmm, interesting...i don't find the F series to buy, but its probably the issue....i will try to find F series of this logic.

 

[dabone]

You can try a HCT in there, not a HC..

 

[demonlg]

I have found the F series! Arrive to home in two three days then i try to replace. If it's working the problem is caused from me!

 

[demonlg]

Ok the 74F32 is arrived and i have installed it on U9 socket, now.....incredible, at first boot with diagnostic chart in 64 mode its boot and execute the test, the result is in image below, i have inserted only cartridge without control port, joyport and tape port test boards. The result is, bad both 6526, bad 6581, bad character rom and bad interrupt, but the sound output is working good, and at the moment i don't see broken character on screen, ram all ok in c64 mode.

But, i have turned off the board and then on again and i have a black screen....after five, six reset button pressing again i have the diagnostic starting correctly! I have turned off again and removed the diag cart and i have black screen in 128 mode, and, with any reset and commodore key pressed i have the c64 boot screen! The diag cart won't start if i set the 128 mode, every time black screen, same if i reset!

Finally i have inserted jupiter lander chart and tuned on again, black screen and after any reset the game start correctly and going to demo without any graphical problem!

At this point, i suspect a bad, or partially bad, 6526 cause the c64 start randomly and c128 won't start every time.

All rom chip are tested and with a eprom programmed i have the same simpthom.

Adding, all my c64 game chart before any resets start correctly!
Any idea?