A600 will not do a warm reset - Ctrl-Amiga - Amiga

[kipper2k]

Hi All,

A strange problem on an A600 i am working on. The computer will not do a warm reset.

I have tried 3 different keyboards (that work fine on other A600). Heres my test, i boot up relokick floppy and then press all keys on the keyboard when it gets to a cli window to see if they repeat on the screen. All printable characters print on screen fine. I press shift, all characters shift, caps lock light toggles fine. enter works great.

Then, i press Ctrl-Amiga-Amiga, i see the caps lock LED do a dim blink and then all keyboard input goes screwy, keys dont respond, keys do weird characters on screen, screen clears, its like someone else is pressing the keyboard. The computer will eventually freeze and then a hard reboot is the only option.

I checked out the schematics and as far as i can see the input from the keyboard goes to the keyboard MPU (U13) and the 8520 CIA (U7).

Games seem to work fine on it and i haven't come across any problems actually playing a game. The PCMCIA slot, floppy, and CF card appear to work fine. Display looks great on composite.

Has anyone had this kind of problem b4. Anyone any ideas on possible causes ?



edit, i also tried 3 different rom chips, 1.3, 2.05, and 3.1 which are also tested good

[rkauer]

check for dry joints on the keyboard connector on the motherboard.

[Photon]

Yeah, if you tried different keyboards and their contact strips all look fine, then it's on the mobo.

[kipper2k]

Originally Posted by Photon

Yeah, if you tried different keyboards and their contact strips all look fine, then it's on the mobo.

thanks for the reply Photon, you edited your post b4 i got a chance to read the neat little trick with the duct tape to try to repair the ribbon cable fully. I'll have to remember that.

I have played around a little more today and i did put a piece of sellotape on the back of the contact strip so that when its inserted into the slot and the top pressed down it actually presses tighter on the contacts. I also resoldered every pin of the connector on the mobo to see if that helped.

Results... twice out of approx 100 attempts it did work, still keyboard freezing so i think its time to swap out u13 or u7... not sure which one to try first. mabbe the keyboard mpu first and then if that don't work onto the odd CIA

[Photon]

Yeah, was looking for the keyboard cable thread just now cos I decided to paste it there. So it'll be in there

[Edit: here!]

[rkauer]

also remember that the <CTRL><LeftA><RightA> have a dedicated line to reset the computer! It is a trace on the keyboard who runs to the keyboard connector.

Verify the line just hooking it to ground, who will reset the Amiga. The A600 schematics will come in handy for this!

[kipper2k]

Originally Posted by rkauer

also remember that the <CTRL><LeftA><RightA> have a dedicated line to reset the computer! It is a trace on the keyboard who runs to the keyboard connector.

Verify the line just hooking it to ground, who will reset the Amiga. The A600 schematics will come in handy for this!

Ok, so this is where i am at.

I have removed and socketed Gayle, Paula, the Keyboard MMu, U7 CIA and replaced all chips one at a time, and still no go on the reset. I also removed the plastic keyboard connector off the mobo too to check for bad traces but it looks good under there too.

I can reset the Amiga by grounding either

pin 43 on Keyboard IC, or
pin 20 on the cpu 68000
pin 16 on Agnus.

I have replaced U36 and U14 on the decoupling circuit and the freaking thing still refuses to to a warm reset. All traces for the reset line look good. I promised myself i wouldnt remove any chips off this board as its my only NTSC A600 and i dont want to trash it but this reset thing has really got me puzzled. I have retried 3 keyboards that are proven on other A600s and they definately work

I can cheat by making a little normally open button but i hate having to compromise (ie give up!!)

The only thing i can think of now is that mabbe the resistance of the combined keypresses don't activate the reset signal. Something does happen though as the keyboard will semi freeze for about 10 - 15 secs

Any ideas anyone

[Zetr0]

time to change the keyboard MPU Kipps =(

[rkauer]

Cannot be the reset circuit acting up? IIRC, there is a 555 in the main board for this...

[Zetr0]

@rkauer

it could be my friend, what concerns me is that Kipps tested the circuit and it worked (reset).

the only otherthing I can think of (if not the Keyboard MPU) is a dead resistor somewhere in the circuit not pulling down to reset properly.

What makes me think its the Keyboard MPU is the intermitten action of the keyboard afterward, very tricky this one....

One way to test the MPU would be to use one of those plug-on adpters for PS2 keyboards and then test it by attempting a reset by the keyboard.

[Shadowfire]

The fact that he's already tried 3 different keyboards that work in other motherboards, rules out any problems with the keyboard itself pretty definitively.

Since he apparently has no problems with power-on reset (i.e. the machine comes up every time), we can assume that the U14 (555) timer circuitry and Gayle are operating correctly.

This leaves R624 and U13 (6570 keyboard controller) as the remaining possible culprits (U13 far more likely), unless you have the U36/R623/Q622 devices installed on the machine, of which the most likely to fail is U36 due to potential ESD from the keyboard.

Pin 43 on the keyboard IC is the system reset line, and is almost certainly an input on the keyboard IC. The keyboard controller resets the machine by asserting pin 34 low, which triggers the 555 to pulse the KB_RESET line, which feeds into Gayle (which should assert the system _RST line). Based on the description of the symptoms, I'm willing to be that you don't have R623/Q622 installed on your machine, and the problem is that U13 is toast.

Edit: Wait... you already replaced U13. Please let us know if your machine has R624 or R623 installed.

If R624 is installed, it should be 0 ohms, so check for continuity.

If R623 is installed, check that when you press Ctrl-Amiga-Amiga, that as long as the keys are pressed:
U36.1,U36.2, and U36.13 are LOW (<.6V) and U36.12 is HIGH (>.6V) (validates U36 is working)
If the above passes, check Q622. Q622.3 should under normal operating circumstances be pulled to 5V by R611 (which is 1Meg). Which giving the machine the 3-fingered salute, Q622.1 (base) is pulled high by U36.12, which activates Q622, which discharges C611 and triggers the 555. When you let go of CTRL-A-A, the usual power on reset timings generated by the 555 timer and R612/C612 take over, and these are known good (unless you are having cold boot problems that you aren't telling us about).

[kipper2k]

Hi Zetr0.

I think U14 was the 555 timer, that was one of the components that i did actually swap out. there was the nasty fishy smell when i replaced that and also U36. I did replace the capacitors when i got this board as that area of the board was a little green, not too severe but definately needed. i've looked real close at the board for dead traces and touched up pretty well most joints in case they were cracked/dry.

There was also an issue with this board freezing when loading games (That stopped after i replaced Gayle). mabbe the thing is too tired to play any more. Once i get my spare NTSC Agnus i may just convert one of my other Pal machines and if that works i'll take another board and put the NTSC Agnus off this one and put in on there.

I have quite a few of these boards that i picked up real cheap so its not that bad of a loss (except for this one i'm trying to fix now, it cost $130).

I'm not quite ready to give up yet though

[Shadowfire]

Heh, I should read the whole thread before commenting.

When you press CTRL-Am-Am, check for the following things:
U36.12 rises to >=.6V (note it will be easier to check this at R623).
Q622 actually drives the 555 trigger input low (Q622.3, maybe easier to check at R611).
As you've already replace U36 and U13, it would appear you have a dead Q622 or trace damage on the board.

If I had to guess (and this is pure conjecture) the Keyboard IC is detecting CTRL-A-A, and asserting PA7 low, then running in an infinite do-nothing loop (assuming that it should be hard reset by the system _RST line).
It then probably has an internal watchdog timer reset, and comes back online.

[kipper2k]

cold boots are fine, i have also tried different KS roms, both burnt and official roms. I'll check all the other components you recommend at lunchtime.

Thanks for all the help guys

[kipper2k]

heres a pic of the A600 NTSC board...

hires here

http://www.kipper2k.com/amiga/a600ntsc.jpg


(there are a few strands of cotton Qtips on the board still ... btw, i also swapped out U15

[kipper2k]

it looks like R624 is not installed on this mobo. R623 is installed on the bottom of the mobo. so i'll check the voltages and resistances in a bit

[kipper2k]

Originally Posted by Shadowfire

Heh, I should read the whole thread before commenting.

When you press CTRL-Am-Am, check for the following things:
U36.12 rises to >=.6V (note it will be easier to check this at R623).
Q622 actually drives the 555 trigger input low (Q622.3, maybe easier to check at R611).
As you've already replace U36 and U13, it would appear you have a dead Q622 or trace damage on the board.

If I had to guess (and this is pure conjecture) the Keyboard IC is detecting CTRL-A-A, and asserting PA7 low, then running in an infinite do-nothing loop (assuming that it should be hard reset by the system _RST line).
It then probably has an internal watchdog timer reset, and comes back online.

@Shadowfire.


I changed out Q622 and no go. I am looking at that part of the schematic and i see that the KB_RESET is triggered by Q511.3. Question is, when that transistor is on, does it dump to ground causing a low to appear on the KB_RESET line
Also i notice that there are 3 inputs from PB2, PB3 and PB6 that feed into the triple input NAND 74F27 (hmm, just noticed that when i do a search for the 74F27 it shows as a triple NOR, my schematic shows a AND symbol with the circles on the input that indicates its a NAND?). are those the raw Ctrl, L Amiga and R Amiga inputs from the membrane?
I still have to do the meter checks later tonight (Have to get the missus to hold down the keys for me while i get a read

Thanks for your help



edit... Does anyone have the Membrane schematics, i don't have that part of the schematics ?

[kipper2k]

Solved!!


See if you can find the bad trace.... Answer is below




Ok, look to the right of the 10uf electrolytic capacitor, there are 4 empty pads, i removed the capacitor from the upper and lower pad closest to the capacitor. look at the trace coming into the top of the lower pad... the trace coming from Q511 (currently removed) to that pad is broke. Therefore the transistor would never go low and the KB_RESET would not generate a warm reset!!



YAYYYYYYYYYYYYYYYYYYYYY!!!!!!


Thanks to all for help

[prowler]

It doesn't look broken. Well spotted, kipper!

[kipper2k]

Heres the fix