A2000 (not B2000) resurrection

[Chain]

I have A2000 (very old one) in the loft and ive finally decided to repair it.
This machine has only two cinch connectors on the back (L+R audio), no composite output. Board is rev 4, no freakin damn buster, simplified cpu slot , half size videoslot etc etc.

and now what it does:
powers on, dark gray screen goes to light gray and to white. power led responds from dim to bright but thats all. It restarts after two seconds to light gray screen.

All custom chips and cpu are fine, so i assume problem must be in some logic ic. I can replace all of them, except four GALs :/ Is there any chance to obtain "content" of those?


so questions are two actually:
1) Where to get those GALs, atleast content ?
2) Where to look for some bad IC (if it behaves like described)?

thanks for any constructive help, no thread hijacking etc

[rkauer]

I don't know if the pre-4.x boards have it, but when the very same thing happened on mine 6.0 board the culprit was the TTL 74LS near the crystal and a problematic PSU (not enough voltage on the +5V rail: 4.7V).

Check also the tick signal.

[Chain]

There is a PC PSU used (old one blowed years ago), so no tick signal
5V line reads 5.12 Volts on all ICs (double checked).

found a PAL equations for that old lady, now how to get it to new PALs?
I can probably borrow PAL/GAL programmer from workmate, but howto translate it???

http://chain.3dgrafika.cz/amiga/A2k_PAL_equations/

is it a palasm or what?

[rkauer]

Looks like assembler, probably the PAL one. Just one way to find out...

BTW: if you'll follow the burn a new chip, your choice will be GAL, not PAL. PAL chips are obsolete, too hot to be reliable and sucks too much power. GAL ones are so much better!

[Chain]

yea, time to build my own galblast then :/

[Chain]

some measures around U29 (PALEN)

A17, pin 6, there is some action


PRI/W, pin 7, pulse on restart


C1 and C3, pins 13 & 14 looks same. 1/deltaT gives me something around 3.54MHz


/VPA, pin 15, there is something too


/RGAE, pin 18, looks almost opposite as /VPA


/RE, pin 19, whoa, looks like three floppy clicks (floppy isnt connected)



/OVR and XRDY seems to be logic zero all time, no trigger here

im still figuring out how that scope thingy works so be gentle

[Chain]

is there boot sequence described with timing in those old books? I mean hardware reference guide etc

[Chain]

im still wondering why there is a /RST generated every 3 seconds.
is it possible to be LM2901 bad, or some surrounded capacitor?
or it can be generated from somewhere else? from CPU?

EDIT:LM2901 seems to be ok and it generates /rst after three-key stroke

[rkauer]

...Or when a hardware fault is detected...

[Shadowfire]

I just looked at the A2000 schematic. It appears that the only things that can actually generate a reset are the processor, and the LM3901 @ U20 (which drives the processor RESET* line via buffer U12 and the HLT* line via U16). U20 is an open-collector output style opamp, [which means that it never drives the output high. It only drives the output low if (IN-) > (IN+)] . The circuit around U20 asserts reset if KCLK is held low for a period of time (the first opamp tristates, then C96 charges thru R92, and the reset is tripped if C96 rises above 2.5V, which is somewhat less than 1 RC, or around .18 seconds). KCLK is pulled up by R132 and is routed to the 8520 (presumably as in input to the 8520).

To divide and conquer, sample U12.13 or U20.14 on your scope. If there are any low pulses, then the problem is with your keyboard or the U20 circuitry. If not, the reset is being generated by the 68000, and the problem will likely require a good deal of luck to find, if you don't have access to a logic analyzer.

Since the machine does, however, boot up and flash through the typical startup, you can safely assume that the CPU data and address bus lines (as well as ROM and RAM arbitration) are fine. I would first suggest swapping the 1.2ROM with another one, if available.

The GALS in question, generate control lines for the 8520 and RGA bus, among other things. I would not expect them to be at fault if the machine is able to get to the white screen.

[Chain]

seems to be true, but its generated by CPU?
then how cpu detects hw fault?
There must be some error with rom mapping, as "hand with disks" picture never appears

[Shadowfire]

I doubt that theres a problem with rom mapping, seeing how much of the bootup actually gets started. If you really want to pursue this, you can test continuity for the address/data bus lines from the 68000 to the ROMS (should be <1 ohm), check the CS* and OE* lines on the ROM for activity with the scope. If those things check out, then the only problem you might have, if the ROM was the culprit, would be a bad ROM (note: I've had EPROMS pass the verify test in a DATAIO programmer, but not work in a circuit, due to damage in the EPROM output drivers that the DATAIO didn't catch.)

Let me reiterate, that I do think that you're barking up the wrong tree here.
http://home.arcor.de/kickstart/TKA/T.../trouble2.html may be of some help. Have you tried swapping floppy drives for a known good unit?

[Chain]

rom works in another machine, floppy drives arent connected.

trided this from troubleshoot already:

# No workbench prompt:
Screen goes from dark grey to light grey but does not display


OK - * 8520 U7 Odd CIA (U300 A2000)
OK - * 8364 U3 Paula (U200 A2000)
Not present in A2000-A, PAULA has power ok * EMI 1301 1ohm resistor (R309 Rev. 6A/7 boards; R200 A2000)
8367R0 agnus here, OK * 8370 - 8372/8375 U2 Fat Agnus (U101 A2000)
Not present in A2000-A * 74LS32 U37 (chip select) U301 A2000)
seems to be fine, checked connections * U2 socket

[rkauer]

Are you sure the PSU is fine? The guide you mentioned is what I use to make one of mine A2000 run again.

[Chain]

psu is PC-AT fitted in old cage. No Tick, but thats supplied from video part

[orange]

maybe you should test it with A2000 PSU that has TICK, just to be sure.

[Paul_s]

What condition is it in around battery area?
No dry join#ts?

Sorry for being obvious, especially with it being you Well, you never know, you may've been drunk when checking it over

[Dimlow]

Whats that 2v pulse at the beginning of of the trace on A17, dont look right to me.

[Chain]

I have only one probe now, so i cannot trigger if from somewhere else :/

that pulse occurs around dark gray>light gray screen change iirc

[Dimlow]

Is your scope dual trace? For this next test you will need it.

Check the OE enable CS lines on the various chips that are connected to a17 when you see that funny signal. Looks to me like one has failed, checking those lines will help you identify the chip. I don't know the hardware so I'm guessing. But this test will help.

Plenty of cheap probes on eBay!