Page 1 of 2

Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 1:17 am
by Equinoxe
I have another eBay special (i.e. project Lazarus) that I'm hoping to resurrect - It's showing the usual memory initialisation pattern at startup (see pic) but goes no further. Voltage at output of 7905 is 4.85v (Input is 9.15v using a 2a modern PSU). The regulator heatsink is too hot to touch within 90 seconds. The board looks intact with no visible damage. I've tried shorting C21 with no change. No chips are hot.

I have a couple of decent multimeters on hand and a logic probe on the way. Any ideas on a systematic approach to try to start diagnosing the issue (preferably with minimal cutting of traces at least to start)?

Many thanks in advance!

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 2:39 am
by kenneth
The power voltage is too low. A chip swallows too much energy. Maybe a DRAM memory is dead? This one is hotter than all others.

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 3:24 am
by Equinoxe
None of the chips are getting hot - just the 6502 and 6522 get slightly warm after a few minutes - any suggestions on how I might be able to determine which chip may be at fault?

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 11:05 am
by Badger
It does sound like the reset circuit isnt working as it should.

This might be a long shot but check that the reset swicth hasnt become clogged and locked closed electrically. It should be normally open.

Once you get your logic probes you'll be able to test the clock signal is working.

Other than that maybe checkout http://blog.tynemouthsoftware.co.uk/201 ... uitry.html. It seems a very similar problem.

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 5:00 pm
by Equinoxe
I checked the reset switch and that's good. I also checked the voltage for the 6502 IRQ and NMI lines and read 4.85v and 4.75v respectively. I also checked the resistance across the 5v rail and read 160 Ohms. I also did a memory chip piggy back test with no change on the display. I have spare ICs on order but they wont be here for a few weeks so if anyone has any other suggestions in the meantime they would be greatly appreciated :)

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 5:42 pm
by Chema
4.85v is too low. A component may be faulty. Most of the time it is a memory chip, but if no chip gets hot, it is difficult to tell.

It could be almost anything. You can start from removing the socketed chips (ULA and ROM) and check if you get something near 5 volts.

You have a good screen image, so the ULA seems to be ok, although it could also gave a partial failure an might not be generating the signals correctly. As there is no activity at all, the next suspect (apart from the memory) is the CPU, but it is not socketed and I don't know how to test it.

Anyway, did you have a look at Mike's repair guide? http://oric.signal11.org.uk/files/pub/r ... st1-00.pdf

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 6:48 pm
by mikeb
+1 on the repair guide link ... ;)

The RESET switch (NMI) is edge-triggered, isn't it? So if it was stuck, open or closed, it wouldn't stop oric working. It just wouldn't warm-reset.

The reset circuit (power up, cold boot) is a different thing, and I've found before that an Oric that is powered up won't start beyond this bars-and-black, if the reset circuit is tired. Other times it starts up but there's a spurious character "typed" as the VIA/6522 didn't seem to initialise right ...

To manually force a COLD reset, once you're powered up, dab a wire briefly from pin 40 of the 6502 CPU to ground: You can find this at pin 1 of the 6502, so basically opposing pins at the notched end of the chip.

If this helps, there's only 2 components in the reset circuit, and one is C21 1uf/10v capacitor, which may be getting old. Try replacing it.

As to DRAM piggybacking -- if a DRAM chip has failed, and is holding an address/dataline down/up, then piggybacking won't improve anything. That only works if the suspect DRAM chip has effectively gone "open circuit" and is no longer contributing. It also assumes only one failed ...

Your system voltage does look low, so something may be overloading the regulator. In the repair notes PDF there are hints for where to cut tracks to isolate the power rails to try and remove the faulty area (block by block), you may need that.

Re: Oric Atmos Won't Initialise

Posted: Sat Mar 16, 2019 6:54 pm
by peacer
Same thing had happened to my Oric in the past. The reason was a failed ram chip. I replaced it and it worked fine.. Maybe you could consider this..

Re: Oric Atmos Won't Initialise

Posted: Sun Mar 17, 2019 6:46 pm
by mikeb
peacer wrote: Sat Mar 16, 2019 6:54 pm Same thing had happened to my Oric in the past. The reason was a failed ram chip. I replaced it and it worked fine.. Maybe you could consider this..
It's likely it is -- but finding which one is hard. If you get lucky, the failed one heats up. We've already established that the memory chips aren't unduly warm/different, the regulator is ready to cook eggs, and the 6502/6522 are "warm". So no easy indicator.

Absent of getting in there on the data/address lines around each DRAM chip and looking for an obviously stuck high, stuck low, intermediate-silly-voltage, it means either replacing all 8 for (excessive good) measure, or tossing an 8 sided coin and changing them one at a time while hoping.

If you do remove any chips, please consider fitting a socket in its place, it will save trouble later if you (or the next repairer) ever have to fault find again.

Re: Oric Atmos Won't Initialise

Posted: Mon Mar 18, 2019 5:29 am
by Equinoxe
Thanks so much to everyone for the help so far - I was hoping there were some easy tests that would avoid putting this machine 'under the knife' by way of cutting traces but if that's what's needed to give another 35 years of life I'm willing to do so.

Before I go down that route, on the topic of RAM testing is there a way to test the RAM 'in circuit' with the logic probe (once it arrives) or any other IC's for that matter? I've ordered a complete set or sockets for every IC on the board definitely with future service requirements in mind :)

Re: Oric Atmos Won't Initialise

Posted: Mon Mar 18, 2019 6:50 pm
by mikeb
Probably not entirely with a logic probe (which will only tell you low, high, high-Z (floating) or pulsing). Again, unless you get lucky :)

Oscilloscope might be handy if you have access to one?

With the DRAM, you have 8 "identical" chips. You can compare the states of the corresponding legs on each chip and see if there is something obviously different about one chip.

The chips will be being accessed by the ULA even if the 6502 is not working (this must be true, you are getting a picture). So you should see reasonably equal behaviour of the data line (Din and Dout are tied together), the address lines, RAS, CAS ... (Google 4164 DRAM pinout if it's not obvious from the Oric Circuit).

Something I have mentioned before (and is in the repair notes linked) is that if the power rails are wrong, all sorts of odd symptoms can appear which you will never fault find, as they are not real faults. 4.85v is on the low side, legal range is 5.00v +/-0.25v, so finding why it's low is useful.

Even if that's "Doh!, it's not low, my battery powered multimeter needed a new battery!" (it happens more often than you think).

Re: Oric Atmos Won't Initialise

Posted: Tue Mar 19, 2019 5:48 am
by Equinoxe
I have access to oscilloscopes at work (or rather, the people I work with have access to them) so I could probably ask one of our techs to have a look at the Oric (which I'm sure would generate some curiosity since they were never sold in North America) I was first trying to do it myself with my limited knowledge and also trying not to leverage too many favours until they were really needed. The techs would happily let me tinker on my own with their gear but I don't have the first clue how to set it up and read the results. Before I go to the next step I'll double check the Oric voltages with my second multimeter just to be sure the batteries are good ;)

Re: Oric Atmos Won't Initialise

Posted: Tue Mar 19, 2019 8:39 am
by iss
+1 that it's bad RAM chip (160Ohms between GND and +5V is low). Else everything helpful is already said.

From the picture we can conclude that 12MHz generator (XT1+IC21) and ULA are working.
When the logic probe arrives you can check the presence of F2 signal and event what happens on the address/data buses immediately after RESET, until then - just a small advice - don't cut tracks, it's definitely better to sacrifice any chip than the PCB!

(additionally you can try to finish solving the Rubik's cube but I doubt that this will fix the Oric ;).)

Re: Oric Atmos Won't Initialise

Posted: Tue Mar 19, 2019 6:36 pm
by mikeb
Just to be clear ... the tracks I was talking about cutting are large, easy to cut, easy to repair, GND voltage rails, and purely for the purpose of isolating the fault, so that chip replacement can be targetted to the (hopefully single) faulty chip.

To remove even a single 14 pin DIP chip from a board of this age risks damaging the through-plated holes, or lifting surface traces off. This is much more work to fix than a simple break in a voltage rail, which you deliberately introduced, and can put back.

Cutting tracks was precisely to PREVENT "sacrificing" a PCB, not to cause it.

OTOH: Cutting off all the legs of a chip (destroying it) in the hope that you might find the faulty one is fine if you have a large pile of spare 6502, 6522, ULAs etc.

Re: Oric Atmos Won't Initialise

Posted: Wed Mar 20, 2019 4:45 am
by Equinoxe
iss wrote: Tue Mar 19, 2019 8:39 am +1 that it's bad RAM chip (160Ohms between GND and +5V is low). Else everything helpful is already said.

From the picture we can conclude that 12MHz generator (XT1+IC21) and ULA are working.
When the logic probe arrives you can check the presence of F2 signal and event what happens on the address/data buses immediately after RESET, until then - just a small advice - don't cut tracks, it's definitely better to sacrifice any chip than the PCB!

(additionally you can try to finish solving the Rubik's cube but I doubt that this will fix the Oric ;).)
20190319_213307.jpg
Done! :) My kids don't believe I used to be able to do these in under 2 minutes - sign of a misspent youth I guess :)

I had a chat with one of our techs today and he said one of them would take a look at it so I'm packing it in an anti-stat bag as we speak (and pondering ideas for a gift card). He even said I could even take home one of their older desoldering stations if I'm feeling adventurous (time will tell on that). This Oric is in great condition otherwise for its age (it even still has the protective plastic on the shiny top strip) so it would be nice to preserve as much of its originality as possible - I've even wondered if it's worth creating a fake serial number sticker for nostalgia's sake :)

In any case, as soon as I know more I'll report back on the results - Thanks again!