Commodore 3022 Printer

[SiriusHardware]

Absolutely, and it isn't, it is staying low, so all of the devices whose RESET pins are connected to 7A/(10) are being held permanently in reset. So that's our first problem to solve.

Normally at switch-on, capacitor C6 is 'empty' and has to charge up through R20, which is connected to +5V. For the short time it takes the capacitor to charge up, IC 7A (5) is pulled low via the relatively low value resistor R5, which in turn means that IC 7A (10) should also be low for the same duration. Once the capacitor has charged sufficiently the state on 7A (5) changes from Low to High and the state of IC 7A (10) should also do the same. The effect of all of this, when it works, is to hold the system in reset for a short time after power is applied, after which the state of the RESET line changes from Low to High and the system starts running.

That's what should be happening and is not, so can you now look at the voltage on and / or state of IC 7A pin 5? Also, measure the voltages on both ends of R5.

[ScottishColin]

7A pin 5 does the same (well it nudges slightly to 376mV).

R5 is 314mV one end 376mV the other.

Colin.

[SiriusHardware]

OK, can you now have a look at the state of / voltage on IC 6A pin 5? And also tell me if the link J8 is fitted, or not fitted?

What sort of IC is IC 6A?

[SiriusHardware]

Have to head off for a bit, back later.

[ScottishColin]

IC 6A pin 5 stays low.

6A is a HD7406P

Link J8 is indeed fitted.

Colin.

[Mark1960]

I’m not familiar with this model of printer, but something to watch out for in general on dot matrix printers is that you don’t want the print needles permanently driven while the print head is trying to move. You might want to consider disconnecting the print head until you have a bit more confidence in the control board.

Also check that if the processor is held in reset, or not initialising the drivers for some reason, that the drive for motors and print needles is disabled.

[SiriusHardware]

Good thoughts, Mark, although bearing in mind that the uP and other ICs are intentionally held in reset for a short time after power on, I would hope that being held in reset should not cause a destructive state of affairs.

The 7406 is an open-collector inverter so a 'low' on its input would allow its output to rise high, provided nothing else is holding the line that the output is connected to, low.

With power on, identify which end of R5 has the lowest voltage on it, power off and measure the resistance from that end of R5 down to 0V. This is to see whether C6 has failed low resistance / short -circuit. I'm assuming you can't easily get to the pins of C6 which will be close mounted to the PCB to make the measurement directly on C6.

Busy at the moment, so I can only dip in every now and then. Anyone else, please pitch in with suggestions.

[Mark1960]

In the schematic posted earlier, looks like the case open switch will hold the reset active, so that might be first place to start if the 6504 is being held in reset.

[ScottishColin]

You're right about not getting to the C6 pins; two of the screws that hold the board down are stuck in place but I guess I'm going to have to sort that to replace any components.

Anyway - infinite resistance from low end of R5 to earth.

Colin.

Originally Posted by SiriusHardware

Good thoughts, Mark, although bearing in mind that the uP and other ICs are intentionally held in reset for a short time after power on, I would hope that being held in reset should not cause a destructive state of affairs.

The 7406 is an open-collector inverter so a 'low' on its input would allow its output to rise high, provided nothing else is holding the line that the output is connected to, low.

With power on, identify which end of R5 has the lowest voltage on it, power off and measure the resistance from that end of R5 down to 0V. This is to see whether C6 has failed low resistance / short -circuit. I'm assuming you can't easily get to the pins of C6 which will be close mounted to the PCB to make the measurement directly on C6.

Busy at the moment, so I can only dip in every now and then. Anyone else, please pitch in with suggestions.

[ScottishColin]

Yes - we thought that too but I simply cannot find a case open switch...

Colin.

Originally Posted by Mark1960

In the schematic posted earlier, looks like the case open switch will hold the reset active, so that might be first place to start if the 6504 is being held in reset.

[Mark1960]

Ok I think I caught up with where you are at now.

Is there 5v on one end of R20? This is not direct from 5v and seems to be via Q1, only when 24v supply is present.

[ScottishColin]

Yup - 5.14V.

Colin.

Originally Posted by Mark1960

Ok I think I caught up with where you are at now.

Is there 5v on one end of R20? This is not direct from 5v and seems to be via Q1, only when 24v supply is present.

[SiriusHardware]

Good spot by Mark, I hadn't followed that 5V line far enough to realise that it was switched, however the 5V does seem to be present.

We can disregard the mythical case-open switch because J8, which is connected across it, has been confirmed to be fitted.

For the measurement of C6 I would expect to see a brief flash of low resistance rising to a high resistance - for a 10uF capacitor that would only take a brief moment but you should see that, so I'm not sure we had the right measurement points there or were making proper contact.

One end of R5 should have direct continuity to 7A (5), find out which one that is and then measure from the -other- end of R5 down to ground. Meter on resistance, not on continuity - as you initially put the probes on those points you should see a brief flash of low resistance, rising to a higher resistance. If you don't see that we may need to do a proper check on C6, although an open-circuit C6 would not cause our reset-low problem.

Final suggestions for tonight, with power off, measure the resistances from 7A (5) to 0V, from 7A (10) to 0V, and from 6B (6) to 0V.

[Mark1960]

Its one of the strangest reset circuits I’ve ever seen. Either someone trying to be very clever, or just plain stupid, there is often not much difference.

Also connected to C6 via 330 ohm is the IRQ of the 6532, which I thought is an output but can also be an input, and also connected to the output of 4B-6.

R19 looks like its intended to make 7A operate as a schmitt trigger, but 1000P direct on the output is a heavy capacitive load, so wouldn’t be much of a surprise if 7A has been damaged.

[SiriusHardware]

Also connected to C6 via 330 ohm is the IRQ of the 6532, which I thought is an output but can also be an input, and also connected to the output of 4B-6.

Having read through the circuit description which annoyingly is primarily based on the circuit reference numbers of the Epson control board and not the Commodore one, it seems the IRQ pin of the 6532 is used in output mode as part of a 'watchdog' scheme, whereby if a timer in the 6532 times out due to not being refreshed within a certain time scale by the firmware, it resets the CPU to reboot the unit.

All of the 'exotic' ICs seem mercifully to be in sockets, so it should be easy to establish whether the 6532 IRQ pin is the source of the pulldown on the reset circuit.

Let's get the answers to questions already asked, first, though.

[ortek_service]

I've now also managed to work-out which of the (Both confusingly labelled EPSON, and not sure what the other '3110' label refers to) diagrams applies. As the manual doesn't make it clear what the 2022 and 2023 differences are.
I've also not managed to work out how a 3022 / 3023 differs, other than they are targeted at specific PET model series. And that the PET 2001(N/B) doesn't apparently have print routines built in, and you must have a printer customised or use with either the 2001N or 2001B?
But hopefully any PET 30xx / 40xx model can be used with a 30xx printer.

The only Commodore printers I recall having much experience of was a (Seikosha? based) MPS-801 (Commodore cut-down IEEE-488) 'Serial' one on the C64, where after fitting a new ribbon, the head wasn't returning to start. But this just used a very long coil spring to do this, and a bit of WD-40 freed-up the slightly-sticking mechanism.


Both 2022/2023 manual schematics seem a bit over-complex, having RAM (and timers + I/O) split between two 6532 'RIOT' IC's. And there's a crucial typo that says the ROM is in a 6332 (So easily confused with a 6532 / an RRIOT, with ROM in it), when it is actually (luckily) in a 'standard' 2332 dedicated ROM IC.
Hopefully these are all OK - But at least there are ways of replacing these if the originals are not available
(Assuming there's a copy of the ROM out there on the 'net, if it unfortunately ever fails / to verify it is still OK)

It seems to me that nIRQ (which is also a feed into their reset circuit) is driven by a (OR) gate of '4B' (Is there a table anyway to translate these PCB positions to real part numbers? - As seems to be lacking on the schematic / 'Cross-Reference' parts list / B.O.M) which is driven by an 'IFC' line from the IEEE-488 port connector (Pin 9), that is Pulled-high with 3k3 and pulled low with 6k8 'termination' resistor networks.
So the printer can also be reset by a signal on this port.

I also see that the switched +5V supply that also goes to their reset circuit, is controlled by Q1, that itself is controlled by the 'Power' rail. So locks-out the logic, if this is not present.

I've not yet found the paper feed switch on the schematic, that is also used for test-mode. But do need to get it out of reset first.

[TonyDuell]

I know this printer in the 2023 form, which is friction feed and otherwise identical.

A few points that I remember.

The little PCB under the mechanism is the motor speed control. The unit has one motor, when running it turns a complex double-thread leadscrew, the printhead will then move left to right and then right to left continuously, the latter being faster. There's a solenoid that engages a paper feed ratchet to advance the paper once per revolution of said leadscrew. It appears from the manual that the tractor feed model doesn't have that solenoid, but rather a separate stepper motor to feed the paper.

One reason for the complex reset circuit is to allow the IFC signal on the interface connector to do a hard reset of the printer microprocessor.

I've never seen a cover-open switch in this type of printer.

Once you've solved the reset problem, another 'gotcha' may be that some Commodore printers do ROM and RAM tests after a rest. If those fail it will not respond to the feed button or commands from the PET. Normally the LED will blink a fault code though

[SiriusHardware]

Thanks Tony - early days yet, but how are we doing for availability of printer ribbons for these? From the illustration, the ribbon looks like a bog standard typewriter ribbon rather than the ribbon 'cartridges' that many later DMPs used.

The gate IC 6B 4/5/6, which buffers the IFC signal on the way to the reset circuit, is on my 'to look at' list.

[ortek_service]

Originally Posted by TonyDuell

I know this printer in the 2023 form, which is friction feed and otherwise identical.
>>

That's very-helpful to know. I did see a link on the schematic for 'friction feed', but thought that there maybe a way of swapping between a detachable tractor feed and friction-feed or an optional sheet feeder, like on later Epson's.

So I assume the 2022 (& 3022?) is the tractor-feed version?

[ortek_service]

Originally Posted by SiriusHardware

Thanks Tony - early days yet, but how are we doing for availability of printer ribbons for these? From the illustration, the ribbon looks like a bog standard typewriter ribbon rather than the ribbon 'cartridges' that many later DMPs used.
>>

Well I suppose the crucial thing to know is what standard Epson models used the same mechanism - Pictures I found looked a bit like an FX80 or RX80?
(I know some of the later MPS-xxx ones, looked just like an LX80, I've got)

It may be possible to re-ink the ribbon, if Epson have discontinued them and compatible ones are NLA. I did used to spray WD-40 into the ribbon cartridge, after taking the lid off, to redistrbute the ink. And worked quite well, after a bit of drying-out excess, as I found some of the compatible ones dried-out very quickly after being opened. No doubt Epson had a better 'oil' amongst the multitude of chemicals they said there's used to lubricate the head properly