Commodore 3022 Printer

[ScottishColin]

Again, 4.97V.

Colin.

Originally Posted by SiriusHardware

Try that again but with the 7B RIOT now fitted (all pins connected) temporarily in the 5B position. Leave the IC originally from 5B in the 7B socket and look again at what happens to the system RESET line.

[ScottishColin]

As a test, I swapped them back into their original positions with all pins connected and RESET is still 4.96V. Now I'm confused.

Colin.

[ScottishColin]

OK - now 1 power-on in 10 I get this. I'm going to see if I can safely poke devices/cabling to see what I can find.

https://drive.google.com/file/d/1-vY...ew?usp=sharing

Colin.

[ortek_service]

I did notice that it should normally be possible for the head solenoids to be permanently driven, by a faulty '5B' 6532 RIOT (or this not being controlled correctly by the 6504 CPU etc) as the '2A' NE555 should provide drive-pulses to the '3B' / '4B' 74LS09 (Open-collector) AND gates that drive the array of darlington-pairs head solenoid driver transistors.

So would presumably just get a loud 'buzzing' from the head, if pins were being continuously driven in ad out by these pulses. That would still be effectively-more than the 'max 5 sequential reverse-image characters' after only a second, but not as bad as having them continuously-on. And hopefully less-likely to damage the head.

I would also expect the circuitry to be design to have all of these drives off by default at reset, and not rely on configuring the outputs of '5B' 6532 RIOT to be different to what they would be at reset.

Originally Posted by ScottishColin

Thanks Tony - that's very kind. I'm hoping I haven't done a bad thing to the print head; I guess we'll see though.
Colin.

Originally Posted by SiriusHardware

Tony, would it be possible for you to measure the resistances of the solenoid coils on your TX80 head? If Colin then makes the same measurement on his that may give us some idea of the state of his solenoid coils.

Originally Posted by TonyDuell

A couple more comments...

The ribbon had little studs/eyelets near the end to trip the autoreverse mechanism (some typewriters do this too). The epson manual describes the reversing mechanism and how to take it apart.

You want to get those PCB screws out, you are going to want to get to the solder side. Incidentally, can you get the mechanism out? If so, then I'd remove the complete ribbon drive unit at this point (it's only 3 screws, one of which is the clamp for the printhead flexiprint) to make it easier to remove the printhead.

And you're not going to believe this, but I've checked my spares stock and it seems I have a TX80 printhead. It may well be the one with the broken pin that came from my 2023, but the solenoid unit would be good. I might consider exchanging it for your dead one if need be.

[/QUOTE]

[SiriusHardware]

I did notice that it should normally be IM?possible for the head solenoids to be permanently driven, by a faulty '5B' 6532 RIOT (or this not being controlled correctly by the 6504 CPU etc) as the '2A' NE555 should provide drive-pulses to the '3B' / '4B' 74LS09 (Open-collector) AND gates that drive the array of darlington-pairs head solenoid driver transistors.

I also thought that no matter what else happened the 555 timer should limit the length of a head 'strike' to a hardware limited amount of time, but maybe the 555 is being continually triggered by whatever the fault condition is.

I guess we should wait and see if Colin can find a physical reason (bad connection etc) for the unit only doing a head seek every one startup in a while.

[ortek_service]

BTW. I think I eventually managed to work out what type numbers all the gates actually are on the schematic for this 'Commodore' Hitachi-built board.
- Which isn't too easy, when reference positions aren't very-legible on the schematic for this (and they didn't include the type numbers at all on that for the smaller gate symbols) !

Plus the manual's 'Cross-Reference' Parts list isn't much help when it is only for the 'Epson-built' board, that uses quite a few different IC's (and different, more-conventional, sequential numeric-only IC designators). Not sure why Commodore didn't include one for their (Hitachi-built) version of the board.
So had to cross-reference Colin's board photo / the manual's 'Hitachi' layout diagram, that shows type number and reference position designator for all of these.

The 'Epson-built' board's schematic is much clearer, being split over 2 sheets and with additional annotation notes / labelling od the +30V unregulated rail etc. And much of the circuitry is functionally similar.
Although the Epson one only seems to use 12 IC's rather than 13 in the Commodore / 'Hitachi' one - Which also has an extra (IFC input) PA7 to nIRQ open-collect gate, that directly-pulls the nIRQ output of '7B' Low, whenever IFC is pulled-Low. Whereas it seems the 'Epson' version relies on the '7B' 6532 RIOT IC pulling nIRQ low in response to PA7 going low (Presumably after reset and the 6504 CPU configuring it to do this). So these two boards may behave slightly differently, if IFC is low when printer is switched-on.

[ortek_service]

Originally Posted by SiriusHardware

Regarding the RIOT ICs, I had a quick look around to see if anyone has made a tester project for these. Here is one commercially available product, but while I respect the development time which would have gone into it, I feel it is overpriced given the limited number of devices it tests.

https://www.pinballsolutions.eu/prod...a-riot-tester/

It would theoretically be possible to make a 6532 tester with an Arduino Mega and a datasheet for the 6532. It would not be an absolutely trivial task, this ongoing forum thread gives some idea of the process.

http://forum.6502.org/viewtopic.php?t=7678

Yes, some of these Arcade IC testers can be a bit pricey, and many are now also NLA. So an open-source IC-Tester would be much better.

It might be cheaper to use an Atari 2600 VCS, if still cheap, to do a quick test of these / provide a working donor- as it seems these have one under the same Atari part number that's actually on the ones in that printer: https://www.arcadecomponents.com/cat...45/2173752.htm
Or see if a cheap '6532' spare is available - Not yet found a UK stockist, but it seems they can be found for < 10 euros / maybe $5, with some shopping around, (but shipping to the UK could rather bump that up).

Plus it does seems that the 6532 is actually used in some of the Commodore PET etc Disk Drives: https://en.wikipedia.org/wiki/MOS_Technology_6532
So maybe able to swap them with ones in those, if the model Colin has does use these (and hopefully socketed): https://en.wikipedia.org/wiki/Commodore_8050

Hopefully, the original 6532's are actually OK and there's just a bad contact (Although there's none of those White IC sockets, it's probably worth cleaning the pins and reseating all socketed IC's, as there seems to be some intermittent issues).

With the ROM now verified OK, then as long as the RIOT's do work OK and the 6504 CPU is OK, then it should hopefully be doing a bit more once reset is working OK.

[Mark1960]

Cricklewood have the 6532 listed. Not as cheap as you’ll find from china, but probably more reliable, less likely to be something relabelled.

[SiriusHardware]

At the moment there is some question as to whether the 6532s are faulty, because reseating them and the ROM seems to have made the RESET line issue disappear.

[ScottishColin]

Started off with some Deoxit, but no help there.

Then, I thought I had a mechanical problem involving years of grease/muck so it's all had a good clean, but no change from intermittent spinning up.

I have removed the motor from the chassis so there is no resistance for the motor, but I'm still getting the same problem. I think I have now narrowed it down to a wiring fault somewhere in the blue connector/motherboard socket, so I'll focus there for a while. A wiggle of the connector sometimes allows the motor to spin up.

I can't get the motor to spin up after power on if it doesn't initially spin; it's like it gets a power on signal once and never again. There is no resistance on the motor shaft if I power on and it doesn't spin.

Colin.

[ScottishColin]

The motor is an SM-E005; Can anyone find a spec sheet so I can what voltages I should be looking for please? I've looked and failed.

Thanks.

Colin.

[ScottishColin]

OK - if I look at the EPSON .PDF linked to in post 13, section 3.3 it says the motor should spin if I short circuit two points on the smaller PCB which doesn;t work.

I've checked the voltage to the PCB and it's 26V; I believe it should be 24V - is that in the margin of error?

There are small wires on the motor where I can check the voltage and it's getting 26V but not spinning. Does this mean I have a problem with the motor or should I still be hunting wiring?

Colin.

[SiriusHardware]

Looking at the motor controller circuit on PDF page 20 of the manual you linked to in the first post it can be seen that power for the +ve side of the motor comes from the +24V rail. I think 26V is OK. If you have 26V directly on the wires going to the motor and the motor is not turning, then there may be a problem with the motor.

The 24/26V power is permanently connected to the +ve side of the motor and its -ve terminal is then connected to GND via the darlington pair of transistors Q3 / Q4. When it wants the motor to run the control circuit applies a 'low' signal to the MD (motor drive) input to turn off Q1, which allows the voltage on Q2 collector / Q3 base to rise, turning on Q3 / Q4 and therefore turning on the motor.

Darlington pair Q5 / Q6 look as though they are connected in such a way as to short the motor terminals together when they turn on, and that is exactly what they are for, they are the 'Brake'. If you just remove power from a running motor it can run on for some time afterwards but if you short its terminals togther it will stop much more quickly, so Q5 and Q6 turn on whenever the motor drive signal is deactivated, in order to stop the motor as quickly as possible.

There is also another motor-like item in there which is driven by the motor / mechanism and that is the 'generator' which looks like it is used for two purposes - some of its output is fed back to Q2 where it is used as a speed correcting signal - if the motor runs too fast, Q2 turns on more and reduces the drive to Q3 base, thereby reducing the speed.

The other thing the generator is used for is to provide a 'tacho' like signal via Q7 / Q8 to the 'Timing Signal' output from the motor control board and this is to enable the control PCB to either measure the speed, or at least to know that the motor is running.

[Mark1960]

Are you able to measure the resistance of the motor? Then slowly turn the motor while measuring the resistance and see how it changes. I think this might check the contacts of the motor brushes and if any segments of the armature of the motor are open circuit.

[ScottishColin]

Brushes. I don't know why I didn't think of them. I've mended enough washing machines and tumble dryers in my time.

They've gone (or at least one has crumbled away enough.

So - does anyone have any good guesses how/where I can source some more - or am I looking at a new motor (which looks tricky to source in the UK)?

Colin.

[SiriusHardware]

Do you have an Android phone? If you can get the most intact brush out by itself and take a photo of it on a totally plain background, try using Google Lens to find online images of similar items, one of them may lead you to a source for them. I've only just started using Lens but it has found me items as diverse as wildflowers spotted while out on a walk and shower rail runners for a rail we didn't buy and had no idea who made it.

Not being an Apple man, I don't know whether iOS has a similar object-ident feature?

[ScottishColin]

Not so far.

They measure as 7.5mm x 2.5mm - length unknown. I guess thats 03." c 0.1" in old money.

Colin.

[SiriusHardware]

Could you post some good shots of a single brush here? It all helps to widen the search. I wonder if anyone does replacement general purpose brush contacts, just the little carbon block itself? Not sure how you would attach it to the metal arm bit, though.

[ScottishColin]

It is laid out flat so would be longer new.

It has a wire/braid embedded in it.

Colin.

[ScottishColin]

This web site says it has a motor if I can't sort out the brushes. I'd like to fix it really if possible.

https://www.radwell.co.uk/en-GB/Buy/EPSON/EPSON/SM-E005

Colin.