Texas Instruments SR60A Repair(s)

 

Quite a while ago I finally saw one of these machines on ebay. The listing said it wasn't working, but it was cheap, very cheap in fact, so I assumed it was indeed not working.

It arrived.

 

The case top was not attached and the display PCB was out of it's frame but that was easily fixed.

 

Power on and:

 

 Do not power the printer without the control signal ribbon attached

It was fully functional, apart from the card reader, which I couldn't test as I had no cards. Excellent.

Roll on a few months and I wanted to exhibit the machine at Retrofest. I pulled it out and turned it on and up popped the 'PROMPTING DESIRED?' message. Excellent. All working.

A day or two later I turned it on again, just to look at the LED dot matrix display. 

'PROMPTING DESIRED?' it said. For two seconds. Then a blank screen. Oh dear.

 Do not power the printer without the control signal ribbon attached

 Fortunately the service manual is available and I had a copy. First thing: check supplies. The PSU is a linear-switching-linear affair with a transformer wired to a voltage select switch (with the most incomprehensible labelling I've ever seen) the output of which goes to a bridge rectifier and a big capacitor. This supplies between 25 and 43V to the input of this PCB:

 

The DC input is at the bottom, there's a switching supply or supplies, it looks like, followed by several linear regulators. The PSU supplies only negative voltages. There's a -20V, -19V -15.8V, -10V and -5V. 

 Do not power the printer without the control signal ribbon attached

 Unfortunately (very unfortunately), the service manual does not include circuit diagrams. That would have made things very much easier.

I disconnected all of the supply connectors and powered up the PSU.  I checked the rails and the -15.8V supply was not present.  After some PCB examination it appeared that that rail came from a TIP30 which was driven from a smaller transistor of a type I couldn't identify. I bought a few TIP30s and replaced that, but the ril didn't re-apper. I started tracing the circuit, but it's complex. I bought a new component tester and tested all of the transistors on the PCB and found tht the smaller transistor that drove the TIP30 had failed. With the tester I also identified what the transistors were:

 

KA2031:  NPN

T1598:     PNP

TIC106Y:   SCR

TIP33: NPN

TIP35: NPN 

I also identified the pinouts, I'll put them up somewhere. With that information I used a 2N3906 that I had to replace the blown transistor. When power was applied, the rail returned. I connected the logic and display PCB and powered up. No joy. When disconnected the rail had gone again. I replaced the transistor and the rail was back. I then checked the resistanc eof the rai on the PCBs and found that the main logic PCB had some old 'bullet shaped' tantalum capacitors. One was short circuit so I replaced it with an electrolytic. I actualy replaced all of the tantalums that I could find. Power was re-applied and prompting was desired. Success.

Do not power the printer without the control signal ribbon attached 

I reconnected everything and powered up. No display. Unfortunatey the wire to the logic PCB carrying the -15,8V supply had fallen off the PCB due to age and a lot of flexing. I soldered it back on and nothing. It turns out that the other end of the wire had fallen off as well. I put that back on and, well, nothing.

At this point I decided to draw the circuit out, but got fed up doing that and decided to test all of the transistors again. The -15.8V rail had gone again and that path worked last time. No failed device appeared. I took the PCB out and ran it off a bench PSU as I could probek the PCB easier that way. When it's in the machine it's not possible to access the conponent side of the PCB. After a lot of testing and finding nothing, probing suggested that the TIP30 output transistor had the correct signals on its base and collector but the emitter was not supplyong -15.8V. I tested the device and it was fine, apparently. I replaced it and the rail returned. The tester was not identifying the TIP30 as bad. 

OK, connect logic and display PCB and... no display. Checking the rails and the -15.8V was there. This time the -5V rail was not there. After some testing of supply resistances on the PCBs I found that the display PCb had a tantalum on it and it was shorted. I then replaced all the tantalums I had missed the first time round. 

Do not power the printer without the control signal ribbon attached 

Power up again and:

 

 Success, it's back again.

The printer is also working:

 Unfortunately I found that the ribbon cable from the logic PCb to the display has an intermittent break in it, so I will have to patch that, but otherwise it looks fine. I now need to put it all back together and case it up. Hopefully that won't disturb too much.