HP Forums

Full Version: Help! HP-19C printer gone mad
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

Hi Guys,

Here I am again asking for help, but this time for an HP-19C.

I’ve just managed to acquire (at last!) an HP-19C.

The seller I got it from (yes, on TAS) was the original owner, but said that it hadn’t been used for 30 years. The battery pack was removed then (allegedly), and everything put in a drawer.
Of course, the battery pack is long dead, and it had leaked, but the calculator was apparently intact (and no trace of white dust from the battery cells inside the calculator).
Using the dead battery pack’s shell, I replaced the dead cells with fully charged NiMH.
I inserted the battery pack, making sure the connections to the terminals were ok (voltmeter in hand), and turned on the calculator.

That’s where things go wrong. The printer head just travels across and comes back, and the paper advances, a small pause, and this repeats itself. There’s no way to stop it. Nothing gets printed. It doesn’t matter that the printer switch is on MAN/TRACE or NORM. It doesn’t matter that the power switch is on PRGM or RUN.
For the short moments that the printer head is not moving between cycles, the display lights up seemingly properly. If I am fast enough (in RUN mode) I can enter some digits, but no function works.

I opened up the calculator, and didn’t find much wrong that I could see. There is what looks like spots of green corrosion here and there on some tracks of the power supply/printer PCB (from battery’s degassing?), and the power circuit’s large resistor has made the PCB black, but that’s all.

Looking around I found this, but the battery *is* working fine and full of juice, so it may be the same symptoms but doesn't look like the same cause.

The HP-19C is in otherwise excellent physical condition. It would be an utter shame if it cannot be made to work again. I sent an e-mail to fixthatcal, but Randy hasn't replied yet.

Should I clean all the contacts etc. (DeoxIT, slight vinegar wash for green spots, Isopropanol, etc.) before I cry for help?

Any idea or suggestion? (no, I haven't got an oscilloscope :-(

Hopeful regards,


It is most likely a power supply problem. Either the Vgg or Vss side of the supply is failing. I have 2 of them on the bench right now trying to troubleshoot the same problem. So far I have replaced one open NPN transistor. I hate to work on this model. I have great difficulty trying to marry the logic board back to the stakes on the power supply board. If you are not very careful the stakes can easily get bent when you try to close the two halves back together, and I have not found a source for stake replacements. I won't even get into the flaky contacts on the flex wire circuit between the keyboard and the logic board. An intermittent contact here can drive you crazy. Good luck with your repair.

Edited: 11 Mar 2011, 5:42 p.m.

I agree that it's some sort of power supply issue, but may not be the internal supply. Before you start digging around in there (these are a pain to work on) try using an external 5 or 6 volt power supply in place of the batteries, one that can supply at least couple of amps. It might be that your current cells just can deliver enough current for this calculator.


Hi Katie, Donald,

Thanks for replying. The fact that you both think that it's still a power supply problem is reassuring, in a way, however I find myself very ill-equipped to proceed; I'd like to have an oscilloscope to be able to monitor various points, and a bench power supply so that I can feed whatever I need - I have neither.

On the power front, the NiMH cells I used are of good quality and were fully charged. At 2200mAh, I would have expected them to feed enough current for the motor.
Of course, there could also be a problem at the PCB contacts with the battery I suppose, however I monitored the voltage on the points' solder and that looked fine, although there was some voltage drop during "printing".

I will try to re-tin the PCB contacts with some solder, and also try to rig up 4xAA new alkaline batteries straight to the PCB tracks. I'll let you know how I get on.

Donald, you mentioned that you replaced one open NPN transitor. Had it failed? Why would that be? Do you have the schematics?

Many thanks,


Before you point out that the battery's capacity of 2200mAh has nothing to do with current delivered, yes, sorry, I realise that; I expressed myself incorrectly. What I meant was that such batteries should be able to output a couple of Amps easily.



see here

cause of transistor failure, unknown.

Edited: 12 Mar 2011, 1:03 p.m.

Thanks. I'm an HPCC member, but I didn't spot these ;-)

Some of Tony's schematics are available for download here, just search for 'schematic', however the 19C is not among them. For the full set of schematics (including the 19C) you need to contact Dave Colver at the link that Donald provided.


I have had some success last night.

I soldered leads directly to the battery contacts on the PCB, reconnected the PCB to the keyboard/CPU PCB, and connected my leads to four fresh AA alkaline cells outside the calculator. Effectively I have an HP-19C without its back, and I have to be careful with the printer (it's not held by anything) and its head's flex connector.

As soon as I switched on, I knew there was progress, as the printer stopped its merry dance.

Printing itself doesn't work. There are just a few pixel columns getting printed (from the right). At one point last night, after wiggling the various contacts between the PCBs, I got the printer to work properly (printing the stack, printing statistic registers, TRACE operating as normal), but since then it hasn't worked anymore.

I am not sure that it was the "wiggling" that made printing work, and I hope that I haven't damaged any electronics. It feels as if the data to print isn't making it to the printer's electronics.

That short period of normal operation at least proved that there is no burnt filaments in the print head. It also proved that the printing *could* work, at least with the assumption that I haven't caused any damage.

The calculator itself (i.e. with printing on MAN) seems to be working fine, STO/RCL works, various operations +-/* Log Ln etc. work, I was able to write program steps and run them.

Of interest: according to my meter, the calculator draws between 100mA and 140mA when 'on' (depending on the number of lit digits). When printing a line, this jumps to 160mA to 180mA. Considering that I'm feeding it 6V or slightly over, this is just about 1W being used, as stated. At one point I had the printer flex connector impaired, and the current jumped to 220mA or so. All this seems reasonable.

Would the concensus be now that I have got a connectivity problem between the PCBs? Surely it cannot be a power problem anymore?


Forgot to mention: I also cleaned the various spots of green corrosion that I could find on the power/printer PCB.

The contacts between the PCB's are all gold plated so they're not likely to be a problem, but since you mention the "green corrosion" I'm betting on that being the problem. You've likely got one or more bad solder joints due to this corrosion. Clean off all the corrosion and re-solder all the nearby joints and I'll bet that your 19C is back to 100%.

[...] and I'll bet that your 19C is back to 100%

That would be nice ;-)

I haven't got time to spend on it anymore right now, but I'll carry on cleaning corrosion and checking tracks and solder joints.

On another tack, I have been trying to get in touch with Randy at fixthatcalc, but he is not replying. Is he a member of the forum, so that I could try to reach him another way? Ultimately I haven't got the adequate equipment to fully repair these things.



I just spent an hour scrapping off any trace of green oxidation that I could find, checking that tracks and solder joints were ok around each, and soldered a wire that fell off. I dusted off everything, and reconnected the power PCB with the keyboard.

Again, for a few seconds, things seemed to be slightly better; I was getting a few characters printed out. For instance the whole of "PRST" appeared as I asked to print the stack...but not the stack values. Quickly, the printing went back to as bad as it was before.

But I noticed something else (I had vaguely noticed yesterday, but hadn't paid much attention): every "bad" printing zaps Y Z and T, and the motor's pitch changes. So for instance if I am in TRACE mode, with 5 in X, and I press 4 and then +, then 4 stays in the display after the printer flounders. The printer only runs once, instead of twice (as it should to print "4 +" and then "9 ***"), and going to MAN and cycling the stack with R« I notice that Y has been set to zero. The addition operation obvioulsy didn't happen either. A similar observation goes for stack printing. If (in MAN mode) I load the stack with 1, 2, 3, and 4, then do PRST, then the printer runs twice (as if to print "PRST" and the subsequent blank line) then stops, instead of printing the stack. Examining the stack, only X remains at 4, with T, Z, and Y, having been reset to zero.

Considering that storage registers 16-29 and the stack registers Y, Z & T are not in permanent memory and are zeroed when the power is turned off, this looks like a power cut (maybe from overload) is effected somehow when the printer runs.

So here is my conclusion: the printing starts, and some pixel columns get printed. Most often though , a power "cut" occurs, which stops everything. The power comes back on effectively immediately, and because the printer's end-stop switch is open, the printer is driven all the way until the switch gets closed again (starting position). There's no printing that would happen during that "run back to start" stage, and that's actually audible - the pitch of the motor is different and the printhead speed is faster.

Now, as to what is causing this power glitch, I don't know. Sometimes it doesn't happen. If it's a bad solder joint somewhere, I haven't found it (it might help to know where in the circuit it's most likely to happen, but I don't know). If it's a bad connector, I don't know where it is; and considering that the power circuit and the printer are on the same PCB, a bad connector is less likely.

Could it be an (intermittently) faulty component? And if so, which one? It could still be the power supply, but I'm pretty sure mine is ok at the moment.