Dead HP-67



I've been using several HP calculators for over 20 years without a single problem. Yesterday I was using my HP-67 to perform scientific engineering calculations at work.

After I performed a calculation I switch the machine off, to conserve batteries. I switched it on to perform another calculation and was greeted by a blank display. I repeatidly switched the machine on and off with no joy.

I noticed that each time I switch the machine on or off the battery low LED flickerd. I have not opened the HP-67 yet, therefore, this problem isn't a result of anything I have done. The machine is in immaculate condition, hence, it has not been used much by the previous user. Even the blank magnetic cards and standard application cards look unused.

Is it possible that this problem could be related to an active component in the switch mode power supply breaking down or a passive component such as a capacitor becoming more resistive through aging?

If it is a power supply problem, then could it cause any damage to the circuitry of the HP-67?

The battery packs been checked and is OK. The machine is only used by me and no one else, mainly, because non of the people in the analogue design section understand RPN.

Any advice would be greatly appreciated.

By the way, this is the same HP-67, mentioned in the forum, with the missing display segment, but it does not affect my calculations because I round of all calculations to 3 significant digits. Nevertheless I still plan to replace the defective diplay module when I acquire a classic series machine. Many, many thanks to all who helped in solving that problem.

Thanks & Best Regards


The problem could be related to a dirty power switch contact, charger plug bypass strap (that gold strip that connects the two outer pins of the charger port when the plug is removed), corrosion, etc (often invisible) on the battery contacts, etc. The battery touchs the EDGES of the contacts near their tips. Does the machine work on the charger alone? Perhaps the battery is dead (can be rather unobvious if the charger has problems).


Could you tell me more about this shortening strip? I have a dead 67 that does not have such a strip in the charger port. Do all 67s have this thingy? What does it look like?



Andreas -

A good description of the shorting strip is over in the "Collecting & Using" section on this website:

Scroll down to the Classic series. There's an illustration, as well as a link to a photo of the battery compartment, which shows the charger port at the top (in an HP-67 the battery compartment is not adjacent to the port, of course).

The strip just connects the two outboard pins when the charger connector is not in place. You should be able to fashion a strip out of a small piece of thin brass or beryllium-copper stock, folded over to give the proper "spring" action. Unfortunately, to stay in place it has to clamp onto a plastic lip in the calculator. This lip is readily accessible in an HP-35,-45,-55, but may require disassembly of a -67 to reach. There are articles in "Collecting & Using" on getting to the innards of a -67, and I'll let the experts weigh in...

Good luck,



Your fault could be one of many things, some easily repairable, others would need parts from another HP calculator.

Here's what I would do. Open up the machine. To do this, take out the battery pack, flick out the 2 plasitic plugs near the charger connector, and peel up the rear (battery pack end) corners of the back label. Take out the 6 screws (2 under the label, 2 under the plugs, 2 under the front feet which can be lifted to get to them). Take off the bottom case.

Now power it up from a bench PSU clipped onto the battery contacts. I am pretty sure the rear one, nearest the display/charger connector, is -ve, but you can check by looking for continuity between the battery contacts and the middle pin of the charger connector. The one that's directly connected to the middle pin is certainly -ve. Use about 3.75V to 4V.

Now check the power converter outputs. The easiest place to find these is on the pins of the ACT (CPU) chip. This is the 22 pin, 0.4" wide chip on the logic PCB. The common reference point for all measurements is the -ve battery terminal. Anyway, pin 1 of the ACT (Vss) should be at about 6.2V, pin 2 (Vgg) at about -12V. Check these, and if you have a 'scope, look for ripple too. They should be stradyt DC voltages.

If those are right, the PSU is working. If not, then it's not. In the latter case it can most likely be repaired (FWIW, 2N3094 transistors are suitable replacements for the ones in the power converter).

If the PSU is fine, I'd then look at the 2-phase clock outputs from the ACT. These, IIRC, are pins 16 and 17 of the ACT. If they're missing, then most likely the ACT is dead, but check the LC tank components connected to pins 13 and 14 of the ACT.

If you've got this far, then most likely you have either a dead ACT (HP custom, you can get a replacement from an HP21, 22, 25, 25C, 29C, 19C, 91, 92, 97 or 97S -- NOT a 27 though), or ROM0 (the 18 pin chip on the logic board, and this has to come from another 67).


Thank you Tony, that's extremely valuable information. I'll report after the operation!




Thanks for the info Tony. During the weekend I opened my HP-67 and probed the test points: -

pin 1 6.199v
pin 2 -11.659v
Frequency measured on pins 16 and 17 - 177.9KHz

During probing I had an earth strap around my wrist to prevent any damage to the ICs. Futhermore, I used a fixed battery supply of approx. 3.7v to power the HP-67 during test. I have an analogue oscilloscope but did not use it to view any of the above signals. If you think it necessary then I will do so.

Are the above voltages and frequency within the design limits. During test I found that the display intermittently switched on but I was not able to use the keys. Other than that it returned to the state where the low batt led would flicker when the machine was switched on or off.

Is it possible that none of the ICs are damaged and the capacitors may have aged outwith their designed specifications? John Garza mentions this in another forum message. I know that ageing electrolytic capacitors may cause product failure but does the same apply to tantalum capacitors? Do you think that replacing the tantalums may help? Are there any electrolytic caps on the display board? Is there a separate clock that is used to drive the LED display board? (this clock may rely of a capacitor)

I cannot make out all of the capacitor values on the PCBs. Would you or anyone have a list of the capacitors and their values? I will not replace the capacitors until I am absolutely certain they need to be replaced? (I am very tempted!)

I have not disassembled the HP-67 any further and will only do so if necessary.

I would also like to hear from Marais - what were your findings?

I know I sound over cautious, but, thats because I realy admire the HP-67 and don't wish to do anything in a hurry that may cause it serious or irreversible damage.

Thanks & Best Regards
Imran Farooq


Those voltages and clock frequencies sound spot-on. I would be suprised if the problem was a capacitor, but you can replace them if you really want to. The tiny ones (2 in the Vgg voltage doubler, one in the CPU reset circuit) are 2.2uF. The one on the Vss line (often yellow or brown) is 47uF. Tantalum caps do fail with age, but normally you either have problems reading cards (noise on the supply lines) or the Vgg line is missing/incorrect. It can't hurt to look for ripple on the Vss and Vgg lines with your 'scope. If they're smooth then the caps are almost certainly fine.

There is no separate clock on the display board. The display board contains 2 chips. One is HP custom and is the display cathode driver (scanner). The other is a heptal (7 stage) transistor array (it's either a CA3081 or a CA3082, I forget). The only clock in the machine is the LC tank circuit on the ACT.

I hate to say this, but your machine is showing the classic symptoms of ACT (CPU -- the 22 pin chip) or ROM0 (the 18 pin chip on the logic board) failure. The easiest way to check is to replace them. You can try figuring out what's going on by grabbing the ISA line with a good logic analyser, but without a ROM source listing you are not going to get very far.



You speak very often about Tantalum caps on the motor driver card/board, but: what are the values of these caps? Thanks,



The only bad caps I have ever seen in an HP machine are the little blue dipped tantalum cap on HP65/67/97 card reader board. When it goes bad the reader will write but not read.

The only other one was a filter cap in an HP29C power supply. It was a physical break in the cap lead bonding.

If it ain't broke, don't fix it... you can easily do more harm than good.

ACT chips in the HP67 and HP97 almost never go bad. These machines have real power supplies unlike the chip gobbling travesty in the Woodstock machines.


I have had bad caps in HP power converters. In particular I've seen an HP29C that would intermittantly mis-write to program memory (it always read consistently, but sometimes the wrong keycode was written) that turned
out to be a bad cap on the Vss or RAMVss line (I forget which, I replaced the pair). But it's not a common fault.
As regards dead ACTs, I've had at least one 67 and one 97 that needed a new ACT. Actually, I think I've seen every chip on a 67 logic board fail apart from the card reader controller. I've certainly replaced ACTs, ROM0s, other ROM/RAMs.
There is also little difference between the power converter in the Woodstocks and the 67s. Apart from the well-known problrm in the -C models if the battery pack goes open-circuit with the charger connected, there is _no_ reason why a woodstock should kill mroe chips than a 67.


At least the HP67 charger has some electronics in it. When running off the supply it has filtered and regulated juice. The woodstocks get slammed with whatever that unregulated beast throws at them. I have had one bad ROM0 chip. I suspect that this chip seems to fail because it sucks a lot of power to drive the display. Also had a bad 8 pin ROM in an HP97 and a bad PIK.

I have collected over 20 bad HP25 machines. Almost all with bad ACT's, many also with bad RAM chips. Strangely, HP21 machines seem much more robust. Out of a couple dozen only two or three had problems. .

The bad cap in the 29C power supply caused the power to be a square wave. The display was dim and flashing. A meter showed proper voltage. A scope showed the evil.



In my previous posting when I measured the voltages on pins 1 & 2 and the frequency on pins 16 & 17, on the ACT chip, everything checked out OK. I used a true RMS digital multimeter.

This time I think I'll probe the supply rails with my analogue oscilloscope to see if there is any noise. Maybe, just maybe, a bad cap or contact may be inducing noise on the supply rails. This noise maybe enough to force the digital logic into some corrupt state.

Thanks & Best Regards


>I would also like to hear from Marais - what were your >findings?

I won't be able to report before mid-june. My test equipment is in my parents' house; and I don't go there very often. I'll spend a week there in June, with my 67, and another week in July with my most recent acquisition, the 9810. I'll report then! (And thanks again to Tony)


Possibly Related Threads...
Thread Author Replies Views Last Post
  Previously dead WP 34S is resurrected! JayV 2 638 04-18-2013, 12:41 AM
Last Post: Walter B
  Does anyone want dead 82001 battery packs? Michae Altmann 6 977 03-15-2013, 07:32 PM
Last Post: Harald
  Dead 50g after 3 months Chris Smith 13 1,561 03-13-2013, 12:28 PM
Last Post: chris smith
  More or less dead HP-67 Bernd Ulmann (Germany) 7 1,068 02-09-2013, 12:53 PM
Last Post: Ignazio Cara (Italy)
  HP-29c keyboard dead 2nd row problem Xavier A. (Brazil) 6 1,065 07-17-2012, 07:09 AM
Last Post: Xavier A. (Brazil)
  WP 34S and 'dead' HP20B Rob Willett 17 2,115 10-20-2011, 06:49 PM
Last Post: Egan Ford
  hp-12c from arrives with dead batteries Masaki Adachi 4 785 09-30-2011, 05:25 PM
Last Post: Masaki Adachi
  The calculator is dead - viva la HP 15C Limited Edition Joerg Woerner 38 3,984 09-04-2011, 08:44 AM
Last Post: hpnut
  offtopic: is HPCALC.ORG dead? x34 2 578 03-15-2011, 06:44 PM
Last Post: DeboT
  30b Dead 5 Key Norman Dziedzic 10 1,213 02-03-2011, 03:36 PM
Last Post: Norman Dziedzic

Forum Jump: