HP 97 latest PCB with blown capacitor : Alive !

[Chr Yoko]

Hello,

I just got a very nice HP 97 mechanicaly speaking, but unfortunately non-functionning.
It is a model with the same PCB as the 97S one, so quite different to the initial 97.
The initial version of the HP97 Service manual is easy to find and to download.
It is unfortunately showing a different PCB with a slighly different power circuit than on my HP97....

Here is the PCB type (a picture i found from a Katie Wasserman message years ago) :



And on her picture, we can see a "yellow-purple" condensator. This is the one blown on mine :



Does anybody know its value ? color code seems to say 47µF or maybe 4.7µF ... ?
How to check and fix that power circuitry ?
I would like to get it working our of a standard 5V DC USB style power, would be great.

Thanks for advices

[AndiGer]

I once draw the new layout and published on this very site for download:
https://www.hpmuseum.org/forum/thread-10...#pid107514

[Chr Yoko]

(12-06-2020 05:26 PM)AndiGer Wrote:  I once draw the new layout and published on this very site for download:
https://www.hpmuseum.org/forum/thread-10...#pid107514

Fantastic !
Many thanks, what a great forum 8-)
=> hum, i would then recap all of thoses capacitors

[teenix]

(12-06-2020 05:12 PM)Chr Yoko Wrote:  Hello,

Does anybody know its value ? color code seems to say 47µF or maybe 4.7µF ... ?
How to check and fix that power circuitry ?
I would like to get it working our of a standard 5V DC USB style power, would be great.

Thanks for advices

The power supply circuit is probably the same, so C6 is listed as 47uF and is a tantalum type.

There will be a fair bit of current draw if the printer is used so the USB will have to capable of supplying this.

cheers

Tony

[Chr Yoko]

Hi Teenix,

Correct, i will first try to revive it with a bench power supply after recaping it with some Tantale capacitor.

If it works again, then i will see for the new 5V power. Should not be an issue custom or off the shelf.

That 97 was sold without any transformer, the previous owner was trying to running it with an external old phone supply .... hope he did not fried the Roms or else.

There are also a bunch of transistors on the PCB, i need to find their modern references and see if i need to replace those as well.

[teenix]

(12-07-2020 08:47 AM)Chr Yoko Wrote:  Hi Teenix,

Correct, i will first try to revive it with a bench power supply after recaping it with some Tantale capacitor.

If it works again, then i will see for the new 5V power. Should not be an issue custom or off the shelf.

That 97 was sold without any transformer, the previous owner was trying to running it with an external old phone supply .... hope he did not fried the Roms or else.

There are also a bunch of transistors on the PCB, i need to find their modern references and see if i need to replace those as well.

Are there part numbers written on the transistors ie 1854-0071 ?

cheers

Tony

[Kostas Kritsilas]

if you are going to replace the tantalum capacitor, you may want to use an electrolytic instead of tantalum. Tantalum capacitors are not tolerant of either reverse voltage, or voltage higher than what they are rated for, and electrolytic capacitors are less sensitive to either. Tantalum capacitors were used because they offered a lot of capacitance in a small size, but with time, electrolytic capacitors have also improved in that regard.

[Chr Yoko]

Thanks guys for your support !
I already ordered a pack of Tantalium capacitors, maybe i will look also for some good Aluminium-Electrolyte ones too.
My plan is to re-open the 97 this WE, investigate , take some pictures, change some caps.
Will keep you posted.
Maybe this will be the reason to finaly get an oscilloscope ;-)

[Kostas Kritsilas]

(12-09-2020 12:10 PM)Chr Yoko Wrote:  Thanks guys for your support !
I already ordered a pack of Tantalium capacitors, maybe i will look also for some good Aluminium-Electrolyte ones too.
My plan is to re-open the 97 this WE, investigate , take some pictures, change some caps.
Will keep you posted.
Maybe this will be the reason to finaly get an oscilloscope ;-)

Sorry, I should have put this into my previous post.

The tantalum being blown may be an indication that the power supply was hooked up backwards. I could be wrong of course, but hopefully, if it was, the tantalum blowing up may have saved the rest of the electronics. It really depends on the tantalum failure mode, how long it was powered on, and if it was hooked up backwards, where the tantalum is in regards to where the power input is relation to the rest of the electronics.

Not looking to alarm you, but as far as I know, the tantalums used in the HP-97 are not a known problem area.

[John Keith]

(12-08-2020 07:56 AM)Kostas Kritsilas Wrote:  if you are going to replace the tantalum capacitor, you may want to use an electrolytic instead of tantalum. Tantalum capacitors are not tolerant of either reverse voltage, or voltage higher than what they are rated for, and electrolytic capacitors are less sensitive to either. Tantalum capacitors were used because they offered a lot of capacitance in a small size, but with time, electrolytic capacitors have also improved in that regard.

Electrolytic caps have improved a lot in recent years but you still have to choose carefully if you are replacing tantalum caps. Low ESR types with good high frequency performance, e.g.solid electrolyte caps should be used.

For USB power a good choice is the Raspberry Pi 4 power supply. 3A output, USB-C and reasonably priced.

[[kby]]

-97s use essentially the *&!@? Cheap power bricks thst are only A/C transformers; no rectification or regulation, but it does mean you can’t hook it up backwards.-kby

(12-09-2020 11:07 PM)Kostas Kritsilas Wrote:  

(12-09-2020 12:10 PM)Chr Yoko Wrote:  Thanks guys for your support !
I already ordered a pack of Tantalium capacitors, maybe i will look also for some good Aluminium-Electrolyte ones too.
My plan is to re-open the 97 this WE, investigate , take some pictures, change some caps.
Will keep you posted.
Maybe this will be the reason to finaly get an oscilloscope ;-)

Sorry, I should have put this into my previous post.

The tantalum being blown may be an indication that the power supply was hooked up backwards. I could be wrong of course, but hopefully, if it was, the tantalum blowing up may have saved the rest of the electronics. It really depends on the tantalum failure mode, how long it was powered on, and if it was hooked up backwards, where the tantalum is in regards to where the power input is relation to the rest of the electronics.

Not looking to alarm you, but as far as I know, the tantalums used in the HP-97 are not a known problem area.

[teenix]

(12-07-2020 08:47 AM)Chr Yoko Wrote:  Hi Teenix,

Correct, i will first try to revive it with a bench power supply after recaping it with some Tantale capacitor.

If it works again, then i will see for the new 5V power. Should not be an issue custom or off the shelf.

That 97 was sold without any transformer, the previous owner was trying to running it with an external old phone supply .... hope he did not fried the Roms or else.

There are also a bunch of transistors on the PCB, i need to find their modern references and see if i need to replace those as well.

The original charger (as noted by Kby) is only an AC transformer intended to charge the battery. I'm not 100% sure, but I don't think it was intended to power the entire calculator especially the printer and card reader. If this is the case, connecting a USB power supply to the charger terminals may not work as intended. Also, there is a bridge rectifier in the circuit path, which which will rob 1.2 volts from the input supply, so 5V will become 3.8V. This rectifier would have prevented a reversed DC voltage from affecting the calculator if somebody did connect this voltage to the charger input.

The high current battery rails are accessed from two terminals inside the calculator and are marked (+) and (-) near the charger input. I have connected a 5V 3A supply to these pins (with the battery removed) without problems. Care is needed here, as there is no reverse voltage protection and a mishap will cause damage.

Note that the HP-97 service manual states that, even though the calculator will not run properly, it is ok the run the calculator from the AC charger without the batteries installed, however, the owners manual states that "damage" may occur in this situation.

There is a flow chart for fault finding the power supply in the HP-97 service manual on page 4-14 (Figure 4-8)

cheers

Tony

[AndiGer]

@Chr Yoko. What I wonder: how did you come to the conclusion C6 is blown?
What happens when you try to power the calculator from the batteries? Does it show up with the battery low LED lit and nothing else?
Andi

[Chr Yoko]

(12-11-2020 04:33 PM)AndiGer Wrote:  @Chr Yoko. What I wonder: how did you come to the conclusion C6 is blown?
What happens when you try to power the calculator from the batteries? Does it show up with the battery low LED lit and nothing else?
Andi

Battery was at 0V....
C6 was blown (burned), i removed it and ordered a set to replace it.

So today :

- I removed the broken power connector.
- I replaced the 47µF C6

and the 97 works again.... :-)

... well not yet at 100% since it is still an old lady :

- Works on power coming from a bench power supply @4,8V and 1A limitation connected to the battery connectors with crocodiles (need be carefull about the + / - orientation !!). Will need to install a new connector replacing the old one, maybe USB style so i can use a standard USB charger. Or just run it on the battery that i would charge outside of the 97.

- Battery pack supplied with the calculator seems almost alive after i charged it a bit with the bench power supply. I need now to build a "charging stand". However, i noticed that one of the cell is hitting over 45°c if i send in over 0,7A (slowly @ 1A and very quickly @1.5A !)..... It must have got a high resistance symptom, maybe dead. Any idea about the max A i can inject into those obsolete Ni-Cd cells ?

- The LED display works except one missing element on one digit : "8" => "6" : A connection issue between the MB and the display driver ? Something else ?

- The magnetic card reader makes a strong noise (like a chain saw) and display shows "error" ... i disconneted it, so it is quiet now (flat connector). Will have a look, but no big deal if dead.

- The Printer is very slow/sluggish to move paper and lines overlaping : maybe i need to check its mechanical parts/wheels.

So i still need a bit of work, but i am already very statisfied ! 8-)
Will post pictures soon.

And i will get a cheap vintage oscilloscope tomorrow , a Tektronix 2205, so i can check other things if necessary 8-)

[teenix]

(12-13-2020 08:34 PM)Chr Yoko Wrote:  - The LED display works except one missing element on one digit : "8" => "6" : A connection issue between the MB and the display driver ? Something else ?

Most likely a dead LED segment in that display. If so, not repairable.

cheers

Tony

[mfleming]

It's not terribly difficult to rebuild the battery pack, the hardest part being the separation of the two halves which seem to be heat welded together. A little time with a jeweler's saw or Xacto knife will do the trick. Substitute subC NiCds with flattops and welded tabs can be found. I soldered tabs together and sealed the case with electrical tape on a couple of packs, but was thinking of rigging a pressure contact arrangement so the cells could be removed and charged separately.

Restoration is a great way to while away the time

[[kby]]

(12-13-2020 08:34 PM)Chr Yoko Wrote:  [quote='AndiGer' pid='140090' dateline='1607704433']

- The Printer is very slow/sluggish to move paper and lines overlaping : maybe i need to check its mechanical parts/wheels.

The plastic wheels on the feed roller are probably glazed. There have been articles posted elsewhere on this forum, although I ended up using the same square o-rings I like for card reader rebuilds, which it sounds like you also use. I did have to file those down a bit; did those by putting the axle pin of the roller in a Dremel chuck to spin it then held a small flat file to allow it to sand it down. Unfortunately it doesn’t work so well on both ends; too much play so end furthest from the chuck has too much wobble. It would be nice to do both sides at once since they have to be the same diameter. The actual diameter isn’t that important but if they are not the same the paper could track crooked. It didn’t seem too hard to get it to work though. You can test for being level by rolling the roller down a flat slight slope and observing that it doesn’t veer to one side or the other; that seems good enough. Make sure you remember the orientation (left right) of the roller when you take it out. It us ratcheted to roll in one direction internally so won’t work if inserted the wrong way. There are some gaps in the plastic that surrounds the central metal axle; you want the smaller gap on your right as you face the calculator (the “home” position of the calculator is to the right as you face the front of the calculator with the printer on the right hand side and keyboard facing you.

[teenix]

(12-13-2020 08:34 PM)Chr Yoko Wrote:  - Works on power coming from a bench power supply @4,8V and 1A limitation connected to the battery connectors with crocodiles (need be carefull about the + / - orientation !!).

- The Printer is very slow/sluggish to move paper and lines overlaping : maybe i need to check its mechanical parts/wheels.

1A supply is probably not enough to drive the printer properly. Each print head element draws around 0.5A and there will most likely be more than one active while printing, and then there is the motor current and calculator CPU etc.

I used one of these...

https://www.jaycar.com.au/5v-dc-3a-slim-...s/p/MP3480

cheers

Tony

[Chr Yoko]

Hi,

Hereby some pictures about he HP97 revival progress made so far :





I just fixed the printer paper advance issue by replacing the cracked old elastomeric friction wheels by sections of rubber tube just pressed in. To do so i had to desassemble completely the printer mecanism shafts. It was not difficult. Cleaning and added some silicon oil on the wheels and shaft. Works as new now.

I disassembled the battery pack and its NiCd obsolete cells from ... 1986 that were showing high internal resistance and ordered some fresh NiMh same size cells from a vendor in Italy. Just waiting for the delivery.

For now, the 97 is working from a bench power supply connected via crocodiles to the battery connectors.

[Chr Yoko]

Hereby the missing display segment (5 instead of 9) and the printer fix before/after :


And an high tech home made independant battery charger :