HP-97 troubleshooting

[Kees Bouw]

Hi,

Some general questions about HP-97 troubleshooting.

My newly restored and cleaned HP-97 shows only the low battery light when switched on. The battery is not actually low, and besides the same thing happens when I connect the charger. What would you check?

Is it safe to operate the HP-97 with the printer unit and the card reader unit not connected or removed?

The card reader is located in a such a position that it is impossible to measure anything of fine-tune it. Is it possible to have the card reader unit connected from outside the calculator? Can it still work correctly if you connect it with a long flat cable (let's say about 20 cm)?

Thanks very much for any help!

Kees.

[AndiGer]

See here. Step 19 (no. four from below).
Do you have the printer boarder mounted?
Andi

[Paul Berger (Canada)]

(04-06-2020 10:34 AM)Kees Bouw Wrote:  Hi,

Some general questions about HP-97 troubleshooting.

My newly restored and cleaned HP-97 shows only the low battery light when switched on. The battery is not actually low, and besides the same thing happens when I connect the charger. What would you check?

Is it safe to operate the HP-97 with the printer unit and the card reader unit not connected or removed?

The card reader is located in a such a position that it is impossible to measure anything of fine-tune it. Is it possible to have the card reader unit connected from outside the calculator? Can it still work correctly if you connect it with a long flat cable (let's say about 20 cm)?

Thanks very much for any help!

Kees.

As Andi mentions the printer card must be in place as the LC tank for the clock is on the printer card. The low battery LED is turned on by the cathode driver chip so i would suspect it includes circuitry to check battery voltage. Battery voltage is connected directly to this chip, so I would suggest measuring the battery voltage at the cathode driver in case there is a bad connection somewhere. Are you measuring the battery voltage with it installed and the unit turned on? That would be the only way to reliably know if the batter voltage is sufficient or not since something inside may be pulling it down when you turn the unit on.

I suppose you could operate the card reader outside the unit but I am not sure why you would want to do that. I would avoid unplugging the cable to the card reader if possible, those cables are now all pretty old and are prone to falling apart. If you do unplug them you need to slip a piece of thin plastic in between the cable and the connector contacts before pulling out the cable otherwise the pins will dig into the cable and almost certainly wreck it. I use pieces of PET cut from plastic bubble packs.

Paul.

[Kees Bouw]

(04-06-2020 04:13 PM)AndiGer Wrote:  See here. Step 19 (no. four from below).
Andi

This way the calculator is upside down on the bench. That makes it very difficult to operate.

[teenix]

The battery LED is controlled by circuitry inside the cathode driver chip and a voltage set resistor connected to pin #15 to ground.

This is a measurement I took on a 97, although this may vary a bit between calculators.

The low battery LED begins lighting at 4.77 volts and is fully lit at 4.6 volts

The battery voltage for this circuit comes from pin #4 of the 9 way connector on main circuit board just below the CRC chip. There is another unused 3 pin connector to the right of this connector. (Be careful playing around here with probes if the calculator is turned on or off as any inadvertent short could kill the calculator). Looking left to right, pins 7 and 8 are live battery connections.

It is difficult to fault find the 97, but with the rear cover removed, you can put a heavy elastic band (or similar idea) around the battery to hold it in place when the calculator is upside down. You could also construct some clips and leads to an external battery.

There is some bare circuit board space just to the left of the 9 pin connector, I would try putting light pressure, on and off, here with a finger to see if the LED changes. If so it might just be a poor connection and requires some cleaning. Do not try to bend the connecting pins to get a better connection, they are brittle and will snap apart.

The voltage set resistor that feeds the cathode driver battery sense pin may have poor connections or has changed its resistance. It is listed as 200K or 330K depending on what was installed.

If these check out ok, then the fault may lie in the cathode driver. Even so, you might be able to fiddle with the resistor value to get it to work. It is difficult to get to it though, requiring removing the main circuit board and underlying plastic panel.

cheers

Tony

[Paul Berger (Canada)]

In addition to Tony's posting above, when I was doing serious troubleshooting on a 97 with the back off I had a stiff plastic strap that I ran over the battery pack and fastened it down with a couple cover screws to hold the battery in place. I also operated one with the top cover removed by detaching the display and keyboard from the top and just having it all sit on the black mid frame, providing insulation where needed.

Paul.

[Kees Bouw]

Thank you all who replied. A short status update:
I connected the calculator to my lab power supply at 2,5 V just to see what the current would be. The current was 0,5 A which is about twice as much as it would normally use at 5 V.
So there was no bad contact, but instead someone was using more than their fair share of current! There was no magic smoke or even the slightest smell, just like when I had connected it to the battery. I suspected the printer because that uses a lot of current if it is in use. So I disconnected the printer and also the printer pcb, knowing the calculator would not work without it. Normal current of about 130 mA at 5 V. Then I connected the printer pcb back but without the actual printer, still normal current and a working calculator!
So what could it be? Easy to suspect the printer motor or the thermal head, but if either is bad would that be an issue if it is not activated from the pcb? Maybe the pcb is bad and it continuously sends current through all 7 thermal points? I will have to do more measuring, there are some hints in the service manual.

Kees.

[Paul Berger (Canada)]

(04-07-2020 07:44 PM)Kees Bouw Wrote:  Thank you all who replied. A short status update:
I connected the calculator to my lab power supply at 2,5 V just to see what the current would be. The current was 0,5 A which is about twice as much as it would normally use at 5 V.
So there was no bad contact, but instead someone was using more than their fair share of current! There was no magic smoke or even the slightest smell, just like when I had connected it to the battery. I suspected the printer because that uses a lot of current if it is in use. So I disconnected the printer and also the printer pcb, knowing the calculator would not work without it. Normal current of about 130 mA at 5 V. Then I connected the printer pcb back but without the actual printer, still normal current and a working calculator!
So what could it be? Easy to suspect the printer motor or the thermal head, but if either is bad would that be an issue if it is not activated from the pcb? Maybe the pcb is bad and it continuously sends current through all 7 thermal points? I will have to do more measuring, there are some hints in the service manual.

Kees.

Was there paper in the printer? If there was and current was continuously flowing through the print head there should be a mark on the paper, also since the print head is not designed for 100% duty cycle I suspect it would not last long. Another possibility is the home position switch is not working. When the 97 is first turned on if it detects the print head is not homed it will turn on the motor in reverse to bring it home and if the home switch never closes it will drive forever, mind you if your printer still has the nylon idler that will soon take care of itself by stripping teeth off the gear. This switch is a reed switch and is activated by a magnet on the print head carrier.

Paul.

[teenix]

If sounds like print head element(s) as you don't mention the motor keeps trying to turn.

By turning on the power then off shortly after, you should be able to feel the print head and see if it gets warm or hot.

If it is heating up then...

one or more of the print head driver ICs' may have failed.
the PIK chip could have partially failed and is turning on the heads in error.
there is a short one of the boards

All the printer circuit board pins #8 - #14 should be close the 0 volts. If they are higher than (guessing) 3 volts then the PIK chip may be at fault.

If the print head flex cable is removed, then you can take your time testing these pins. Test with the printer board connected and not connected. It may point to which board is at fault.

Pins #3 and #4 are the ground pins for the printer board.

cheers

Tony

[Kees Bouw]

Hi All,

The mystery is solved.

I did a lot of testing and measuring, everything seemed quite normal with paper in the printer except for one thing: the paper advance button had no effect at all. Inspecting the connections on the switch and on the main board and comparing it with the service manual revealed that I made a mistake when connecting the switch to the main board. In my own defence, it was not really a stupid mistake. I am very careful with these things :-) I had looked at a photo from another HP-97 and connected the wires accordingly. But in that particular HP-97 the wires were connected differently from what the service manual says, something I had not noticed before. In the service manual it shows that the wires should be R B W (red black white) from top to bottom on the main board. But in my older HP-97 it was connected as B W R. To prove this I attach pictures of both switches which were wired differently as well of course. Left is the "irregular" one and at the right as it should be according to the service manual. The way I connected it the F1 from the ACT was connected to the OPS all the time. I do not remember exactly but the strange behaviour may indeed have been related to whether or not paper was present in the printer, meaning that way too much current was used when there was no paper and more-less normal behaviour with paper in it.
Anyway: a word of warning, be aware that the paper advance switch may not be wired as shown in the service manual!
Fortunately, no damage was done and after connecting it correctly everything worked fine.

Kees.

[teenix]

Well done, good news.

That makes sense now. Wiring the switch incorrectly would mess things up.

The switch has two functions.

When pressed it pulls the CRC F1 flag line low which tells the microcode you want to start doing paper advances until the button is released.

When not pressed, that ground signal goes to the out of paper switch which is normally open circuit when there is paper in the printer. When the paper is removed, the out of paper switch shorts and places this ground onto the FWD drive signal from the PIK chip, effectively stopping the motor until it reverses.

It makes sense to wire it this way as the out of paper switch is prevented from stopping the motor when the paper advance switch is pressed without paper in.

Not sure why the excess current as you didn't mention the motor was running. Depending on the way the switch wring was the PIK and CRC pins could have been shorted, or shorted to ground which may have cause a problem.

Anyway, all seems good.


cheers

Tony