Service required for HP 9160A Optical Card Reader

[Jurgen Keller]

I was so keen to connect an HP 9160A Optical Card Reader to my HP 9100B. The result was that the 3A fuse in my 9100B failed. After replacing it the calculator seems to work as before (luckily). I have neither the electronic knowledge nor the equipment to track down the problem in the card reader. However, a working card reader would be a super cool addition to my HP 9100 system.

So if there is anybody out there who could service such a device, please contact me.

[TASP]

As always, wait for a consensus before doing anything, however, I've seen fuses blow from apparently not much more than old age, and I'm thinking your fuse in the 9100B was ancient indeed.

Maybe there isn't anything wrong with the card reader ??

As in human medicine, the first rule is to do no harm. Maybe ohm out the optical card reader and see if it measures anything other than a dead short, then reporting back here for the rest to weigh in.

If it does measure a dead short, that is significant, and potentially easy to find.

[Jurgen Keller]

(01-03-2016 06:48 PM)TASP Wrote:  As always, wait for a consensus before doing anything, however, I've seen fuses blow from apparently not much more than old age, and I'm thinking your fuse in the 9100B was ancient indeed.

Yes, that fuse was ancient. However, a few months ago I've successfully connected other peripherals to the 9100B without blowing the fuse.

(01-03-2016 06:48 PM)TASP Wrote:  Maybe there isn't anything wrong with the card reader ??

At least the transport roll was totally rotten. I couldn't even measure the correct diameter. So I replaced it with a gummy ring that might not fit well. However, I've no idea if that can have a bad impact on the correct operation.

(01-03-2016 06:48 PM)TASP Wrote:  As in human medicine, the first rule is to do no harm. Maybe ohm out the optical card reader and see if it measures anything other than a dead short, then reporting back here for the rest to weigh in.

If it does measure a dead short, that is significant, and potentially easy to find.

I have to admit that I'm an electronic nobody. My interpretation of "ohm out the optical card reader" is to measure the resistance of each pair of the interface pins. Correct?

[TASP]

I was hoping you had the appropriate pin out diagram, and we'd be interested in the resistance between Vcc (or what ever it's called on that particular device) and ground.

Some where on this site, or the available materials, there should be a pinout diagram for your equipment, although, I have not much idea where specifically.

[Jurgen Keller]

(01-04-2016 04:16 AM)TASP Wrote:  I was hoping you had the appropriate pin out diagram, and we'd be interested in the resistance between Vcc (or what ever it's called on that particular device) and ground.

The diagram is available in the 9100AB interfacing manual. I will do the measurement tonight and report the values.

[TASP]

Pins 20 and 22 would be the money shot here.

[Jurgen Keller]

I've measured 0 kOhm for pins 20/22. As far as I understand that's the "dead short" you mentioned before. Any ideas how to proceed?

[TASP]

Do we have a circuit diagram for the OCR ?
Are there any caps across the +15 and GND pins ??

Also, any crud in the connector? Frayed wire(s) visible ?


That 50mA output current limit isn't much, BTW.

[Jurgen Keller]

No, I haven't found yet a circuit diagram for the OCR.

I have inspected the connector and the device. The connector is clean, no crud. No lose or frayed wires as far as I can see. However, I found a resistor that looks damaged, see this picture (it's the small one in the middle with a crack). Probably that's not the root cause, just the consequence of another defect. Anyway, it's a starting point.

[jebem]

(01-04-2016 10:21 PM)Jurgen Keller Wrote:  No, I haven't found yet a circuit diagram for the OCR.

I have inspected the connector and the device. The connector is clean, no crud. No lose or frayed wires as far as I can see. However, I found a resistor that looks damaged, see this picture (it's the small one in the middle with a crack). Probably that's not the root cause, just the consequence of another defect. Anyway, it's a starting point.

In addition to what TASP has mentioned:

If you are able to measure Zero Ohms between the power supply pins, it should be kind of easy to find the failed component thanks to the vintage easy to work with PCB.
While a schematic is always useful, for these kind of vintage equipment where we can clearly see the discrete components and read its references and values, it is like your schematic is the PCB itself.

I assume you are using a digital auto-range multimeter. If not, select the lowest resistance scale you can get from your equipment.

Is the reading resistance value steady at "zero" ohms?
Connect the multimeter probes and wait a few seconds. Did the value change?
Reverse the probe leads and repeat the reading. Is it the same?

The reason for this test is simple:
1) Depending on the multimeter type, a "0Kohm" value is normal when a large capacitor in parallel with the power supply lines is discharged - it will start charging from your multimeter and it will take a few seconds before you see the resistance reading start to increase.
2) Semiconductors shows resistance variance (even after failure) dependimg on polarity.

Assuming the zero ohm (or close to zero) is a steady value, then it is time to do a visual inspection and follow the copper traces to find what components are in parallel with the power supply lines.
Usual suspects are: capacitors, diodes, voltage regulators, and semiconductors in general.

In fact, a probable candidate is a voltage regulator IC (integrated circuit) that will have two of its legs connected in parallel with the power supply lines, so it is very easy to spot it.

It can be just a simple 3-leg component (like the TO-220 79nn or 78nn series, or maybe a TO-3 case). These IC regulators can have catastrophic failure and the result is what you are experiencing on that 9160A.

Usual technique to test those components: Use a cutter to remove it from the PCB (or use a soldering iron and a desoldering tool to keep the component leads intact) and test it out of the circuit. If the component is good, solder it back.

You may post pictures showing all the PCB if you like. Maybe we can spot the suspect there.

[quantalume]

(01-04-2016 10:21 PM)Jurgen Keller Wrote:  I found a resistor that looks damaged, see this picture (it's the small one in the middle with a crack). Probably that's not the root cause, just the consequence of another defect. Anyway, it's a starting point.

That's an inductor, 240 uH I believe.

[Paul Berger (Canada)]

(01-04-2016 11:35 PM)quantalume Wrote:  

(01-04-2016 10:21 PM)Jurgen Keller Wrote:  I found a resistor that looks damaged, see this picture (it's the small one in the middle with a crack). Probably that's not the root cause, just the consequence of another defect. Anyway, it's a starting point.

That's an inductor, 240 uH I believe.

...and it is discoloured as well which would suggest it was subjected to current beyond its design. The one right beside it has a small crack as well between the violet and brown bands. Where does the edge connector contact these two pins are connected to go? Is that a connection back to the 9100? According to the interface manual the 15V supply is only rated for 50mA which does not seem like a lot for a reader which I would suspect has a motor to feed the card through, does the reader derive all its power from the 9100 or does it have its own power supply?

Paul.

[quantalume]

So wait, that PCB picture is from the inside of the 9100B or the card reader? Is that the card edge that faces the rear of the 9100B or a card edge from an internal card of the 9160A?

[quantalume]

(01-05-2016 02:45 AM)Paul Berger (Canada) Wrote:  does the reader derive all its power from the 9100 or does it have its own power supply?

From the 9160A manual:

Quote:The 9160A takes its power directly from the Model 9100 Calculator. The only connection required is to insert the 9160A plug into the connector at the rear of the calculator.

Also:

Quote:When using other peripherals, such as the 9125A Plotter, in conjunction with the Card Reader always plug the 9160A signal connector directly into the 9100A Calculator. Insert signal connector from second peripheral device into the back of the 9160A connector.

This would suggest that peripherals are "daisy chained." If that card edge is the rear-facing connector on the 9160A, then I suspect someone has plugged something into it that damaged the card reader.

[Jurgen Keller]

Thank you all for your tips and suggestions!

The picture shows one of the PCBs in the card reader, not in the 9100. And as quantalume noticed, there seem to be two defective resistors (or inductors?).

Furthermore, as Paul already wrote, the 9160A takes its power directly from the Model 9100 calculator, it has no power supply.

I've taken some more pictures so that you have a better idea of what the device is made of. I've also opened the connector so I can try to follow the wires from the connector to the PCBs. Actually, there is not that much electronics inside. There are two PCBs, the motor to pull through the cards, and the lamp to detect the marks on the cards.

PCB1
PCB2
Connector Top
Connector Bottom
Device Pic1
Device Pic2
Device Pic3
Device Pic4
Device Pic5

So I think my next tasks are:

• Follow the wires so we know which connector pin corresponds to which PCB pin.
• Do some measurements. Suggestions and assistance highly appreciated.
• Buy replacements for the broken parts. Suggestions to determine the exact kind of these parts is welcome. Shall I open them to determine if they are resistors or inductors?
  

Please apologize my lack of electronic knowledge. I really appreciate your help, and I think it's really worth to try reviving this device.

[quantalume]

Those are nice images! I always enjoy looking at quality, old boards with their gold-plated traces.

(01-05-2016 09:30 PM)Jurgen Keller Wrote:  So I think my next tasks are:

• Follow the wires so we know which connector pin corresponds to which PCB pin.
• Do some measurements. Suggestions and assistance highly appreciated.
• Buy replacements for the broken parts. Suggestions to determine the exact kind of these parts is welcome. Shall I open them to determine if they are resistors or inductors?
  

Please apologize my lack of electronic knowledge. I really appreciate your help, and I think it's really worth to try reviving this device.

I'm pretty sure that those components are inductors. I believe that position on the card edge (#2 from the right) is where the -15 volts power enters. If you look at the other (presumably not defective) board, you will see another similar inductor, also connected to position 2, the other end of which connects to the negative terminal of an electrolytic capacitor. What we have here is an LC filter to clean up any noise on the supply line (or keep noise from the device from affecting the supply line). The defective board has two inductors which presumably connect to the two adjacent capacitors (you can verify this by following the traces on the back side of the board). My guess is that there are two switching circuits, one of which is the motor driver, which utilize that supply voltage, and they wanted to keep them isolated from each other and the supply.

One of the first things to check would be those electrolytic capacitors. They should initially indicate a low ohm value when you connect the meter, which rises over time (seconds) as they charge. If the reading doesn't reach megohms after a short while, then they are likely bad. In fact, it's good practice to replace electrolytics that old even if they seem good. You should unsolder one end of the capacitors when checking them.

If the capacitors appear good, I would suspect that the motor is drawing too much current. You could apply a DC voltage to the motor from a power supply or battery and measure the current. Let's hope it's not the motor since they are likely impossible to find now, although you could probably rig up something else to mount in its place. I can't tell from the photos; does the rubber wheel spin against another wheel, or does it grip against that metal plate? If there is too much friction, the motor won't spin freely and will draw an excessive amount of current. Unfortunately, because of the gearing, you can't just turn the wheel to see how much friction there is. You could, however, remove the motor and check the current again with the wheel spinning freely.

[quantalume]

I was just reading one of your earlier posts:

(01-03-2016 09:33 PM)Jurgen Keller Wrote:  At least the transport roll was totally rotten. I couldn't even measure the correct diameter. So I replaced it with a gummy ring that might not fit well. However, I've no idea if that can have a bad impact on the correct operation.

Did you originally have the card reader connected before replacing the drive wheel, and was it blowing fuses then? I'm wondering if this new wheel is creating too much friction.

[jebem]

Looking to the posted pictures, it seems fairly easy to do the diagnosis now.

- Test for power supply short circuits again, but without plugin in the two PCBs, then one OCB, tehn the other PCB, to find out where is the short-circuit.

- Also make sure that the burned inductor is not in open circuit.
If it is, temporally use a piece of wire to bypass it in order to allow the current to feed the electronics and the motor.

- Most probable what you see as a "0Kohm" reading would in fact be a lowish value, like 5 or 10 ohms, and that will be the normal internal resistance of the motor.

When the motor is energized, this lowish resistance value cause a large initial rush current, but this is normal and should not blow the fuse, as long as the motor can start rotating to let its impedance to rise fast to the normal value and let the current decrease rapidly.
(fuses can stand larger current than the specified value for a short time, specially if they are slow blow).

Now, if the motor can not rotating because it is mechanically blocked, its impedance will not rise and the current will remain high resulting in the issue that you describe (blown fuses and series resistors/inductors burnt).

[Jurgen Keller]

(01-06-2016 03:07 AM)quantalume Wrote:  I was just reading one of your earlier posts:

(01-03-2016 09:33 PM)Jurgen Keller Wrote:  At least the transport roll was totally rotten. I couldn't even measure the correct diameter. So I replaced it with a gummy ring that might not fit well. However, I've no idea if that can have a bad impact on the correct operation.

Did you originally have the card reader connected before replacing the drive wheel, and was it blowing fuses then? I'm wondering if this new wheel is creating too much friction.

No, I didn't. The drive wheel looked so bad that I thought I try to replace it first, before connecting the OCR. Fore sure it would have been a good idea to try to connect the OCR just with the old wheel removed.

[Jurgen Keller]

(01-06-2016 10:32 AM)jebem Wrote:  - Test for power supply short circuits again, but without plugin in the two PCBs, then one OCB, tehn the other PCB, to find out where is the short-circuit.

The short circuit is on PCB1, the one with the burnt inductors. That's the expected result I guess.

(01-06-2016 10:32 AM)jebem Wrote:  When the motor is energized, this lowish resistance value cause a large initial rush current, but this is normal and should not blow the fuse, as long as the motor can start rotating to let its impedance to rise fast to the normal value and let the current decrease rapidly.
(fuses can stand larger current than the specified value for a short time, specially if they are slow blow).

Now, if the motor can not rotating because it is mechanically blocked, its impedance will not rise and the current will remain high resulting in the issue that you describe (blown fuses and series resistors/inductors burnt).

That makes sense to me. However, I don't have the impression that the wheel is blocked, I can rotate it manually without much force.

Furthermore, I wouldn't expect that the wheel turns right from the beginning, only when a card is inserted. There is kind of a switch above the wheel which lifts the silver metal strip you can see in the middle of "Device Pic5". I think that's the mechanism to start the motor.

I've also verified that my replacement gummy ring lifts the metal strip a bit, maybe enough to start the motor right from the beginning. So I think the diameter of the replacement ring is definitely too big. I should have understood the mechanism better before inserting a replacement wheel. It's easy to be wise after the event.