HP Forums

Full Version: HP 82143A / HP 82162A foto coupler (encoder) - substitute
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2
First of all: Happy New Year 2017 to anybody!

Need some information maybe somebody can help. I didn't find any information - neither here in this great (old and new / archive) forum nor on the web at all.

I discovered that in a HP 82143A thermo printer the foto coupler is defective. In the printer service manual it is referenced as encoder to be replaced in whole.
I found out it is not working (properly) when I changed it with a working one.
The effect is that the print haed is running into the left wall and does not go but to home position till I cut off power (switching off).
Quite sure the same principle is used in the HP-IL printer 82162A.

Is there any substitute I can somehow use - with or without any modifications - for replacement?
Any suggestions appreciate.
Thanx.
Andi
My printers are sitting about 24 miles away at the moment, so I can't check it physically, but I may have a part available salvaged from a printer that was in a flood.

If you can determine if the LED or the phototransistor is the bad component, it may be possible to fit a new one. I don't recall how embedded they are in the part.

Edit: It turns out I hadn't saved that part, just the encoder wheel.
The foto coupler is one component welded together.
Didn't try to open up until now. See picture attached.

If I compare the not working one with the working one with a DMM I can not find many differences. Maybe I did not use the right measurement methods.
My findings:
Working one and not working one both give me only one result when I use the diode tester:
- on black, + on red/white 1524 and 1644 resp.
Using resistance tester:
- on green, + on red/white: if I cover the diodes gives open circuit, if I give direct light from a LED torch gives about 5.4 kOhms one the working, 3.2 kOhms on the not working one.

Andi
According to the service manual, when the encoder pulley reflects the infrared light onto the photo transistor, the green wire (at the connection with R11 and the positive in put to U3C) will have a more positive voltage than on the red wire (at the connection with R10 and the negative input to U3C), as referenced to ground. Page 4-5 has some trouble shooting steps for determining if U3 or U8 needs to be replaced, or even if the encoder pulley needs to be cleaned. U3 is a quad comparator, which should be an LM339.
Thanx Cruff,
I looked into the service manual at the mentioned page. And the electric diagram as well.
Do you think that one of the integrated circuits is bad? U3 would be replacable, ok.
Why do you think the printer works properly when I switch the photo coupler with the other one from the working printer? Vice versa the other printer runs into the left side wall with the supposed to be defective photo coupler.
Am I correct you mixed the red wire (thru C5 to ground) with the black one (negative input to U3C)?
As soon as I have time I will measure the voltages on green, black and U3 output / pin 38 of U8 and will report.
(01-02-2017 05:18 PM)AndiGer Wrote: [ -> ]Do you think that one of the integrated circuits is bad? U3 would be replacable, ok. Why do you think the printer works properly when I switch the photo coupler with the other one from the working printer? Vice versa the other printer runs into the left side wall with the supposed to be defective photo coupler.

I don't know myself, just mentioned the troubleshooting table, but it does seem that from your testing, the issue is with the problematic encoder. If the window of the phototransistor is dirty, or the phototransistor has degraded, it would result in a lower voltage reaching the + input of U3, which would prevent the comparator from triggering. Take a look for debris covering the phototransistor. Alternatively, the output of the IR LED could be degraded, or the LED current could have changed, which would affect the comparator threshold.
Also, if the values of resistors R10 and R11 are not correct, which seems unlikely, this would affect the signal comparision.

(01-02-2017 05:18 PM)AndiGer Wrote: [ -> ]Am I correct you mixed the red wire (thru C5 to ground) with the black one (negative input to U3C)?

No, the red wire supplies +5V (Vp) to both the IR LED and the photodiode. C5 is a bypass capacitor, which is present to help prevent noise from affecting the phototransistor output to U3. If C5 is shorted, which seems unlikely, it would prevent the circuit from functioning properly.
(01-01-2017 01:59 PM)AndiGer Wrote: [ -> ]I discovered that in a HP 82143A thermo printer the foto coupler is defective. In the printer service manual it is referenced as encoder to be replaced in whole.
I found out it is not working (properly) when I changed it with a working one.

(01-02-2017 04:39 AM)AndiGer Wrote: [ -> ]If I compare the not working one with the working one with a DMM I can not find many differences. Maybe I did not use the right measurement methods.
My findings:
Working one and not working one both give me only one result when I use the diode tester:
- on black, + on red/white 1524 and 1644 resp.
Using resistance tester:
- on green, + on red/white: if I cover the diodes gives open circuit, if I give direct light from a LED torch gives about 5.4 kOhms one the working, 3.2 kOhms on the not working one.

(01-02-2017 05:18 PM)AndiGer Wrote: [ -> ]Why do you think the printer works properly when I switch the photo coupler with the other one from the working printer? Vice versa the other printer runs into the left side wall with the supposed to be defective photo coupler.

Hi,

Sorry, this is a long report.

If I understood correctly: the faulty photo-coupler does not work in any of your two printers, while the good photo-coupler does work on both of them.

From all your description, the photo-coupler is defective or the LED/photo-sensor windows are blocked/dirty.
One other possible cause is the encoder wheel reflective surfaces. Is it of the same type on both printers?

I don't have this printer, so I can't give you the correct voltage readings. However, from your description, I could calculate the approximate voltage values based on the HP service guide schematics.
I'm attaching two diagrams that may help.

Assuming the photo-coupler windows are not blocked, a faulty LED is the mostly common cause, although I have seen defective photo-transistors showing anomalies in their dynamic behavior.

In this case, theoretically both photo-sensors should work, as either measured static value (5K4 and 3K2) will set the positive input (pin 11 of U3 comparator) at a minimum of +4.2VDC (5K4) or +4.5VDC (3K2), well above of the estimated +3.5VDC threshold voltage present at the negative input (pin 10 of U3).

Why estimated 3.5VDC at pin 10 of U3?
Because the typical IR LED will have a forward voltage of 1.5V to 1.8V.
So 5Vp - 1.5Vf = 3.5V. This will be the worst case scenario. If the IR LED operates at 1.8V, then the threshold voltage will be just 3.2V.

However, what you really need is a oscilloscope to measure the U3 C comparator input at pin 11 and output at pin 13.
This will show the dynamic behavior of the photo-transistor.

You can test the suspected LED by comparing it with the good working photo-coupler LED in a safe way:
Use a external 5 or 6VDC power supply and apply the voltage to the LED terminals using a 220 ohm current limiting resistor and a current meter in series with the power supply.
Take note of the current reading on the good photo-coupler and compare it with the faulty one. The faulty one will show a reduced current.

Should the IR LED be faulty, there are plenty of 1.5V IR LEDs in the market that can be fitted inside the photo-coupler capsule.
Same as for the photo-transistor: virtually any NPN type will do and there are plenty to choose from.

As mentioned by cruff, Despite all diagnostics done by you are pointing to a faulty photo-coupler, component values in the PCB can also trigger this issue specially when using the 5K4 photo-transistor item (or using the 3K2 one, not sure which one is not working, specially if this is a dynamic behavior issue).
One obvious thing to check here are the two resistors R10 (120 ohm) and R11 (27Kohm).
The U3 section C IC can be faulty by presenting a lower impedance at pin 11. It will work with the 3K2 photo-transistor but not with the 5K4 one.

If all the above does not apply, you can try a simple test: cut one leg of resistor R11 and insert a additional 10K or 22K in series for a total of 37K or 49K at pin 11 of U3 C.
This will allow to use the 5K4 photo-transistor if it is out of tolerance.
First off all: Thanx a lot for explaining and your efforts - especially the diagrams by Jose.

This evening I did some measurements on the defective printer.
Red <--> ground: Battery voltage, about 4.87 VDC.
Green <--> ground: If I rotate the pulley manually voltage will change with each movement, but no regular pattern. Voltage varies between 0.4 to 4,65 VDC. If I cover the coupler's "window" voltage drops to 0 VDC.
When I turn the pulley slowly from home position to left hand wall I have about 20 high voltage of 4.82 VDC at pin 13 of U3. In case of high voltage at pin 13 the voltage at the green wire is between min. 3.65 VDC and up to 4.65 VDC (sometimes goes up to the higher value when slightly turning of the pulley). This is the 3k2 one.
Black <--> ground is 3.65 VDC constantly.
Measurement of the 2 resistors give required values: R10 = 120.2, R11 = 26.7 k.
Pin 3 of U3 has 14.6 VDC.

I will check the count of high voltages at pin 13 of U3 in the working printer and report. Maybe this will help.

Andi
Just tested the working coupler, green wire vs. ground.
It gives as supposed a voltage of 4.65 every light turn of the pulley dropping to less than 1 volt in between.
So I will test the LED as described. Will report in a few days.
Tested LED current. Both have the same value.
So can we conclude the defective component is the photo diode/transistor?
As mentioned and visible in the picture: LED and receiving component sitting inside a welded case. Should I try to crack it open? Anybody did that before?
(01-07-2017 12:40 PM)AndiGer Wrote: [ -> ]So can we conclude the defective component is the photo diode/transistor?

That seems probable.
What is printed to the side of the part? It does look like a motorola or murray? logo IIRC. Just wondering if there is equivalent part that could be used as a drop in replacement... In theory Vishay IrDA module trans/recv. module could be used as a replacement, but that wouldn't be pretty.
Can tell you the printing on the coupler when I have access to the printer again.
Or is the picture in post 3 sufficient?
just a wild guess, this pony would probably do the trick:
Reflective Opto Switch
Dear Hans, looks very close to the original one. Starts with OPB as well. There are similar one's, OPB70.. I have to check the detective distance, 0.15 or 0.2 or 0.25 inch.
Maybe someone who is more familiar with transistors than me can tell the right one according to voltages. 25V or 30V collector emitter voltage? Or doesn't matter? Or need a Darlington version? For example:
http://pdf1.alldatasheet.com/datasheet-p...asheet.pdf
Thank the electronic experts for further help
Andi
(01-12-2017 10:26 AM)AndiGer Wrote: [ -> ]Maybe someone who is more familiar with transistors than me can tell the right one according to voltages. 25V or 30V collector emitter voltage? Or doesn't matter? Or need a Darlington version?

That 25V or 30V VCE is the maximum the phototransistor will tolerate. With tensions of around 5V floating around in there you're well within tolerance.

Without knowing how the printer is built I would guess that the output from the phototransistor is pulsed (alternating reflective and non-reflective segments passing in front of the opto-coupler) and put into a cleaner square wave form by the LM339, which is essentially doing the job of a Darlington couple (so no need for that type of phototransistor specifically). The software counts the pulses and thus knows the horizontal position of the print head.

There's a good chance that the component linked to by Hans will work as a drop-in replacement as long as the transistor is the same polarity as the original's.

G. Stewart
Thank you very much for explaining. So I will try to get a substitute.
The linked to transistor is NPN. That's the version needed as posted by jebem.
Thus only the right reflecting distance is important.
After measuring and building in I will report.
Hope it will work and with common efforts will also help others
Andi
out of curiosity, i dug through the hpmuseum USB-stick and checked the HP82143 printer service manual.
according to the manual, this so-called encoder is indeed an IR-LED / NPN combo.
part number printed on the component is OPB2607, but seems to be unobtanium now.
i tried some quick & dirty testing with one of mine (fully functional), using a handheld DMM and found this:
1) checking the IR-diode:
put DMM to diodes test and connect DMM plus to red wire (it's actually 2 wires, red and white). connect DMM minus to black wire. DMM reads approx. 1 V.
2) checking the receiver NPN transistor:
put DMM to OHMS mode (seriously) and again connect DMM plus to red wire. connect DMM minus to green wire. DMM should read "overload" or "infinity" or something that indicates an open circuit. now, take your TV IR remote and hold its transmitter LED directly against the encoder's window. press and hold down a button on the remote. DMM should now give irregular readings in the kOHMS range.

note that DMM readings may vary with meter brand (i.e., OHMS sourcing current). so, ymmv, but you get the picture.
(01-17-2017 08:53 PM)Hans Brueggemann Wrote: [ -> ]put DMM to diodes test and connect DMM plus to red wire (it's actually 2 wires, red and white). connect DMM minus to black wire. DMM reads approx. 1 V.
That's consistent with the junction threshold of an LED being around the 1.2V mark.
After some searching I finally ordered a Honeywell HOA1405 photo coupler.
For the printing picture itself it works well. All dots there, nice spacing. Printing numbers everything fine.
BUT: When I print alpha the printing is done completely but the print head stays in the left side. When I switch the printer off and on again the print head runs back to home position. Whatever I print after that (number or alpha) the printer performs a line feed before printing the next requested output.
If it is alpha again the print head behaves as described before. If it is a number the print head works as desired and comes back to home position.

Any idea solving that?
Pages: 1 2
Reference URL's