Simple circuit with Photointerrupter has got me baffled...

bertus

Joined Apr 5, 2008
22,276
Hello,

I know that it will be very difficult to get hold of the schematic of laboratorium equipment (what this autosampler is).

Bertus
 

GetDeviceInfo

Joined Jun 7, 2009
2,196
Can you determine the signal line. What would the point of measurement be? And with everything connected?
point C is your signal line for that switch. Going on the measurements you've indicated, your source is being pulled low somewhere, or not being sourced.

I would trace that line back to the motherboard connector, looking for a pull up resistor along the way (or another device thats 'or'ing the line low). If the line is a single run, disconnect your connector at the motherboard and then;
measure for your +5vdc at the motherboard, and, check your resistance from the appropriate cable trace, back to point C (looking for continuity), and again to common (looking for a short).
 

eblc1388

Joined Nov 28, 2008
1,542
I agree with the advice by GetDeviceInfo but I think the problem is something(faulty) upstream all the way to the motherboard pulling the signal line HIGH instead of low.

Therefore although the Z sensor operates correctly, its voltage can't drop down to about one volt(only drops from 5 to 4.2V) so the motherboard doesn't consider it a valid signal.

Is there yet another board situated between the sensor board and the motherboard?

If it exists, have you checked that too?

You'll need to trace the wire at point "C" (or the Z-axis sensor line) all the way back to the motherboard and try to see what component it connects to.

Make a sketch of the route and all the components that signal line "C" connects to and then maybe we can isolate the possible problem or at least give some advice.

 

Attachments

Last edited:

Thread Starter

pureanalog

Joined Oct 8, 2010
36
I will give you a full sketch and hi-res photos of the situation here, and will get back soon. There is a similar board to the one we discussed right before the motherboard with another 2 opto-sensors. Seems approximately the same in configuration.

I also noticed that the resistors measure double their nominal resistance when the power is on. 330 ohms when power is off and 660 when power is on. Are these the pull up resistors?

Anyhow, I will get back to you soon.
 

GetDeviceInfo

Joined Jun 7, 2009
2,196
From what I've been told;

With everything connected at the indicated point C, the voltage swing is between 0 and 0.7 Volts for blocked and unblocked respectively. For all other points it is 0 and 5 Volts.
I'm going with a faulty source or signal pulled low somewhere. However, there seems to be much contradiction, so I'll just watch and see if the information stabalizes.
 

GetDeviceInfo

Joined Jun 7, 2009
2,196
Inverse convention, who would have guessed.

If the OP would remeasure at point C, using common as a reference and repost his findings, I'd be glad to help.
 

eblc1388

Joined Nov 28, 2008
1,542
You don't need to guess.

The reason why positive is used as a voltage reference is due to the fact that only +5V and GND is powering the board, being so to isolate the location of the fault.

With this arrangement, all the photo-interrupters output becomes open collector and voltmeter reference to 0V could not show any voltage changes to confirm the operation of the interrupter.

It is not a convention, just a method to find out if the interrupter is working fine or not.
 

Thread Starter

pureanalog

Joined Oct 8, 2010
36
I was basing my observation on this, which also applies to the measurement result you have quoted in post#45. The voltage is measured w.r.t. +5V.

I am sorry. I am not professional and apparently not as systematic as I should. It seems that I caused confusion. To sum up, here are the proper and correct results as measured again now:


This measurement quoted above was conducted with everything connected between +5V and each one of the A,B,C, and D indicated points.

But now upon a new measurement, I find that now point D measurement gives 0 Volts regardless if the sensor is blocked or not.

With only the 5V and GND connected, all was working properly. All 4 points are giving out 5V and 0V unblocked and blocked respectively.



and here is a new observation:

WHEN ALL IS CONNECTED:

I noticed that point +5V and point D are short circuited. No resistance whatsoever.

Between point 5V and A, B, C I measure around 1000ohms resistance.


If anything is unclear, please let me know to clear it up!! Or if you feel like I should re-do any measurements just let me know.



getdeviceinfo:

You need a measurement between the common (GND) and each of the A,B,C and D points? With all connected?
 
Last edited:

eblc1388

Joined Nov 28, 2008
1,542
WHEN ALL IS CONNECTED:

I noticed that point +5V and point D are short circuited. No resistance whatsoever.

Between point 5V and A, B, C I measure around 1000ohms resistance.
Never measure resistance when power is ON, the results you are getting will be meaningless.

So, when you said +5V and point D are short circuited in above, is the power of the apparatus ON or OFF?
 

Thread Starter

pureanalog

Joined Oct 8, 2010
36
Never measure resistance when power is ON, the results you are getting will be meaningless.

So, when you said +5V and point D are short circuited in above, is the power of the apparatus ON or OFF?
With all cables connected and apparatus off, point D and +5 are short circuited.

Point +5V and GND are also short circuited when power off.
 

eblc1388

Joined Nov 28, 2008
1,542
With all cables connected and apparatus off, point D and +5 are short circuited.
Reverse the test leads(i.e. red and black) while measuring the same location to see if the reading changes.

Second rule of resistance measurement of a component still soldered in the circuit is that the value measured is *largely* affected by the other components in the circuit. You may not get a true picture.

Point +5V and GND are also short circuited when power off.
As above.
 

Thread Starter

pureanalog

Joined Oct 8, 2010
36
Reverse the test leads(i.e. red and black) while measuring the same location to see if the reading changes.

Second rule of resistance measurement of a component still soldered in the circuit is that the value measured is *largely* affected by the other components in the circuit. You may not get a true picture.



As above.

Alright, I reversed the test leads and the result remains the same.
 
Top