Dirt Sensor Circuit Misery

[Carl E]

I need BIG help. This is a vacuum dirt sensor circuit made of 2 boards. I've reverse engineered the circuits as best I can. The circuits are floating AC powered, I cannot attach a scope to see the traces as it causes the LEDs to turn on very bright and I even smoked a 220 Ohm resistor when I attached the ground lead to the scope.
As dirt interrupts the IR between D1 and D2,  LEDs  labeled Gled (Green Led) and Rled (Red Led) switch logic states, then switch back after 3 -4 seconds. The circuit does work, except the transition from a lit Rled to a lit Gled is slow taking 2-3 seconds to turn the Gled on to full brightness.. A good circuit changed states in 1/2 sec.
I don't understand Op amps well and cannot figure out where the cap is that controls the transition of the Leds. Í also don't understand what the D6-D9 bridge does.
There are x2 3 pin SOT23 devices controlled by the OpAmp, I suspect one is an NPN, the other is ?? maybe a MOSFET ?? The pinout on Q2 does not match NPN, PNP transistors. I can't ID the caps as they are ceramic SMD devices.
Any thoughts on this circuit (attached) would be appreciated as I'm going bald from scratching my head. Thank you 

[ace1903]

Resolution of the pictures is so low that barely anything can e seen.
Probably you can power circuit with some transformer for test  instead of capacitive dropper.
Transformer will provide isolation so you will be able to use oscilloscope.
Zener diode will give you ideas of what secondary voltage needs to be. 
Please provide pictures with better resolution.

[Carl E]

Thank you ace1903. I will retake or redraw the pictures and re-post them. After the capacitor dropper there is a 78L05 regulator but the voltage I measure is 4.0v not 5.0.
Carl

[Carl E]

Hi, I've redrawn the schematics and attached them. If anyone has any idea about how this circuit works, I certainly will be all ears. As far as I can tell, there is no documentation on the net for this, it is in a sears kenmore power head. There are very few boards out there.
Thanks in advance
Carl

[ace1903]

I don't know why pictures are 400x300pixels only.
I guess chips are LM358 , and one is oscillator that drives IR LED and second one is detector circuit to average filtered signal from receiving led.
That voltage that is 4 volt instead of 5 says that somewhere is short that causes excessive current.


If you have regulated power supply put 5 volts after 78L05 instead of mains power.
Check with scope signal at transmitting LED then goin back  to the output of operational amplifier.
It can be that there aren't oscillation and IR LED is constantly on and consumes lot of current that causes voltage to drop.

I can not see positive feedback around that operational amplifier. Maybe it is error in schematics or I simply don't see it.
If my theory of operation is correct IR LED should be driven with signal in tens of KHz range.
If that is correct next step will be to check receiving circuit. But let first see transmitting part.
If you check with phone camera can you see it IR signal from IR LED?

[Carl E]

The circuit did work, but the brighness of the red and green leds were dim. The change from green to red is instantaneous when the ir beam is broken, but, the change from red turning off to green turning on is slow and almost looks like a dimmer switch is being used. Sometimes it looks like there is an ocillation between the red and green leds being on.
I will try your suggestion of injecting 5v in the circuit, I also considered using a 24 VAC as a source instead of 110VAC, I dont have an isolation transformer.
Which LM358 would be the  oscillator,  pin 1  on the emitter board ?
I will also recheck the schematic to verify I didnt miss anything as far as positive feedback in the circuit.
I suspect that the problem is on the emitter board as I have 2 control boards and they both act the same (which are now red always on..I must have toasted something,)
I dont know why the picture are 400 x 300 pixels....sorry

[ace1903]

Yes pin 1 of LM358.

[Carl E]

I connected 7 VDC to the + and - sides of the bridge rectifier and got 5V out of the 78L05 chip. With this input voltage the circuit does work as it should. I was able to see wave forms on the scope and measure voltages, time  etc. There is square wave generated out of pin 1  on the emitter board but it's only there when the IR beam is broken.
I have 2 control boards, both exhibit the same problem when connected to 110 VAC. The voltage after the bridge is about 3.5V. I need at least 7V to allow the LM 78L05 to work properly. Is it possible that 2 boards from 2 sources have the same problem? The only thing I can think of is to change the x4 1n4007 diodes in the bridge and change the .47 uf 250V cap on the ac line. The 220K and 220 Ohm resistors seem to be OK. The only other component after the bridge and before the 5V regulator is a 47uf cap which I replaced with a 100uf but that made no difference.
If that doesn't work, what value can I change on the AC line to jack up the voltage to 7V after the bridge ?
I wish I was better at AC circuits
Thank you for your help.
Carl

[ace1903]

Positive feedback with IR transmitter/receiver diodes, smart indeed.
Learned something new today.
Bridge with diodes D6/D9 makes no sense to me unless there are placed zener diodes instead of regular diodes.
Chane with increased capacitance can make problems with increased current over 1n4007 diodes.
I would replace all of them with new ones. An open diode can explain the lower voltage.

[Carl E]

 I added a 100 uf cap in parallel with C1, 47uf. This actually brought the 78L05 output DOWN from 4.07 to 3.96V.   I tested the diodes in the bridge, they tested good.  D6/D9 are melfs, there's no way of knowing what they are. I did measure the voltage at D6/D7, the change is 0 to 1.25V when triggered. The voltage at D8/D9 is 1.25 to .5v when triggered.
then,
I tried to increase the .47 250V capacitor by adding another .47 in parallel. The leds were much much brighter, the 78L05 output was 4.99V, the circuit works ( at least it reset) but I don't understand how this affects the rest of the circuit and I dont want to toast anything.  Could the 110v supply side be some kind of tuned circuit?
I removed the extra .47uf and  tried adding a .1uf in parallel with the .47,  this brought the 78L05 output to 4.5V but the circuit never reset, ie, the red leds stayed on. Maybe I didn't wait long enough for a reset after I added only one .1uf. I added 2 more .1uf to the .47, the 78L05 output is 4.98v and the circuit works. I'm still at a loss to understand what would cause the circuit to require a value change from .47uf to 7.7uf on the AC input in order to work correctly
I now have a bald spot on my head. LOL

[cozza]

Sounds like the original 0.47uF 250V dropper cap has lost it's capacitance. Replace it with a new one and see how it goes.

[Carl E]

Changed the .47 uf to .68uf. Circuit works fine, no smoke.
Appreciate the feedback from you all. This is a GREAT forum.
Now its OTTNO.  (on to the next one)
Carl