Hello,

I am a farmer not an electrical engineer, but I have recently been trying to learn more about the electronics we use in farming. I have been trying to troubleshoot the grain loss system on our combine that hasn't worked in years. Keeping in mind that I am a farmer, as I understand it the sensors are just a piezo sensor glued to a piece of plastic over a piece of foam that isolates the sensor from the vibration of the machine (first picture). Any grain striking the plastic will send a signal to a monitor in the cab telling me that the machine is overloaded and dumping grain onto the ground with the chaff. There are four sensors on the machine @ $800 a piece. Could we fix these sensors for around ten bucks?

The resistors seem to have tested as good. The pair on the bottom left (second picture) @ 46k ohm, top center @ 22k ohm, and pair on bottom right @ 1k ohm.

The number on the transistor reads 2N 4401 422. I found a clip on YouTube on how to test small transistors using diode mode, and it showed a reading from base to emitter and from base to collector. I'm assuming it is good, but as I said, I am a farmer.

Now the problem, I could not get any reading from the piezo or capacitor. As I understand (again from YouTube), a good piezo should flash a voltage reading when tapped. I could not get any reading. And a good capacitor should build a charge (?) in diode mode or build resistance when ohmed. I could not get either reading. The number on the capacitor is 473K 100N.

Now my questions:

1) The piezo and capacitor are wired directly in series and both seem to be bad. Would one going bad cook the other? This sensor is original off a 1998 machine, so it is old.

2) The capacitor number is 473K 100N. I read that the top number on a small capacitor is usually the capacitance in Farad and the lower number is the maximum voltage it can handle. So what's with the N? Is it in Newtons? The closest I could find on Amazon is a 473K 250V metallized film capacitor; would that work? I forgot to mention that the power to the sensor is 8V.

3) Putting quality of manufacture aside, is a piezo a piezo a piezo? I couldn't find any product that suggested a difference in the voltage put off by a sensor when struck, but none look like the one I have on the sensor. Would any generic piezo do?

This machine is old and the components are getting harder to find. Many sensors are getting ridiculously expensive. Do you think this is possible to fix? Maybe I'll have to bite the bullet and drop the $3200? Thank you all in advance for your help and insight. Thank you. Thank you.

 

A Piezo-Element produces a tiny, very high-frequency, Voltage-Spike when something hits it.
It will also operate as a crude Microphone.

Placing a Piezo-Element in an environment where it will be subjected to
continuous bombardment by fast moving particles is a guaranteed way to destroy it.

Your Circuit-Board appears to have been subjected to a moist and dirty environment,
it's amazing that it's still reasonably intact.
The Circuit-Board is simple enough to be "reverse-engineered",
and a duplicate made, but the hours required to do this will come with a price-tag,
which could easily be over ~20-times the cost of the actual new components.

Piezo-Elements,
which could probably directly replace the existing ones,
are readily available, and cheap,
but will probably have a short lifetime in a harsh environment.

Since Piezo-Elements are cheap,
it might be worthwhile to simply try replacing the existing ones with new,
but there are no guarantees that the "system" will perform as expected afterwards.

In my opinion, this is a stupid design,
and it has very questionable usefulness/accuracy,
but on the other hand,
maybe it's the only solution the Engineers could come up with.

Sorry that I can't be more helpful.
.
.
.

 

Here in Australia these grain loss systems have been around for many years, they work as far as i know with a filter that differentiates between the sound of grain & straw landing on the sensor. The unit on my family's farm harvester lasted the life of the harvester. The only problem i recall was the od cable connection fail due to vibration.

 

The voltage generated by a piezo device that size would be in micro volts when tapped. Tor testing i would use an oscilloscope to test its output. I doubt that capacitor is faulty as they are pretty reliable & it wont test using ohms scale unless it is short circuited internally. The ohms scale test on a capacitor is really only relevant on electrolytic capacitors.

 

Welcome to AAC!

What you have described appears to be simple enough for most of the experienced members here on AAC.
We could get this working again for under $10.
Don't despair. We just need to spend some time and troubleshoot the system that you have.

Can you post a photo of the piezo sensor?
Also can you draw up a quick circuit on how things are normally wired together?

Edit: Sorry, I did not realize that that tiny whitish circle is the sensor. What is the diameter of the sensor?

 

Welcome to AAC.

What does the output of that board connect to?
How did you conclude it was the sensor and not what it connects to that is faulty?

 

Wow, thank you for your responses! Reddit ignored me.

LowQCab: I think part of the reason the board may look dirty is because it was totally encased in rubber that I carefully removed. In the background of the first picture you can see the yellow plastic; the piezo was sandwiched between that and the black foam. So the components weren't just hanging out in the open. But you are right, this design has been around for a long time. It isn't used to quantify loss so much as let the operator know they're over the machine's ability to separate.

debe and Ya'akov: Right on. The wiring harnesses on the shoe and separator return pan were a mess, so that's the first thing I addressed. I talked to a mechanic that knows these machines at our dealership, and he said 99% of the time problems are those rear wiring harnesses or a faulty sensor. We have already replaced two sensors since owning the combine, and over the years they have quit working one by one as opposed to all at once. I have 8V and a good ground to each sensor.

debe: Entry level oscilloscope is on the shopping list lol

MrChips: Diameter of the piezo is 0.5in/13mm. Now I'm going to look like an idiot because I can't remember which outside terminal was which. The center is 'sensor common' which is the ground. Would it be a good guess that signal out is the left terminal on picture 2 and the right terminal on picture 3 since the capacitor would be following the sensor? Which would leave the right terminal on picture 2 and left on picture 3 as 8V power. If needed I could take another apart. Will post a couple wiring diagram pics in a sec.

 

I think the sensor is not polarity sensitive.
What we are trying to figure out is how everything is wired. By doing so we will be able to reverse engineer the circuit.
It is not a complex circuit and we will be able to test the circuit design on a breadboard on the workbench.

 

Also I forgot Ya'akov: The output hooks up to a module in the cab (first picture) that has a selector switch (shoe sensors, rotor sensors, or both and a test fuction) and a potentiometer that adjusts sensitivity (how much loss you're willing to accept). From there LEDs light up on the dash telling you how intensely (according to your settings) you are losing and from what area of the machine.

 

That scope is going to be very helpful. You should also get a bench power supply so you can provide things you are testing with the proper voltages.

The first thing I would so is connect the sensor board to 8V and see if there’s any signal. This would be the first step in signal tracing which is a standard practice for troubleshooting. You follow a signal from one end to the other checking levels and sometimes waveforms to see where it stops looking the way it should.

In this case you have strong reasons to suspect the sensors, but no proof. Unfortunately, we don’t know what the sensor signal should look like, but at least we can make a guess. That transistor is acting as either an amplifier or a switch. It is going to be taking the very small signal of the piezo and making it some portion of the 8V.

It’s got to make it to the cab through relatively small wire, so it should be quite large relative to the scope’s sensitivity. You will probably have to start with the scope on roll and see if you can provoke a reaction on the trace, then adjust the timebase and do a single trigger capture. (Details for when you get it...)

It’s great to have the wiring diagram, I wish whoever drafted it was someone who needed to use it. That section with the selector and sensitivity control is laid out horribly. Better than not having it though.

This really should be just a matter of systematic work. Also, you could possibly get yourself a decent side business making and selling sensor replacements for $400! This is a joke, sort of. If you wanted to do it, you could contribute to the right-to-repair movement by laying out a new board, specifying a BoM (Bill of Materials) and open sourcing it.

The parts would be a few dollars at most, the PCB would be something like $5 for 10, plus shipping for much better, modern versions. You could spend a bit more for some moisture resistant PCB material. And, you could add a connector for the sensor to make replacing it much easier since it seems to fail regularly.

I don’t know how much life is left in your combine, but given the way this is constructed it seems like you could “upgrade” other electronic systems as well if they are prone to failure. I wonder if there are any right-to-repair advocates in your area. They might be of great help if there are .

 

MrChips: Diameter of the piezo is 0.5in/13mm. Now I'm going to look like an idiot because I can't remember which outside terminal was which. The center is 'sensor common' which is the ground. Would it be a good guess that signal out is the left terminal on picture 2 and the right terminal on picture 3 since the capacitor would be following the sensor? Which would leave the right terminal on picture 2 and left on picture 3 as 8V power. If needed I could take another apart. Will post a couple wiring diagram pics in a sec.

If you trace out the connections on that PCB and draw a schematic it will be easy to tell what is going on. It’s a very simple circuit.

 

Here's what I came up with. If it doesn't make sense (I'm having to learn all this as I go) I can try to just trace it out exactly as it is on the board; it would just be slightly more convoluted.

Ya'akov: I'm a big fan of the idea of right to repair. That's why I'm trying to learn how to troubleshoot these problems. Seems a shame to spend hundreds on a new component and throw something in a junkyard that may just have a bad trace, blown resistor, etc. If something isn't working there has to be a reason that can be remedied if I can learn to find it. I have another $400 sensor off a sprayer in the shed that would normally be thrown out. I still have a long way to go before I start making design improvements lol. A Massey Ferguson 8780 also isn't an especially common combine which might be part of why the parts are getting expensive.

 

Good job in the schematic!

I think you have the transistor upside down, recheck the pinout and connections. I think the 2n4401 is American style pinout which is EBC looking at the flat side with the leads pointing down.

The circuit is a simple high gain AC voltage amplifier with likely gain in the 100 to 200 region.

 

Here's what I came up with. If it doesn't make sense (I'm having to learn all this as I go) I can try to just trace it out exactly as it is on the board; it would just be slightly more convoluted.

Ya'akov: I'm a big fan of the idea of right to repair. That's why I'm trying to learn how to troubleshoot these problems. Seems a shame to spend hundreds on a new component and throw something in a junkyard that may just have a bad trace, blown resistor, etc. If something isn't working there has to be a reason that can be remedied if I can learn to find it. I have another $400 sensor off a sprayer in the shed that would normally be thrown out. I still have a long way to go before I start making design improvements lol. A Massey Ferguson 8780 also isn't an especially common combine which might be part of why the parts are getting expensive.
View attachment 298264

Transistor leads are reversed C, E .
Everything else is right..

 

Checking a piezo element is not so complicated if you happen to have a high impedance headphone and good hearing. The headphones of an old crystal radio would be a perfect choice, even a crystal earplug phone would work. Connect the phone across the piezo sensor and tap the panel with a key or similar hard object. If you hear a sound in the phone then the piezo is working. Like others have stated, the wiring is the critical part because the signal is small. And shielded cable is easy to damage.

Several of my ancestors were farmers, some of them also brilliant designers of things to make farming more productive and less hard work.
The development of a grain loss sensor is brilliant, but it post-dates my last experiences with a threshing machine. (1964) WORKING ON MY UNCLES FARM.