Help with diagnose a PCB

[cs.dk]

Hi,
I’ve got a control PCB from a Manitou telehandler, which seems to be defect.

When K16 (the relay at the bottom after an amplifier) is switched to radiocontrol, it operates the hydraulic proportional valve fine.
Since it’s a relay board, there isn’t really much that can go wrong. But..
When K16 is switched to onboard control, it fails to operate the valve.
The only “active” between the thumbwheel and the proportional valve is the amplifier.
I can measure ~30Mohm from the thumbwheel input and the black jumper after the amplifier, when no power is applied to the board. Is this normal?

The chip seems to be grinded or similar, it’s very hard to see some markings, even with a macrolens.
However, I think 358 is marked in it. So I guess it’s a LM358 dual opamp. Does it seem right with an opamp in that place?

First: Check voltage at pin 4 and 8?
Second: I’m clueless here. Can I test if the LM358 is operating correct?

Any other clues?

[fcb]

Crudely put: An opamp will alter it's output so that the (+)non-inverting input and the (-)inverting input will be the same.

So you could check that pins 2&3 are the the same voltage and that 5&6 are the same voltage.  This probably won't be the case if they are using the opamp as a comparator..

[cs.dk]

I don't know if it's used as a comparator.. I'll try to trace the connections out.

[TheUnnamedNewbie]

First step in any troubleshooting: "Thou shall measure voltages!"
Verify the supply rails of the opamp are correct. If this is not the case, try to figure out why (is it shorted to ground?)

I'm going to guess from the circuit diagram that it's going to be a buffer circuit. Signal is applied to positive input, output is fed back to negative input. Hence, you should measure a short between the output and negative input, and no connection between the inputs. When power is applied, there should be almost no voltage difference between the two inputs (a few milivolts is possible due to offset in the opamp itself). If this is not the case, remove the opamp from circuit, and verify that a) the pcb is not damanged somewhere else (output shorted to ground? no connection between negative input and output?) and/or the opamp works out-of-circuit.

[EPTech]

Hi there,

Correct me if I am wrong but I do not see the radio control override in the schematic. this leads me to believe it is connected in parallel, somewhere in the cabinet. Also check whether the radio control is really disengaged and is not putting any voltage on pin 16.

Also check whether the thumb wheel is outputting a changing voltage anywhere between the two rails on it's center tap as you turn the wheel. Since there are some series resistors in the thumb wheel supply, this voltage probably will not go down to 0V or up to the supply rail. If the opamp is indeed a buffer as suspected exactly the same voltage should appear on it's output.

These relays do not look like signal relays to me. They are often not sealed. When they are used to switch low currents such as analog signals they have now way of cleaning themselves. In moist conditions their contacts can become less conductive. Never take a relay for granted.

Happy hunting.

[negativ3]

Sorry to be safety sally but the device the PCB is out of determines that it will need to be replaced or repaired by an authorised repair shop.

It's a forklift truck with the ability to fail catastrophically and cause injury/damage.

[EPTech]

Of course it has to be,

But then again, it also needs to be designed in that way.

[cs.dk]

Thanks a lot for the replies. And sorry for the late reply; haven't had access to the machine, before now.

Output A is indeed shorted to inverting input A - I googled for buffer opamps, which gave a lot of reading. Now that i see the configuration, it makes more sense.

I'll measure more tomorrow.

And just a heads up for safety - There is tilt warning sensors, that overrides the thumbwheel.. The problem here is that the thumbwheel doesn't work at all.

[cs.dk]

If i disconnect the thumbwheel from the PCB the is 3V in one direction, 6V as middleposition, and 9V in the other direction.
However, measured at the LM358 input, there is only ~300 mV. It looks like the LM358 is pulling the signal down, is that possible?

[cs.dk]

Sorry for the late feedback, I almost forgot the thread.

After replacing the LM358, the device is working again. Thanks for all inputs

[cs.dk]

So it's dead again Worked fine for at least a month without trouble.

It's a OnSemi LM358DG; Datasheet

It should handle 32V on the powersupply. However, the datasheet shows this graph, which is extremely close to a 12V battery/alternator system.

[fcb]

Can you add a series resistor and shunt zener to the power supply to the LM358 - this will help with load-dump events (which is one possible mechanism that's killing your IC).

[cs.dk]

I changed the "dead" LM358, when it failed - Maybe it was just a cold joint. The machine is still running good
This time, I took much more time and effort to do the rework, and inspected the joints with a magnifier. There was a few pins that i had to give a go 2-3 times to get them good. I expect that my first approach was to quick.

[cs.dk]

They wasn't in stock when i ordered at that time. So i thought another brand would do the same thing.

Anyway it works now. I think my first attempt at repair was to crappy.

[cs.dk]

Sorry for nekro posting..

The LM358 is dead again after ~1½ year of use. This is really annoying.

The problem is, i don't know if it's overload, overvoltage, or what ever may kill it.. Seems strange though. Manitou have 1000's of machines running all around the world. Could the design really be so bad?

[fcb]

Could well be a 'load-dump' event that is spiking the power supply and killing your LM358. Doesn't look like there is much protection on that board.