I am currently in a summer internship at a company after my first year of university. After only one year of intro level engineering classes, I do not have much experience or knowledge and blame my ignorance on that. I am currently working on updating the design and implementation of a DC Motor Armature tester that was designed nearly 50 years ago. I have a working circuit, which is (for now) powered by 2 24V power supplies, and eventually will be adapted to run off of drill batteries. The signal from the Transmitter circuit is sent through a 2-arm wand which is placed on the commutator of the tested motor and induces an AC signal into the windings. This signal is then picked up by a 2-prong receiver which measures the received voltage across two bars and reads it. If the voltage is over a certain level, there is a bar-to-bar open, if in a normal level, the bars are good, if minimal voltage is picked up, there is a bar-to-bar short. The Receiver is a separate circuit, but the op amp blows without the receiver connected, so I omitted that.

Now for the issue, my circuit seems to work fine for the most part, but occasionally, the Op Amp can blow, and I read the signal on my oscilloscope as -24VDC. I am thinking that it could be due to too much current flowing through the op amp, or possibly back emf from disconnecting the load. I am not sure if the back emf idea is plausible here, as it is an AC signal that is sent through the wand. I am wondering what the reason for the Op-amp blowing is, and any other critiques or suggestions are greatly welcomed and appreciated.

I will attach pictures of my transmitter circuit as well as the previous circuit design.
I am using an OPA445AP op amp, and TIP41C and TIP42C transistors. The "resistors" labeled Pot1 and Pot2 are two sides of the gang potentiometer I have included to adjust frequency. (Also have never posted on a forum before so excuse my mistakes)
Thanks.

 

Welcome to AAC.

It looks like you've drawn this in LTSpice. If so, can you please post your .ASC file.

Do you have an idea of the inductance/resistance of the winding under test?

I read the signal on my oscilloscope as -24VDC

Where are you reading this. At the wand (SIG) I assume?

 

Thank you for the reply,
Yes sir, I am reading that at the wand, which is when I know the op amp has blown. Replacing the op amp gives me the regular signal I am looking for (a roughly +-20V sine wave)
I am not sure of the inductance under test, as I do not have access to an LCR meter. the resistance between the wand probes is in the realm of about 1 ohm. I have a smaller DC armature (Commutator diameter ~ 3.25 inches) set up on my test station, but in reality, it would be used to measure larger motors (Commutator diameters up to a foot) I will attach a picture of one of the larger motors that I have tested on previously.
One more note, the Oscillator doesn't simulate correctly on LTSpice due to it using noise to start oscillations but can replicate the oscillator signal with a 2kHz 20V pk signal.

The wand can be used as a reference for size.

 

the Oscillator doesn't simulate correctly on LTSpice due to it using noise to start oscillations

If you use the UIC transient option, the oscillator should work:

 

What you are describing sounds like the typical 'Growler' This is a simple device for testing armatures , usually no need for complicated electronics!
The growler has a metal Vee-shaped trough that it sits in and a magnetic field is set up an induced voltage between adjacent comm bars.
Then the induced voltage is measured between the segments for consistent voltage, a shorted coil will show as very low voltage.

 

What you are describing sounds like the typical 'Growler' This is a simple device for testing armatures , usually no need for complicated electronics!
The growler has a metal Vee-shaped trough that it sits in and a magnetic field is set up an induced voltage between adjacent comm bars.
Then the induced voltage is measured between the segments for consistent voltage, a shorted coil will show as very low voltage.

The challenge with testing very large motors is that rotating them is a very big hassle. For a growler, you usually sit the armature on top of the device and spin it freely, but there are some growlers that are made to be handheld, and you place on the armature. However, even with a handheld growler, they need to be placed on the windings, and not the commutator. This means, at least once, you would have to turn the motor to test all turns. The advantage of my design is that it is completely done on the commutator. The wand and probe are both placed on bars on the commutator and are tested thusly.

 

usually no need for complicated electronics!

Also, the circuit is very simple. Just an oscillator tied to a push pull amp to amplify current.

 

@crutschow Setting UIC on seemed to make it worse and throws error messages! What did seem to work was putting some series resistance in the supply (8m is a well charged SLA)



Zooming in...



I can't see any obvious reason why the opamp fails, but try putting some catch diodes from drv to + and - rails to catch anything coming back... like this, or at the junction of the output transistor emitter resistors.

 

try putting some catch diodes from drv to + and - rails to catch anything coming back... like this, or at the junction of the output transistor emitter resistors.

I tried adding diodes from drv to supply rails, but was getting a weird reading on my o-scope. However, adding them to the output, after the resistors, doesn't change the waveform and gives me the impression that there is some added protection there. Thank you for the reply, and I will give updates if anything changes!

 

I do not think that I have seen the amplifier circuit schematic, and the pictures of the assembly I really do not find useful. And with the transmitter circuit, I see one output, but to deliver any current will require two connections. AND, I am guessing that it is the transmitter op-amp that is getting damaged.
Adding power supply bypass capacitors to the supply common from both the V- and the V+ IC pins will be a good choice.
Also, with the circuit built on a open breadboard there is a real chance of momentary opens happening,
Likewise with any amplifier intended to sense the voltage across the segments.
NOW A QUESTION: are the power supplies for the transmitter and the amplifier isolated from each other?? Or not??

 

I do not think that I have seen the amplifier circuit schematic, and the pictures of the assembly I really do not find useful. And with the transmitter circuit, I see one output, but to deliver any current will require two connections. AND, I am guessing that it is the transmitter op-amp that is getting damaged.
Adding power supply bypass capacitors to the supply common from both the V- and the V+ IC pins will be a good choice.
Also, with the circuit built on a open breadboard there is a real chance of momentary opens happening,
Likewise with any amplifier intended to sense the voltage across the segments.
NOW A QUESTION: are the power supplies for the transmitter and the amplifier isolated from each other?? Or not??

The transmitter circuit is composed of an op amp oscillator and a push-pull amp. The output is taken from the SIG line and ground. I am using two separate 24V supplies for the rails, connected the + of one to the - of another for my com, so it is two supplies in series with the com between them (this is shown in the schematic). For the bypass caps, you suggest connecting a cap from com to + and com to - correct? What size? Why would this help?

 

For the bypass caps, you suggest connecting a cap from com to + and com to - correct? What size? Why would this help?

Its not a bad suggestion, as batteries are not great when dealing with pulsed currents, so having some additional short term energy storage might help. Try around 200uF across each rail, connected close to the collectors of the output transistors. I would seriously consider rebuilding this on a piece of stripboard/perboard as those breadboards are not good with high/pulsed currents and momentary open-circuits could be having a negative effect. Use a DIP-8 socket for the opamp. Also add a 100nF capacitor from opamp V+ and V- pins to ground as close to the opamp as possible.

 

Power source bypass capacitors, adequately connected, reduce the effective source impedance of the supply, which usually improves the circuit functioning, and always provides a more stable supply voltage.
I have not looked at the op-amp's data sheet to see what the maximum voltage is, but with the scope traces showing 20 volts peak both positive and negative, that may be operating close to the limits.

One other thing is that with the plug-in prototype assembly shown, the resistance of every connection is subject to change from any motion caused by any vibration. So that could easily be the source of occasional failures. Those prototype boards are notorious for that.

 

Power source bypass capacitors, adequately connected, reduce the effective source impedance of the supply, which usually improves the circuit functioning, and always provides a more stable supply voltage.
I have not looked at the op-amp's data sheet to see what the maximum voltage is, but with the scope traces showing 20 volts peak both positive and negative, that may be operating close to the limits.

One other thing is that with the plug-in prototype assembly shown, the resistance of every connection is subject to change from any motion caused by any vibration. So that could easily be the source of occasional failures. Those prototype boards are notorious for that.

I believe I will be adding bypass caps, thank you. The OPA445 is rated for max +-50V.

 

OK, then it is not the problem I had guessed. There does exist the possibility of an intermittent connection in the proto-board assembly.