lm324 oscillations

[kinarfi]

I've been working on a circuit to control a 12vdc motor in both direction have had some problems with the circuit breaking into oscillations, I changed the design around and on my latest, a ramp signal was oscillating and it got worse with a capacitor and I finally killed it by adding a load resistor, 1K. Is this common?
I used CLD to charge the cap on a PUT oscillator, but I couldn't load the ramp out put at the cap, so I ran it through one of the op amps so I could use the signal.
This is the first time I've posted here and first time for using .png, I like it. I'm also looking for advice on 12vdc motors, my project is using a power wheels 19 tooth motor and the gear box with a 15 to 25 tooth chain drive out put, (lots of machining) Is there a better motor with more power out there, the motor used with FIRST, sorry I don't know what that is, looks interesting. What I want to do is string out the gear train and have about the same power as the power wheel set up described. Also, is there a good source of gears similar to those in the power wheels gear box?
I've seen some real wisdom here and hope I can partake of some of it and maybe even share some of mine.
Thanks all,
Kinarfi

 

[mneary]

Maybe you need more capacitance on the +10Vreg, especially near the LM324 power pins.

I don't see a capacitor on Vin for the LM317, could it be making noise that the LM324 is just echoing? And, unless your load current is really low, 12V-10V (2V) is really pushing the Vin-Vout limits of the LM317.

 

[kinarfi]

Maybe you need more capacitance on the +10Vreg, especially near the LM324 power pins.

I don't see a capacitor on Vin for the LM317, could it be making noise that the LM324 is just echoing? And, unless your load current is really low, 12V-10V (2V) is really pushing the Vin-Vout limits of the LM317.

I'm using this on a UTV off road vehicle, so Vin is about 14.4, and the oscillation did go away with the 1K load resistor, the load without the 1 K is just the 2 op-amps with 10K input resistors.
Thanks,
Kinarfi

 

[kinarfi]

Another weird thing

Another weird thing that I noticed was that as I added the load, the oscillation of the green trace would back down the ramp as I turned the pot to a lower resistance, at 1 K, it was gone. measured with DVM.
kinarfi

 

[mneary]

Since the wires to the battery are probably long, the capacitor on Vbatt is very important, and capacitors on the LM324 power are important too.

Loading the LM324 output is not a robust fix. It'll disappoint you at the worst possible moment. Trust me.

 

[MikeMl]

Also I wouldn't call that 1K resistor a "load" resistor because the only current that flows through it is the input bias current of the LM324. A load resistor would be connected from the output pin of the LM324 to either ground or +10V.

I seem to remember (audiguru will chime in on this) that an LM324 is a particularly crummy opamp for audio because it has a lot of crossover distortion in it's totem-pole output stage. A common fix for this is to connect a "pull-up" resistor from output pin to the positive rail, effectively turning the opamp into a class A amplifier such that it is always sinking current at its output pin, meaning that the top half of the totem pole never turns on, making it much more linear (high fidelity in audio terms). Try a 10K resistor as a pull-up.

 

[Hero999]

Either I'm misreading the schematic or you've got two op-amps in the same feedback loop, which will make it oscillate.

What's happening is you've got two 90° phase shifts with a total gain greater than one at the frequency of oscillation. Negative feedback, now becomes positive feedback which causes oscillation.

 

[kinarfi]

Thanks Hero999

Thanks Hero999 Here's the whole schematic and it's oscillating bad now that I have relays connected.
Here's the whole story, I have a Joyner Trooper UTV and someone suggested power steering for it and this is the design I have come up with after a few other tries. Started out with a SPDT switch (home made) to activate relays to power the drive motor in the correct direction and short the motor out when not turning to create drag on the steering so it wouldn't bump steer. We were having a tough time figuring out how to sense the turning of the steering wheel and are ending up using a strain gauge on the steering shaft to sense torque on the steering wheel, that is 1/2 of a wheatstone bridge and the difference is amplified by OP3 so pin 14 moves up or down from 5.000 volt dc. The Programmable Unijuntion with constant current gives me a flat sloped saw tooth wave form that I can adjust all over the place for voltage and freq, but the output needs to be buffered by OP4 and that is where I got the oscillation until I added the 1K resistor, in red circle. OP1 & 2 add or subtract to that signal so peak is at 5.000 vdc, OP2, or the bottom is at 5.000 vdc, OP1. OP 5 & 6 control power relays are set to provide a dead band of .3 vdc above and below 5.000 vdc as the out put for OP3 moves up and down. OP 7 & 8 control PWM, as the out put for OP3 moves up and down, and is fed to the power relays. The read switch senses when current spikes as the motor stalls at it's end of travel and pulls in C4 and C4 is held in by it's N O contact until the steering wheel is relaxed allowing pin 14 of OP3 to return to 5.000 vdc and turn off the FETs. C1 opens the motor circuit in the event of loss of power, removing the drag created by the shorted motor.
I haven't had this out and tested it yet because of some surgery but it seems to work in the garage. If you have any advice, it would be appreciated.

 

[kinarfi]

More info, I put this thing all together and looked at the wave forms and everything looked real good, then I connected the relays up and they started chattering, I believe this because of the unwanted oscillations. If someone has more knowledge of op-amps and can help, I'd sure appreciate it.
some more info, scope pictures and other stuff can be seen at http://good-times.webshots.com/photo/2679056260104282158JiVyGo
**broken link removed**