isolated motor feedback circuit

[ante1]

Audioguru,

With only one digital input I don

 

[audioguru2]

Ante,
I revised your circuit only because Bob needs to detect a motor current of 1A and more. Your circuit's LED current stays zero until the motor current exceeds about 2.8A.

What I meant to say is that the motor does require a minimum amount of drive in order to move with it's load and that works out to about 1A or more of current flow.

Regarding the opamp type, anyone using an old 741 in our circuits here will be very disappointed.

 

[bob_s]

Wow! Thanks you guys for all the help your giving me.

I have a question in regards to frequency. I am driving the fet at 50khz and somewhere between 25 and 100% duty. I am using a PS2501 optocoupler. What kind of delay can I expect from input pulse to opto output?

Bob

 

[audioguru2]

Bob,
The motor, opamp and opto delays are:
1) The motor is an inductor which will have an unknown delay.
2) The opamp has delays of 10 microsec. for rise, and 10 microsec. for fall.
3) The PS2501 opto has delays of 3 microsec. for rise and 5 to 20 microsec. for fall.
Since these delays exceed 1 cycle of 50KHz, the output will not be pulses, but will resemble an average voltage that has a lot of ripple.

 

[bob_s]

Ok. Then the output is not a digital signal but rather an averaged DC output. This is not exactly what I was looking for. I only have a digital input available. Could I expect a cleaner signal if the drive signal was 20khz (50us)?

Bob

 

[audioguru2]

Bob,
20KHz will be much better and we could choose a faster opamp. Also a logic opto will be faster than the PS2501 linear one.
How fast does the current rise in your motor? Do you have a spec for its inductance?

 

[bob_s]

Audioguru,

I've found from experience that the motor requires an inrush of full current for at least 10ms to get started. I'll have to wait for the spec sheet on the motor until monday.

I've managed to free up an analog channel for the sensing circuit and I am planning on using it to check the sensing circuit once every 100ms.

You said that the circuit you changed would output 3.03v if R2 was 910k, that works for me. I've wanted to get an optocoupler to work in it's linear range but could not figure it out. How do I get the output to follow the input?

Bob

 

[audioguru2]

Bob,
Yes, the inertia of the motor causes a high start-up current, but its inductance will cause a current ramp-up for each current pulse while running at a constant speed/load. This will also delay the pulses and reduce their averaged level.

I am glad that you managed to free an analog channel to read the averaged voltage that will be equivalent to motor current.

The PS2501 has a broad range of coupling, for current that flows through its LED, to the current that flows through its coupled transistor, of 80 percent to 600 percent.
So the PS2501's transistor will need an opamp with an adjustable gain, for you to match the PS2501's coupling. Then the output will follow the input with a predictable scale. See page 6 on the data sheet here:
http://www.csd-nec.com/opto/english/pdf/PN10225EJ01V0DS.pdf

 

[bob_s]

Ok that's sounds great! But how do I make an adjustable gain op amp circuit?

Bob

 

[bob_s]

This seems to be an impossible task. Trying to adjust for the wide range of CTR, loss of LED integrity and temperature changes. I seems to me that tranfering the information is still best done digitally and then converting it to an analog signal or a timed digital signal on the other side. So I've decided to add a small micro (12F675) to supervise the circuit and send the data digitally.

Thanks for all your help. I've learned alot from your comments and suggestions.

Bob