Voltage sensing resistor

[arber333]

Hi Johannes

I am somewhat ahead in my design of dual motor board for Volt/Ampera inverter. Both motor work on its inverter side. But now i wanted to try them together and with working logic. Here i hit a problem.
By erroneus ommision of 10K resistor inline with voltage sensing line on BOTH sides i obviously toasted one Olimex as well as opamp!
Clearly i failed to use resistor and i know that now. My question is more like why did you decide to use 10K resistor? Did you figure this just as ROT or is it required by fact? I want to use 4K7 resistor with two inputs forked on the line after resistor. Would i still be ok? Or is 10K neccessary for proper gain?
So whatever resistor i use, i need to put it before the fork in sensing lines obviously.

tnx

A

[johu]

I don't yet know which resistor you mean? Which board, which ref designator?

[arber333]

I don't yet know which resistor you mean? Which board, which ref designator?

R47 of the PC3 voltage sensing line of rev. 3 board.

Damn! i just found 3 of my driver lines are pulling the signal down and Fault is constantly on! Both sides! I think my Volt drivers just died. I never noticed they are off. Do you think 470R pullup would be too much for them?
I will try to repair them.

EDIT: I think they are just optos. Drivers are ok since i get good Fault signal on the line. I will need to replace optos.

[johu]

ok. By "omitting" R47 do you mean you directly bridged it? It is 10k for low pass filtering and pin current limiting reasons. But 4k7 should also work, question is, why change it?

Why would you need to pull up the optos? Isn't it like 5V -> anode -> cathode -> resistor -> stm32 ? Thats how it is with the Leaf inverter, the resistor is dimensioned for a sane forward current. Should be around 200Ohm = (5V-1.2V)/20mA.

[arber333]

ok. By "omitting" R47 do you mean you directly bridged it? It is 10k for low pass filtering and pin current limiting reasons. But 4k7 should also work, question is, why change it?

Why would you need to pull up the optos? Isn't it like 5V -> anode -> cathode -> resistor -> stm32 ? Thats how it is with the Leaf inverter, the resistor is dimensioned for a sane forward current. Should be around 200Ohm = (5V-1.2V)/20mA.

Instead of using 10K resistor i just put one wire bridge. It was OK as long as i used only one Olimex online. With two chips it obviously caused OpAmp meltdown as well as causing fault in isolated amp. Of course even DCDC went beserk.... This evidently caused one Olimex chip to release magic smoke as it does not connect to STlink. I replaced them all and tried again with 4K7 resistor and i can now get accurate voltage sensing. So this circuit should not cause any inaccuracies with R47 being 4K7 instead of 10K?
I hope rest can learn something from this at least....

My setup is wired from PWM pins directly to drivers with 470R pullup resistors on each PWM line. Drivers have optocouplers online which have to maintain open collector when INACTIVE! If i remove power driver opens IGBT!
However that chip near death must have directly released voltage and burned through 3 of the driver opto isolators down on the driver board.
I can hear IGBTs in the motor winding when i start inverter. Since thair line is 0V they come online. Of course there is immediate Fault signal.
I should add 220R inline to PWM with 470R pullup but that would create a divider and i wasnt sure in draining capability of optos.