Current sense signal lags current on PWM controller

[djroseman]

Hi,

I am using a UC2842 PWM chip to control a boost converter, boosting to 120Vdc with a 70Vdc input. The issue is that the Isense input to the PWM chip (no slope compensation fitted) seems to lag the current through the Inductor and sense resistor (0.01R fitted in -ve line)- the two opamp inputs are connected to either side of this resistor.

Could anyone give me any pointers on how to align the Isense signal with the current signal?

I have attached a scope trace, colours are: Blue=UC2842 oscillator (52kHz), yellow = gate drive output, green = inductor current, purple=Isense signal into PWM.


Regards
Dave

 

[djroseman]

I wasn't able to include a schematic with my first post as well, so it is attached to this one

Regards
Dave

 

[zappedagain]

It looks like U2 may be going unstable; what looks like lag may be the recovery time to get out of oscillation.

What else is connected to U1-3? You didn't show the entire net. Do you have the rest of the filter circuit (resistor to ground) shown on p. 5 of the UC2842 data sheet?

A 0.1uF or 0.01uF cap across C1 might help too; C1 has series inductance that can make it work poorly at higher frequencies. Make sure the new cap is mounted as close to U2-8 and U2-4 as possible, with as short as leads as possible.

You might also try reconfiguring the RJ/CB filter for less capacitance. The LM2904 data sheet states it can drive 50 pF; you might be getting close to that with the unspecified load of the UC2842D and other parasitic capacitance on the board.

Back to your original lag question, have you calculated the lag through your filter circuit?


Let us know how this all works out.

John D

 

[djroseman]

After a bit more playing about with this, I managed to get it working by replacing the LM2904 opamp with an analog devices part OP213, which is a bit of a sledgehammer to crack a nut, but other opamps did not make it stable.

--dave

 

[zappedagain]

I'm glad you got it working. Op-amps get really interesting when they can't drive the load capacitance, especially when you only monitor them with a multimeter! The oscillation just looks like an odd DC drift as the duty cycle varies with a multimeter.

John D