Hi team,
My customer is having an issue with UC3638-
On the first attached png file the amplitude of the triangle wave (scope C1) at UC3638DW pin CT drops to about 60% of normal for the two cycles on either side of the center of the trace. The red and blue traces (C2, C3) are top and bottom of the load driven by the PWM that the UC3638DW is controlling and F1 is the difference of C2 and C3. C4 is the current through the load.
Glitches in the triangle wave can be seen on the previous and subsequent cycles surrounding the two runt peaks. The glitches are at an amplitude near the 60% of the normal level.
They have used this UC3638DW circuit configuration and component values on two previous production boards, the difference here being that the amplitude at the load is higher with this board and they changed the gate driver IC that the UC3638DW is controlling. The old gate driver is IR2110SPBF, the new gate driver is 2EDF7275FXUMA1.
The problem shown in the waveform is occurring near the maximum Voltage that they send to the load. The prototypes have 160V maximum output Voltage and the production boards are at 140V maximum. The prototype amplifier gain is greater now providing the increased Voltage output range (along with 170V vs the old 150V power supply).
The output currents between production boards and prototypes should be similar at the same PWM frequency and output signal rise time.
The first scope traces are only showing a fraction of the horizontal waveform compared to the second image showing the upper part of the rising edge of the output Voltage and the green current waveform. The first waveform shows the PWM signals.
The second image's voltage waveform is measured differentially so you don’t see the actual amplitude, instead you see the amplitude across the load which is biased about +20V at the bottom terminal. The F1 rising edge should be more linear and the C4 current waveform should not have three distinct peaks.
The second image waveform is with 400kHz PWM where the first image waveforms were operating at 200kHZ PWM.
What are possible/likely causes of the reduction of the triangle wave amplitude?
I also have the UC3638DW circuit connections- if you need those as well please reach out to me internally as I don't want to post their schematic on E2E.
Thank you,
Lauren
