Make sure the power xfmr you bought is the right one

[Dave H]

Actually, it's not. From the 0mA baseline, center graticle on the scope, you'll see the flat part of the waveform is ~ 150mA, then we're seeing this negative-going charge. Initially, I had the current probe coupled in the opposite direction, but that put the flat part of the curve at -150mA.

Measurement between bridge +ve and first filter cap is after (positive) rectification therefore the whole waveform must be positive. The negative offset has to be caused by the probe. I think your initial measurement with the flat part moved to 0mA is what the circuit is actually doing. Edit: i.e. like this from your last post.

[nevetslab]

Measurement between bridge +ve and first filter cap is after (positive) rectification therefore the whole waveform must be positive. The negative offset has to be caused by the probe. I think your initial measurement with the flat part moved to 0mA is what the circuit is actually doing. Edit: i.e. like this from your last post.

[ATTACH=CONFIG]53006[/ATTACH]

This image was made on the charging current flowing into ground from the (-) bridge connection, which obviously flows into the cap. No DC Current level, as that potential is at ground. When I flipped over the current probe on the wire from the (+) side of the bridge to the cap(s) in the doghouse below the chassis, the flat portion is about 130mA below ground...with the same peak current rising positive like this. That's not something the current probe is doing, or its' mating probe amp. I'm likewise confused by this. We can see the peak charging current stops as the ripple passes thru 0V. Now granted, that is AC coupled X10 probe where I had set the baseline at the center graticle. At any rate, the same peak current shows the same....around 760mA Peak thru the supply cap (100uF/500V part). The RMS value of the 10V P-P ripple measures about 2.9V RMS.

[nevetslab]

Charging Current and discharge current on the (+) bridge connection feed to C30 filter

Measurement between bridge +ve and first filter cap is after (positive) rectification therefore the whole waveform must be positive. The negative offset has to be caused by the probe. I think your initial measurement with the flat part moved to 0mA is what the circuit is actually doing. Edit: i.e. like this from your last post.

[ATTACH=CONFIG]53006[/ATTACH]

In rethinking the current measurement on the (+) connection between Bridge and Filter Cap, where there is a negative 150mA steady current below the 0mA baseline, I believe that -150mA is showing discharge load current into the O/T Primary & thru the choke to the rest of the amp, while the 120Hz charging peaks are indeed positive current into the cap.



On the second image, where the current probe is sensing current flowing into ground from the (-) bridge connection, there isn't current flowing out of ground to the load, so the baseline remains at 0mA. The charging current into the cap(s) still remains around 760mA Peak, measured from the baseline of the current waveforms. The third image is the load current flowing into the O/T Primary C/T.

Is this a sensible explanation for the -150mA current line below 0mA on the first image? The current flowing thru the choke is about 20mA, feeding the screens and preamp/driver tubes.

Remember, a DC Current Probe senses current flowing thru a wire. In this case on the Twin Reverb, there is only one wire to the filter caps below in the doghouse, while the load current from it (O/T Primary, Choke-connected screens and preamp/driver tubes) all connect up on this main center PCB, so the load current is flowing in the opposite direction as the charging current thru that single wire.

[Dave H]

In this case on the Twin Reverb, there is only one wire to the filter caps below in the doghouse, while the load current from it (O/T Primary, Choke-connected screens and preamp/driver tubes) all connect up on this main center PCB, so the load current is flowing in the opposite direction as the charging current thru that single wire.

Right I didn't realize it was wired like that. It explains the steady -150mA. It would have been better with two wires to the cap, one to the bridge and one to the OT, choke etc.

Edit: I've changed my sim to have only one wire to the cap and it does reproduce the effect.

[nevetslab]

Right I didn't realize it was wired like that. It explains the steady -150mA. It would have been better with two wires to the cap, one to the bridge and one to the OT, choke etc.

Edit: I've changed my sim to have only one wire to the cap and it does reproduce the effect.

[ATTACH=CONFIG]53066[/ATTACH]

I love it when math & physics work! There are days here when I question that though.

I had another Fender Twin on the bench yesterday, dealing with different issues, but, while it was on the service cradle, and I hadn't yet checked/adjusted the bias (plate current on the tube set was around 24mA each), I was seeing the same trend with the current probe. Once I raised the bias for 33-34mA on the tubes, I had the same charging current, though with this set of caps, (still series-connected 220uF/300V), the ripple was less....8V P-P). So, same trend.

I'm still running the previous Twin with the new power xfmr at idle off to the side, where I can monitor the power xfmr temp/chassis temp. Of course the transformer temp is governed by the plate current dialed in with the bias. After 4 hrs yesterday at idle, the core temp got to 72 deg C, chassis at the rear panel got to 62 deg C. I'll see how it is the end of today.