Power transformer asymmetry and peaks

[anson]

My power tranny secondary HT is set up with a full wave bridge rectifier and is giving different waveforms on each input to the bridge when I O-scope it, i.e. the two places where the diode cathode contacts a diode anode. That's shown on the picture i included, with one wavedorm greyed and the other dark. They both swing to about 480V p-p now. Should there actually be such different shape between the two sides, or is something wrong with my tranny? Also, at the side of the bridge contacting the filter caps, there are 10 to 40V noise peaks that ive been unable to snubber or post-bridhe RC filter away. I cant get rid of them.They are represented in the other picture i included and only go to about 15V in tjis photo, but usually its 40V.

I've been having problems with audible ( 2kHz and beyond ) harshness in this amp, and am persuing this power supply asymetrey/noise problem with a hope it may fix the harshness.

The transformer (classictone 40-18069) has always hummed audibly. The amp has 4xkt88. It draws an idle current of 0.44 A. The wierd waveform assymetry, peaks, and humming exist regardless of whether there are loads on the PT secondary side or not. The rectifier has 4 x UF4007 diodes.

Do you have any ideas of what I could test before i give up and buy a new transformer? The amp goes plenty loud, and seems to rectify to the expected DC voltage. But there is too much harshness at 2kHz+ and i want to get rid of those spikes to see if they play a role.

[Jon Snell]

No photo was attached as I view your post at 07:40. Please try uploading again and add your schematic.
Disconnect the bridge and measure the AC voltage with no load.
If they are the same, nothing wrong with the transformer but without a schematic, I have no way of knowing what else could be causing your concern.
Perhaps there is a bias winding with a half wave load causing it ... I have no idea.

[loudthud]

A Full Wave Bridge is a strange animal. Unless the diodes are conducting, the transformer winding is just floating, it isn't connected to anything... So interwinding capacitance to the primary can make a strange waveform and a different waveform on each side. Try making an artificial center tap for the secondary with a pair of 100K 3W resistors and connect the artificial center tap to the center of two filter caps that are in series, or just a filter cap to ground. The Voltage on the cap should be half the final B+ Voltage.

I think this is the data sheet for the transformer: https://s3.amazonaws.com/tubedepot-c...s/40-18069.pdf

Some the the connections shown on (edit) sheet three could make a really strange waveform. I don't think that is what is causing the problems with your the amp.

[Mick Bailey]

In addition, if you're getting transformer hum with the secondaries disconnected there's perhaps another problem with loose laminations or a loose bobbin or winding, assuming the transformer is not vibrating against the chassis.

[anson]

Here is an image of the waveforms on each of the "input" sides of the FWBR with the only load being the filter caps each paralleled with a 220k resistor. Clearly there is asymmetry between the sides of the FWBR. But I made a virtual center tap as LoudThud suggested (but 2x 220k resistors) and the asymmetry went away and each side of the HT convincingly showed the same amplitude (480V on my scope) and shape. Following through with LoudThud's suggested experiment, the voltage at B+ was 487V (dmm), and the voltage at the virtual center tap was 232V(dmm). Not quite half.

I'm sharing a photo of the spikes seen only on the B+ side of the FWBR. This was really my main concern. Here they are nearing 40V. What's this? No tubes are installed for these oscilloscope images. It shows up for two different bridges, one standard speed anf the other UF. The spikes still show. My scope won't capture the frequency... maybe some decaying ringing that is being rectified? This also ofcourse shows up on the OT primary. I have seen and heard harshness at about 2kHz and above that I would like to get rid of and hence my concern towards getting rid of these spikes.

FYI, I'm using the simpler HT winding arrangments presented on the datasheet LoudThud linked to, which is the 0V and 330V tap, but I used the 380V tap too and still see nearly 40V peaks at the B+.

Thanks for your thoughts Jon. With all secondaries open, one side of the HT is 260V pp, and the other is 680C pp.

[Helmholtz]

This also ofcourse shows up on the OT primary.

Does it?
What matters to sound is what you see on the OT secondary.

How does B+ look?

You should use a 100x probe to not exceed probe voltage limits.

[anson]

Does it?
What matters to sound is what you see on the OT secondary.

How does B+ look?

You should use a 100x probe to not exceed probe voltage limits.

The B+ has 10 to 40V spikes all over it. The spikes are an extremely rapid increase in voltage to about 10 to 40V above B+, then decay that takes about 0.5ms or so. I uploaded an oscilloscope image that shows the spikes on channel 1--I hope it uploaded correctly, looks OK on my end. I measured the voltage on the OT primary and there are 10 to 40V spikes and that as well as the B+. I used a 100:1 probe. I actually haven't scoped the OT secondary yet--I've been using a histogram from a close mic on the speaker and observed harshness at 2kHz and above, audibly and visually on the histogram.

[Helmholtz]

So the right side pic shows the spikes on the reservoir cap with load disconnected?

A schematic would be helpful.

[anson]

So the right side pic shows the spikes on the reservoir cap with load disconnected?

A schematic would be helpful.

That's correct, the pic on the right shows the spikes in the reservoir cap with load disconnected.

I attached a schematic. The preamp and OT have been disconnected, so it's pretty simple right now.

[Mick Bailey]

My thought would be to completely disconnect the secondary and run the transformer into a purely resistive load at the amp's idle current. Maybe a 50W 12K Ohm resistor, though this will get hot fairly quickly. Should be enough time to scope the output and see if the transformer still hums mechanically. At the moment you don't have anything eliminated from the PSU to act as a known-good reference point.

[Helmholtz]

That's correct, the pic on the right shows the spikes in the reservoir cap with load disconnected.

Ok, but without load is no realistic test condition.
I'd expect a strong damping effect by the load.
So the question is if you can see spikes (apart from normal ripple) with load connected.

Nevertheless I would check the ESR values of your filter caps.

[anson]

Ok, but without load is no realistic test condition.
I'd expect a strong damping effect by the load.
So the question is if you can see spikes (apart from normal ripple) with load connected.

Nevertheless I would check the ESR values of your filter caps.

The spikes I'm seeing at the filter cap/bridge rectifier junction exist regardless of whether or not there is a connection to the OT. I identified the presence of these spikes on a fully functioning amp. I then bit by bit disconnected the power amp secondaries and removed the load in an effort to hunt down the spikes. There is a choke between the FWBR and the preamp, and the preamp power supply is all very clean.

Concerning ESR, the filter capacitors (two series 560uF, each paralleled with a 220k resistor) are stated as 296 mOhms on the datasheet. How does that value compare to desired ESR values for a guitar amp power supply? I don't have a pulse generator to use to measure the ESR. Is there another way to measure ESR of my caps?

Thank you for all your suggestions and comments so far.

[Helmholtz]

The spikes I'm seeing at the filter cap/bridge rectifier junction exist regardless of whether or not there is a connection to the OT. I identified the presence of these spikes on a fully functioning amp. I then bit by bit disconnected the power amp secondaries and removed the load in an effort to hunt down the spikes. There is a choke between the FWBR and the preamp, and the preamp power supply is all very clean.

Can you post a scope pic of the voltage across both smoothing caps with amp fully connected and out of standby?
Probe tip to positive end of upper 560µ cap and ground clip of probe to negative end of lower cap.

Concerning ESR, the filter capacitors (two series 560uF, each paralleled with a 220k resistor) are stated as 296 mOhms on the datasheet. How does that value compare to desired ESR values for a guitar amp power supply? I don't have a pulse generator to use to measure the ESR. Is there another way to measure ESR of my caps?

ESR of 0.3R would be fine. But if so it wiould take a 50A current to produce a 30V peak, which doesn't seem possible.
Is your scope probe calibrated?

There are ESR meters starting from around 20USD.