Hum in SVT power amp?

[martin manning]

Interesting, Mark. Transient behavior is not well understood by most techs and maybe the designers as well. Clearly some careful consideration is required in high-power amps such as these.

[Mark]

Here is a bit more info on SVT's, I find that Ampeg dumped the 22 ohm resistors and diodes in favour of a 220 ohm resistor. This is good info to know.

SVT_CL_Service_Bulletin.pdf

This is a more modern SVT circuit diagram, I was looking if there were changes to the supply rail and the points of the circuit that were being fed. Any other diagrams displaying this would be appreciated.

Ampeg_SVT-VR.pdf

[tubeswell]

Some SVT-CL drawings below - more to follow

[tubeswell]

Remainder

[Mark]

Thanks for the attachments, the power supply docs aren't as easy to understand as a circuit diagram. I wonder which genius at Ampeg thought that would be a good idea.

[tubeswell]

Thanks for the attachments, the power supply docs aren't as easy to understand as a circuit diagram. I wonder which genius at Ampeg thought that would be a good idea.

It's not any easier to look at in real life LoL. But if you stare at it long enough, it becomes clear which are the 'euro' voltage connections etc

[Mark]

I'll get back into it this weekend, I'll remove the 12BH7's and see if the waveform is altered. What concerns me about the waveform is that it looks like an inverted fullwave rectified waveform without filtering.

I was never too keen how MTI/Ampeg wired up the power supply. The preamp supply is attached to the same power supply source as the output stage. One side of the dual cap was is used the output stage filtering and the other for the preamp supply. I suspect they did this as the output stage power supply had more serial stages and filtering. Whereas the 12BH7 power supply stages have less filtering stages.

I think I have tried using another cap with a different earth path and there was no
Improvement, the other idea is to put an extra stage on the 12BH7 power supply and see if that has any benefit.

I'll add to this post. Any ideas let me know.

[Mark]

I removed the 12DW7 filter cap (lower half of main filter cap) and used two 100uF caps in series. The earth was isolated so I could put the earth where I wanted it. I initially used the main star point, which is the body of the lower mains filter cap and later I used the earth point for the 12BH7's. There was no change to the noise waveform.

I thought the issue might have been the output stage rail being used, so I used a 1K resistor and the 50uF cap to form a new lower voltage off the junction of R52 and C9b. The the was almost no difference in the shape the waveform. I was so surprised, as I thought there would a significant change in waveform. The only difference was the dip in the lower part of the waveform was flat.

https://ampgarage.com/forum/download/fi ... &mode=view

Back to the drawing board.

[Mark]

I took advice from Aaron, who said I ought to remove all the valves bar the 12DW7. I did this and the noise/hum was still present. On a whim, I thought I should also lift the 15K plate resistor of the cathodyne phase splitter which follows the stage causing the issue. To my surprise the noise/hum disappeared, there was nothing.

I thought if I reconnected the plate resistor to the supply and lift the cathode connection I should get the same results. However, this didn't happen, instead I had the hum at the original level.

I'm surprised as the valve shouldn't be operating and the only thing connected to the plate is the coupling cap and a 20pF cap to the previous plate resistor. Which would be there to suppress parasitic oscillation.

Any thoughts on this one?

[erwin_ve]

What happens when you remove the 20pF capacitor. Ceramics can act as a antenna.

[martin manning]

Doesn't that suggest that the hum is coming from the power supply, possibly the filter section that feeds that plate resistor?

[Stevem]

What's the ripple voltage on the bias supply?

[Mark]

Hi Erwin

Thanks for your input. I removed the side of the ceramic cap going to grid of the first stage (this is the stage where the plate has the noise/hum on it) and it made no difference at all.

Hi Martin

Thanks again for your assistance.

Well I could be wrong about the power supply node, but both the first stage and the cathodyne phase splitter are connected to the same power supply node and earth bus. I'd think lifting the plate resistor from the supply node would mean the hum would still be present on the first gain stage.

I'm going to lift the coupling cap and see if that has any effect on the noise/hum. The plate of the cathodyne phase splitter then would only have the plate resistor on it. Given the cathode also has a coupling cap etc. I don't see why this would make any difference.

[Mark]

Hi Steven

The ripple on the bias supply is bad as there is only one 100uF filter cap, but all the valves bar the first 12DW7 have been removed, the feedback loop has been lifted. I'm looking at the plate of the first stage as that is where the noise/hum originates from.

Update: I lifted coupling cap C3 (0.1uF) and there is no change in the noise/hum. I'll draw a circuit diagram of the current circuit as it looks tomorrow.

[Stevem]

What takes place if you move the bias test point ground right to the grounding point of the recto bridge?