Hum in SVT power amp?

[Mark]

Hi Merlin

Thank you so much for helping me out.

I have previously done something similar to what you said. I disconnected wire carrying the feedback signal and the hum was significantly louder.

I shorted out R2 the 470K grid leak resistor. I didn't put a big cap to ground by I shorted out R7 which is the gridleak to the following phase splitter stage. The noise reduction was dramatic, the noise level was still present but acceptable.

[Mark]

Actually, thank you to everyone who has helped me up to this point. I appreciated the new ideas which kept me moving forward.

[Merlinb]

Can you lift one end of the 0.1u coupling cap (I can't see what number it is)? It sounds like you may be getting heater buzz, coupling onto the high-impedance grid of the cathodyne. Does the humdinger have any effect?

[Mark]

I know this isn't conclusive but I got a Fender Champ power transformer and made DC supply to supply the nine pin valves heaters. Unfortunately it made no difference to the noise, but given this DC supply had no earth reference on the amp. It might not have proved much.

The hum reduction pot works like a charm, I can hear the noise increase and decrease in volume.

I'll lift the 0.1uF cap tomorrow as it is 9:00pm here.

Thanks for your advice.

[tubeswell]

Maybe try pulling out pairs of output tubes and just running the amp on each pair. (To see if it’s coming from a dud output tube putting the power amp out of balance? Have you confirmed that all the tubes are conducting the same?)

[Mark]

Sorry to be dragging the chain on this one, it's a bit daunting. I should mention all these readings were taken with the CRO probe on 10:1, I did this to avoid loud popping noises and loading.

IMG_0868.JPG

Here is the output of the amp with the only 12DW7 in the amp removed. This is a 12AX7 which amplifies the signal and a 12AU7 which is a splitter.

IMG_0869.JPG

Merlin asked what the noise was like when coupling cap C1 (0.1uF) was removed from circuit.

IMG_0872.JPG

The noise on the plate of the first triode with all valves in place. Note the feedback loop has been lifted.

IMG_0870.JPG

The noise on the first triode plate with the two 12BH7's removed, which effectively biases the output valves off. As shown the amplitude is a bit greater.

IMG_0871.JPG

As always all thoughts and suggestions welcome.

[tubeswell]

So it doesn't appear to be output tube related.

Seeing as how its in V1, I'm wondering if its a problem with transients from the PT somehow affected by bad terminal connections. Does the amp have terminal-and-spade connectors between boards? If assuming this is so, have you inspected them for signs of arcing, or looseness/oxidation?

[Mark]

Regarding the images of the waveforms, I should mention all these readings were taken with the CRO probe on 10:1, I did this to avoid loud popping noises and to a lesser extent loading.

All the wiring it soldered to the PCB. There are no spade connectors, it is very much a hum and not sporadic, the amplitude of the hum is constant.

I should post a picture of the waveform of the supply rail as well as the plate wave form.

[martin manning]

Your first photo looks like power supply ripple, with both 50Hz and 100Hz present.

[Mark]

Hi Martin, the first shot is the output of the amp.

Here are photos of the power supply ripple and the grid, cathode and plate. I decided to set the CRO probe on X1 today.

This illustrates what I find frustrating about this fault, just noise on the grid, noise and maybe a small amount of hum on the cathode (which I think would be normal as there is hum on the plate), and the power supply is as clean as a whistle. I have tried additional filtering and using a different earth point and it made no difference. Where do you go next with this thing?

IMG_0876.JPG

IMG_0877.JPG

IMG_0878.JPG

IMG_0879.JPG

IMG_0875.JPG

[tubeswell]

Could be a torn/partly lifted pad on the board that has cracked a trace. Or a bad solder joint on one of those. Or a cracked cable lead

[Mark]

Could be a torn/partly lifted pad on the board that has cracked a trace. Or a bad solder joint on one of those. Or a cracked cable lead

I'd expect spurious noise from that sort of fault. This amp has a consistent hum.

[tubeswell]

I have seen cracked traces cause s range of noises, sometimes crackling and popping, sometimes buzzing, sometimes high pitched oscillation. I wouldn’t rule it out.

[Mark]

Starting on the ridiculous now, I just lifted the plate resistor and to see if by any chance the noise was still present. The noise was gone but I assume the noise isn't due to the socket itself or an oddity around the socket.

The next thing is to lift any thing that goes to earth and attach it to a wire and try different earth locations. I'll do this with a dummy load, if the thing goes into oscillation, my hearing/sanity will never be the same.

I removed everything that goes to the grid and lifted the cathode resistor connection to the cathode. I used another 2.2K resistor and attached it to the cathode and soldered the grid wire to the earth end of the resistor. I attached a piece of wire to this junction where the test point and the earth point supply "B". The plectrum points to the earth location.

No change at in the amplitude or waveform of the hum.

IMG_0880.JPG

The pink section is of concern as it is the grid leak and cathode resistors of the following 12BH7 stages. I thought I had tried removing the 12BH7 valves I think it made no difference, but I will try this again or lift the resistors.

I'm thinking of putting the DC heaters back on the 9 pin valves and see if there is a change in the waveform. The next thing is to pull the 12BH7's and see if there is a change to the waveforms. After that, create a new power supply node and see if that has an effect. The preamp supply is part of a dual can cap that supplies the output stage.

Any other thoughts welcome.

[Mark]

I was surprised to see this in Wikipedia but it looks like it could be dead on.

https://en.m.wikipedia.org/wiki/Ampeg_SVT

The stuff I find interesting is this comment; "The SVT amps with 6146B tubes tend to put out a bit more power as well as have a more pronounced grind in low mids that makes these amps desirable for some bass players. The 6550A version of SVT amps tends to have a more round, deeper bass sound.. "

This is the thing my brother has been chasing since he first heard a vintage blue line SVT. Prior to buying his MTI SVT.

This also is a bit long winded but interesting; "The 6146B tube in itself does not have instability issues and is a reliable tube. The early driver circuit was not properly designed, therefore would blow the 6146B tubes on occasion. The reason most 6146 SVT amps blow tubes and resistors is due to front end 12BH7 voltage amplifier being fed from the 430 V node. During loud transients and overloads this will produce an AC signal that far exceed the 12BH7 follower that has 220 V on the plates. So when the follower grid is driven far over its own plate voltage it saturates on the positive half of the signal and thus takes over the BIAS voltage forcing it very positive. The time for this voltage to come back to normal is based on the time constant of the 150 kΩ mixer resistors and the coupling cap...by this time it is too late, the bias is pushed way positive and the current gets slammed through the 6146 and they tend to blow the tubes.

Some users have an electronics technician re-wire the 12BH7 feed same as on the later 6550 heads. By adding a 1K and filter cap feeding from the 220 V screen supply to the front voltage amp of the 12BH7, this keeps the 6146B tube running reliably with far fewer issues. Conversion of 6146B to 6550 tubes has a dramatic impact on the output power. The amp will produce roughly 225 watts, due to the screen voltage being too low. As such, the power transformer of a 6146B SVT will be about 220 V DC on the screen supply at idle instead of the typical 350 V idle screen voltage normally seen on 6550 amps.

One solution is wire the screens in a voltage double arrangement, which will end up at roughly 400 V screen voltage at idle. This will make for a very powerful 400 W SVT. Conversion from a 6550 tube to 6146B tube is a bit trickier, as the 6146B will not tolerate anything over 250 V on the screens or else it will arc over. Some amp technicians prefer to disassemble to PT and tap the windings from the side of the bobbin to create a lower voltage taps. Other methods are to use voltage regulation."

I don't tend to go along with having an excessive voltage on the screen, the amp May put out greater wattage but at the expense of reliability. I went in the opposite direction with my brother's amp, I replaced the 22 ohm resistors with 1K resistors as recommended by Ken Fischer. The amp was extremely reliable after that mod.

There is also the mentioning of the preamp being noisy, can hardy wait!