• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Strange DC choke behavior

Hi all:

Getting 12-18 inches of snow here in Wisconsin. Record year of snowfall, or at least darn close. I can’t remember snowfalls like this since I was a kid. Back then, we built snow forts you could actually stand up in! That’s cool. But not anymore. Now it’s just another project for dad to strain his back on. Keep that roof cleared, thaw out the gutters, etc. But it is fun to drive in :)

Wondering what your experience has been with the results of PSUD with choke input filters: specifically, the voltage across the choke. A screen shot of my DC filament supply is attached; PSUD shows a fairly clean dc rectified-looking waveform, but this is not what my scope shows. I get a fairly ugly looking wave reminiscent of core saturation.

Currents look correct and sinusoidal, and output voltage is extremely close to PSUD’s prediction, within a half volt. The choke is rated for 35 mHy, 2A DC, so I’m nowhere close to overloading it that I can see. My PEAK choke current is only 1.3A, and never approaches zero, so I have met critical inductance.

Reason I bring this up is my output transformer is picking up the 120Hz garbage. This is essentially a modified Raven, http://www.nutshellhifi.com/Raven-MarkII.gif . So the high impedance primary is picking up the noise like an antenna, and it is present differentially across the winding ends. Even finds its way out to the balanced output. B+ is silent – shunt regulated supply. Turn the filament supply off, all noise pickup is gone. But guaranteed it’s coming from the choke – look at both choke wave and output transformer noise, and they are identical. Pull the tubes, put in resistors as dummy loads – noise still present across the transformer. Definitely the choke. Spacing is about 3 inches between choke and transformer, and there is a copper shield in between the two.

Maybe I just have a bad choke ?!?
I should mention, the problem has worsened over the past few weeks, but appears to have stabilized.
 

Attachments

  • chokeinput.pdf
    26.2 KB · Views: 82
It's almost certainly magnetic coupling. The choke is gapped and leaks quite a bit of flux.

Try temporarily moving the choke farther away from the OPT and/or rotating it. I've had significant magnetic coupling between OPT and input choke at that distance.

You can also put a flux band around the choke (copper foil, around the outide of the core). It _might_ help.

Pete
 
Very good suggestion.

Yes, I did, with no difference. Actually I have an RC network across the choke that really cleaned up any ringing. Whether or not the cap was right after the rectifiers, or across the choke, it performed the same function. Off hand, I think it was C = 4uF, R = 10 ohm.

But this solution seemed to resolve ringing, not the 120Hz hum.

Very odd.
 

kevinkr

Administrator
Paid Member
I'm not sure that it is all that odd, with a choke input filter you have the entire 120Hz ripple current flowing across it to ground through your first filter cap which presumably is quite a large one, and as a result the field will have a very strong 120Hz component in it.

I've had some similar problems with choke input filters in the past and that's unfortunately why I don't use them on the same chassis with the amplifier.

Distance is your friend in this case, and I'm not too sure there is any reasonable solution (I would hope otherwise) other than to remove and relocate much further away short of a very expensive heavily shielded replacement choke or replace this supply with a regulated filament supply. (Maybe CCS based?)
 
kevinkr said:
I've had some similar problems with choke input filters in the past and that's unfortunately why I don't use them on the same chassis with the amplifier.

That's good to know; I suspect the culprit has finally been found.

Distance is your friend in this case

I'll relocate and advise.

Pete: I will try putting in a 4 ohm resistor (according to PSUD, this will give about the same output) and see how improved the radiation is.

I'll take a scope shot of the choke voltage, too. It's uuuuugly.
 
Got it ! In between shoveling out of over a foot of snow (that's an official number, and extraordinary for SE Wisconsin), I went to town on cutting and splicing.

And the final result really surprised me. Essentially, both chokes (B+ and filament supply) were a source of hum pickup into the output transformer. What had me surprised was that the B+ choke was mounted above the earth grounded chassis, which is made of steel. I was expecting nothing on the top of the chassis to produce a field through the chassis, but wrong again was I.

Obviously, I couldn't get rid of the B+ choke. Fortunately, the filament supply choke was the bigger problem, and removing that effectively dropped the 'nasties' out of the output transformer, leaving me with a much cleaner and smaller 120Hz. At this point (I believe) the CMRR of the differential stage did its job, and the secondary of the output transformer is very quiet and clean. Just a tiny bit of 120Hz can be seen, with the scope on its maximum gain.

Now the fun begins. I replaced the choke with a resistor, which worked, but I had a conglomeration of oversized wirewounds in order to give me the value I wanted. The innards were so tight, I also had no room to mount them. So I tried removing the DC filter/regulator, giving me lots of room. Threw in a 6.3VCT transformer, elevated the center tap to 30V DC, and gave her a try.

The output was as silent as I could expect. Ear literally has to touch the speaker, and all I hear is the background noise of the amp (which is DC heated, and very quiet). So, the linestage is completely silent, with AC heating no less !!

This excites me, because it at least confirms I can be successful in my next amp using AC. Saving the complexity and space due to DC is a benefit I look forward to. Also, having the B+ supply in a separate chassis will be an added bonus, removing those frustrating chokes from the equation. Being the next amp will be all transformer coupled, I don't want to fight this issue again. Took me a few weeks just to find the problem.

Thanks for the help, guys !!
 
zigzagflux said:
Got it ! In between shoveling out of over a foot of snow (that's an official number, and extraordinary for SE Wisconsin), I went to town on cutting and splicing.

We really got a crazy amount of snow, didn't we? It was impossible to even see very far most of the day here. Hopefully tomorrow will be a bit more tame. I wish I was able to get some tube work done amongst the crazy weather. Maybe next time!