• 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.

Aikido hum has me stumped !!!!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I spent a pleasant afternoon with Bruce Anderson yesterday, well it was frustrating. We siimply couldn't nix the hum. Cap-coupling PCB ground, etc, etc. did nothing to improve things, although I did learn that high quality tubes are far less microphonic than my Russian cheapies!

We even tried a few things such as disconnecting the B+ to see if hum was comming from the AC. With heaters on and no HT, it was still there. "Ah" we though it's AC heater hum. However, our hopes were dashed. When we disconected the heater transformer and ran the tubes on B+ only, the hum persisted.

So, we are both totally stumped on this one, although it seems to be a ground loop. Now this morning, I tried to test ground connections from all kinds of places to the chassis ground. Something that I did notice was this:

Input and Output jacks are isolated from the chassis. There seems to be no chassis connection between signal ground and chassis via the pot.

So, I lift all connections to chassis/star-earth, and I get no connection between signal GND and chassis/star-earth bolt. I replace the PSU ground and I get 0 ohms from signal ground to chassis/star-earth - so there's a connection.

Now, I swap PSU ground for PCB ground and do the same measurements and I also get 0 ohms from signal ground to chassis/star-earth.

Finally, I reconnected everything, but removed the signal-gnd from input to PCB. Of course I get no sound from preamp/amp. But...no sound also means no hum!

So, could I have a ground loop in my signal wiring???

If so, how can I get rid of it????

Charlie
 
Remove all your grounds from chassis (safety) ground. The only thing that should be connected to the chassis for this go-round is the 3rd wire from the power line. Connect the now-isolated audio/power ground point containing all the other grounds to the chassis via a 20 ohm resistor. See if the hum goes away. If it does, put the lid back on and enjoy the music.
 
Sketch doesn't show here.

Without the boards in front of me, I can't remember exactly how the grounds are set up, but basically, ground the green wire from the power line to the chassis. Run all other grounds to the usual star point, but isolated from the chassis. Then connect the star point to the chassis via the resistor.
 
Sy,

OK, any particular wattage resistor???

Also, I did have a brief period of quiet last night. It's a little like Awakenings. I redid all my solder joints, anyway I still get hum this morning. I could also try to couple the PCB to the chassis via the PCB mounting bolt and a cap or resistor. Broskie does say that it can be done this way. It would move the chassis/PCB/PSU GND away from the IEC GND, and maybe that would be a good idea.

I tried to post my circuit layout, but I don't think it got put with my postings.


Charlie
 
Charlie,

seems as if hum is being really pesky to you...

Not that I have alot of experience with hum but having built almost TWO ( ;) ) headphone amplifiers I know my side of the story.

After trying almost everything in the end I found a major ground loop (I grounded a shielded input cable on both sides...duh). Another thing to check when you are using a metal top plate...Are the rca's really isolated from your top plate? Sometimes the punched holes are too small and make contact with the inner metal part of the rca...

And last: carefully wiggle all cables or even bigger parts (with an isolated screwdriver) and check whether hum changes somehow...

good luck,
Oliver
 
I'll stop at my local electronic store and pick up a 20 ohm 1/4W resistors. I thought I had a pretty good collection, but nothing low.

As for my jacks, my chassis has wood sides, front and back with a metal top plate which carries all components except for the IEC, switch and jacks which are located on a back plate. It is couple to the top chassis via the same bolt as my earth bolt, which is actually one of the transformer mounting bolts. Maybe I need to move this bolt away from the transformers???

My jacks show open-circuit to ground, so they are likely not the problem.

Thanks for the replies.

Charlie
 
Thanks to Sy, who gave me the clue to hum removal from the Aikido. I had been connecting my PSU ground directly to the chassis ground. There is another ground wire from the Aikido PCB to chassis ground. The bolt for the chassis ground was the same used for the 6.3VAC transformer.

So, Sy's suggestion was to lift my PSU/PCB ground and connect them to the chassis ground through a 20 ohm resistor. Well, I don't have the resistor in place yet, but it got me thinking.

Now the Aikido PCB has a pad marked GND. It also has a pad marked chassis which connects to the chassis via one of the PCB mounting bolts. There's also a jumper to connect the GND pad with chassis pad. So, I connected PSU ground to the PCB ground pad, and jumpered to chassis directly.

The hum is discernable, but only just. Maybe swapping the jumper with a 20 ohm resistor or 0.1uF cap will be even better. Now I have a Wima 0.1uF cap, which Bruce and I used on Saturday in our search for hum reduction. Of course, one leg finally gave up the ghost and snapped clean off. I will buy a cheap polyester cap tomorrow when I buy my 20 ohm resistor.

Sy, when you said that some folk use back-to-back diodes for the earth connection, which kind would you suggest, and do they go cathode-to-cathode or anode-to-anode.

Right now, my Aikido is humming (in the sense of great sound, no hum) I have already tried a number of music CD's - Michael Hedges, Chanticleer, and it all sound bl**dy amazing.

Of course, I'll wake up in a couple of minutes to find that I've been dreaming......

Anyway, in light of tonight's Golden Globes, I would like to thank the following people without whom, this production would not have been possible: Bruce, Sy, BurnedFingers, JoshK, GingerTube. This is a great forum and in the next couple of weeks, I'll be asking for someone to check my resistor values for wiring a blue led to my B+ voltage.


Thanks,
Charlie
 
cbutterworth said:
Maybe swapping the jumper with a 20 ohm resistor or 0.1uF cap will be even better.

You mean resistor and cap, not resistor or cap?

Seems like any combination of resistor, cap, antiparallel diodes, or each alone, except for a single cap, would work. In fact, stout diodes (fuse blows before diode approaches it's current rating) allow the use of a low watt resistor, so you wouldn't leave things unearthed (so to speak) if a fault burned out the resistor.

Sheldon
 
Sheldon,

You mean that a single 0.1uF cap is not a good idea? What about a 20 ohm resistor paralleled with a 0.1uF cap?

Or 1N4007 dioides (as per Sy) parallelling a 20 ohm 1/4W resistor? Would the 1N4007 diodes be strong enough? I already have two fuses on the system, one immediately after the IEC, the second between transformer and rectifier.

I have to reinstall the tube-delay circuit (and hopefully keep the hum from returning), and the source selector. Once I do this, I'll take some pics and send them in.

Bruce was a great help, and gave me a couple of better quality tubes - two Magnavox and two GE's. My Russian cheapies were incredibly microphonic - I would lightly rub the chassis and they would sing. Bruce reckons that the Magnavox and the GE's are cheap tubes, but I'll certainly stick with them until I can afford better ones.

Charlie
 
Jayme,

I have no idea which Russian tubes I was using. They cost $4.95 each new from Antique Electronic Supply. I have looked for branding on them and found none. Of course, what can you expect for that price - maybe not all that much?

I'd love to find some really good, but affordable tubes - maybe RCA's, but until I do, I will thoroughly enjoy my music with the Magnavox and GE tubes.

Charlie
 
The point of the resistor or diodes is to allow a path to earth ground, but make it "less desirable," so it will only be used if there is a short to the chassis.

No need for a cap. You are not trying to filter anything.

A 1N4007 will be fine for the diodes. They are rated for 1A, but usually around 30A for instantaneuos peak, which is the condition we are talking about here (short to chassis). Some people use a rectifier bridge here for the higher current and ease of mounting.
 
Brain,

OK, I'll try and get 1N4007's. However, my chassis is earthed already, so a short to the chassis is already grounded directly to the IEC earth.

For the PCB/PSU ground, I can certainly use a combination of diodes, or even buy a 22 ohm 1W resistor. Broskie in his instructions suggests that a 0.1uF cap between PCB/PSU and chassis works and can get rid of hum.

I am getting confused again, but at least I am not frustrated this time, as the Aikido works and the hum is only negligible.

Charlie
 
Bruce (I assume that's you, tinear9). Of course, I could have chosen a similar email name, but tinitus9 may be more appropriate for me - although I can distinguish tinitus from preamp hum I think :)

Yes, I am very relieved that things are going well with the Aikido. There are a few immediate mods to make - solder all PSU wiring, rather than use spade lugs. Add the source selector, and re-install the tube-delay circuitry. At the very least, I need to do the source selector, as I have no room to fit my Aikido as well as passive on my HiFi stand.

Your donated tubes work are so much better than my originals. My dogs can bark and run around without causing stress to the preamp anymore.

I hope things are recovering at your place, after Saturday's solder-fest.

I really appreciate your help, and we should get together once things settle down.

Regards,
Charlie
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.