• 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.

Debugging my Aikido

I don't understand why this one channel has so much humm. Before the cap (measuring on the 2mV setting on the scope) I get no humm. After the output cap I get humm 10mV setting? Did not use a flash you there are a few extra sinus waves ;)

Probably simple to some of you ...but I don't get it.

[EDIT] To save you from reading this whole thread...the 2mV reading was bad..for some reason....in the end I had the 100k resistor soldered to pin 6 instead of pin 7

[IMGDEAD]http://basenjes.de/tubes/images/diyaudio/aikido_schematic_debug.jpg[/IMGDEAD]
 
Hi Bas. Kind of weird isn't it. I'm definitely not the electronics expert but I did build the Aikido and it worked perfectly first try. NO noise! GREAT sound! I think we can say the 1 meg output to ground resister is unlikely to be generating this noise. Everything to the left of the output cap must be ok. So, what is left? Can it be a bad solder joint on that 1 meg resister rectifying and picking up external noise? OT: Just curious as to why you chose 390 ohms for the 12SN7. At 300 volts B+ you must be running it fairly hot. At 876 ohms it runs 5 ma which I thought sounded about right.
 
Hi Bruce,
chose 390 ohms for the 12SN7

I think the example John Broskie had was 430R I took the nearest value I had..;)

There is a lot less than 300V B+ I think...I just used back to back transformers..nice n cheap..just to see what the aikido could do. I was thinking that because of the inherent noise reduction..the b+ did not have to be perfect...
 
Hi EC,

I don't have any scope experience or knowledge the probe is homemade. But listening tests confirm the humm.. the other channel has some noise on it but it is a lot less audible and has less amplitude. I think it might have been triggered..... but the exposure time of the camera caused that overlapping look on the photo.
 

EC8010

Ex-Moderator
2003-01-18 7:57 am
Near London. UK
Hello Bas,

When the scope is triggered, each trace perfectly overlays the previous one, and you could have an exposure time of hours without seeing anything untoward. It's just that that "hum" looks a bit too sinusoidal to actually be hum, and I'm wondering if you've got RF oscillation, which often then causes hum due to the excessive current drawn from the power supply. Have another go with the oscilloscope...
 
if you've got RF oscillation, which often then causes hum due to the excessive current drawn from the power supply.
I actually think that the humm is caused by excessive current drawn from the psu ...but somehow because of the little transformers? What I don't understand though is why I can't spot it on the scope.

[IMGDEAD]http://basenjes.de/tubes/images/aikido/slides/DSCN6389.JPG[/IMGDEAD]
 
Bas:
I am considering building an Aikido as a headphone amp, so I am not interested in having it hum.

I was able to look at the contruction pictures of your Aikido last night and I could not discern as to where the heater filament wiring ran in relation to signal and power supply in your construction. Is the twisted blue wiring filament? I thought I saw it sandwiched between two resistors at the socket.

When I went to look at it again the link did not work.

Cheers,
dr._sleep
 
Hi Bas. Still off topic (maybe). I think you are looking at the "safer version" example with 6n1p and 5687 tubes. That uses 220V and 430 ohm resisters. I wonder if 300V and 876 ohms (as in the first example he gives) isn't more appropriate for a 12sn7. I'm not familiar with ECC99 operating points. What is your operating current?
 
I was able to look at the contruction pictures of your Aikido last night and I could not discern as to where the heater filament wiring ran in relation to signal and power supply in your construction. Is the twisted blue wiring filament? I thought I saw it sandwiched between two resistors at the socket.

Yes it is the blue twisted wire.

When I went to look at it again the link did not work.
Mmm are you sure? Here is the link to it
http://www.basenjes.de/tubes/images/aikido/index.html
 
I think you are looking at the "safer version" example with 6n1p and 5687 tubes.
Correct

I wonder if 300V and 876 ohms (as in the first example he gives) isn't more appropriate for a 12sn7
You could be very right...I just knocked my stage together to prove the concept...I might build a better version up once I get it to quit humming. :)
 
Bas, how is the 1M resistor returned to ground? Does the ground return pass by anything that could induce hum into it?
I runs straight to earth...with theoretically no hum inducing wires...

Except maybe for this...the first cap after the rectifier is grounded on the exact same point...but what does that all matter because it is star earthing? Then again the currents might be pretty big whereas the current through the 1m is zilch...then again how does noise get through the 1m resistor to the cap?
 
Then take a look at some of the star ground photos in the Morgan Jones books to see the right way.
Bummer..just lent the book out to a buddy.

It doesn't look like a very good star from the photo, more like a small bus- if the 1M are the little light brown jobs, desolder the ground end, then reattach it at the other end of your mini-bus
Ok...will do...
 
No difference after resoldering to the other end of the bus..I move the 1m'S for both channels...and the results were exactly the same....with the noise on the other channel being a little spikey but lower in amplitude and on the other channel the big sine wave or hum...

I'm going to switch around the filament supplies..each channel has its own transformer now..not visible in the current pics..because swapping tubes made no difference either.

[edit] swapped the filament supplies..no change whatsoever..
 
Hi Bas,

You twisted the R and L hot input wires, which isn't good. Use one twisted pair per channel. Even better, use shielded cables for an enviroment as noisy as yours appears to be.

All your transformers are saturated, IMO. The filament transformer is IMO overloaded with 5 tubes.
If you are using two transformers to get the B+ high voltage, the first one should have higher wattage and higher secondary voltage. The resulting power should be under 50 % of the VA rate of the secondary transformer.

About the scope:

Your measurements look like an RF signal. For measuring 50 or 100Hz signal, you should set the time base to 1/50 or 1/100 sec to see one circle or 1/1000 to see more circles in one widow.
Also, set the trigger to AC and appropriate source channel if your scope has more than one channel.

Also, you should take care not to induce noise/hum by connecting the measurement probe to DUT. Connect the probe ground to the star ground, and (if possible) connect the earth pin from your IEC connector to the star ground.

Regards,
Milan
 

EC8010

Ex-Moderator
2003-01-18 7:57 am
Near London. UK
SY said:
Dumb question for EC- if this were an oscillation causing the power supply to hummmmmm, then why doesn't it show up on the anode side of the output capacitor?

Good point. The capacitor is a short-circuit at RF, so RF should show up at either end. Trouble is, even at 100Hz, hum ought to show up at both ends, and it also ought to have a much more ragged waveform. Something's not quite right...

Bas: I suppose you've tried swapping valves between channels just in case its heater/cathode leakage? Another possibility is that the hum is induced by leakage flux from the mains transformers.
 
You twisted the R and L hot input wires, which isn't good.
Ok..did not know that..

The filament transformer is IMO overloaded with 5 tubes.
I've added another one for the other channel..so the problem channel is feeding 2 tubes now.


you should set the time base to 1/50 or 1/100 sec to see one circle or 1/1000 to see more circles in one widow.
1/100 would be .1 sec/div right?

Also, set the trigger to AC
Ok..have done so now..

and appropriate source channel if your scope has more than one channel.
It only has one channel..that is Y right? or should it go on X..x says trigger..