Noisy DC filament power supply problem on an Aikido

[Captn Dave]

I just finished an aikido preamp and it has an audible hum that the circuit doesn't seem to cancel. I disconnected the filament and ran it off a car battery and it's dead dead quite. I scoped the DC power supply and it has more than a tenth volt of ripple. BTW, it's just as noisy with AC direct from the transformer.

The power supply is a set of schottky diodes in a bridge with a pi (CRC) filter using two 10,000uf caps and a .82 ohm resistor. The noise might be the ripple or it might be coming from the schottkies; hard to tell. PSUII predicts 34mv ripple and it's far from that.

Has anybody done such a PS with good results? Any ideas for a fix?

I could easily switch to a linear regulator but I'd first like to find out what went wrong.

Thanks

[Geek]

What tubes are you using? Maybe you have tubes with noisy h-k leakage?

Cheers!

[ikoflexer]

Sounds like a bad grounding problem to me. Check the ripple with a dummy load, on the psu output. Is there a reason for not using a larger R in your CRC?

[DualTriode]

I had like issues with application of the 9 pin Aikido PCB. Not with a point to point version. Could be the PCB is acting as a antenna and picking up the heater transformer / rectifier noise. There is a lot more power going into the heater transformer than the B+.
I am not quite sure what the direct cause was. It is dead quite now. I corrected a few things. The power supplies are now remote, in a separate grounded box. The PCB is in its' own grounded box. The PCB is grounded. The B+ power negative rail is grounded. The heater - is tied to B+/4 volts as shown in the Aikido users Manual. Ground is the U slot in the wall power outlet, the green wire or bare wire in the wall outlet.
keep one hand in your pocket and live

[Captn Dave]

Originally Posted by DualTriode

I had like issues with application of the 9 pin Aikido PCB. ... It is dead quite now.

You say it's dead quiet now. Tell me about your power supply. AC? DC?

BTW- I didn't say but mine is wired PTP.

[Captn Dave]

Originally Posted by ikoflexer

Sounds like a bad grounding problem to me. Check the ripple with a dummy load, on the psu output. Is there a reason for not using a larger R in your CRC?

Well, I've checked grounding carefully but I took a hint from your larger R comment and I've made some chenges. The transformer has a pair of 6.3 windings so I have wired them for 12.6 CT. Now the schotkky bridge is putting out close to 20 with half the current and I should be able to use about 20 ohms which should be an improvement. I'll also have a center tap for the 1/4 B+ hum circuit. I'll also rewire to get that B+ between the caps in the pi filter rather than from the clean B+.

I will need to rewire the tube sockets for series operation (I'm using 6DJ8 and 6CG7) and that will be a bit of a chore at this stage. I'll post when I have results.

[DualTriode]

Captn Dave
I do not understand PTP.
My B+ high voltage is a transformer fed 120 volt voltage doubler then a resistor (adjusted for 250 V DC at the Aikido input) into a capacitor, choke then a capacitor; a pi filter.
The Heaters are powered by a separate center taped 12.6 V transformer full wave rectifiers into a 12K uf capacitor followed by an adjusting resistor (set to 6.3 V DC) into the Aikido. There are no regulators used in this one.
DT

[alexg]

I just did an aikido phonostage with 12at7/12au7 for the first stage then riaa then a 5751/12au7 for the output.

B+ is around 300VDC using a CLCRCRC. I am using a 3.15-0-3.15 heater, AC heaters with voltage divider giving me 1/4 of B+ for the heater ground.

Negligible hum, I can hear the hum around a foot away from my Fostex speakers.

I just ordered a toroidal 12VAC transformer, 5A so I can try DC heaters to totally try to remove the hum.

But as it is, it is tolerable and is quite ok for a tubed phonostage.

BTW, the sound is excellent, better than most of the SS phonostage that I have tried.

[ChrisA]

Originally Posted by Captn Dave

Well, I've checked grounding carefully but I took a hint from your larger R comment and I've made some chenges. The transformer has a pair of 6.3 windings so I have wired them for 12.6 CT.

Now that you have 12V available you can use a _regulated_ DC supply. Just get an LM317 adjustable voltage regulator and set it for 6.3V feed the reg with your CRC filter but use an even bigger R so the LM317 sees about 10V on it's input. Look at the ripple rejection spec on the reg. DC should be as quiet as battery power. It is also easy and cheap to bypass the diodes with cermaic disk caps. They should be about 10x the diode's junction capacitance.

Should be both easier and better to use regulated 6.3V then to rewire for series heaters.

[Captn Dave]

OK, an update. The primary problem was that I was getting some AC ripple from the heater transformer CT onto the B+ circuit via the 1/4 B+ voltage divider. This I corrected with a .1uf cap to sink the AC ripple. That took care of 90% of the hum. Geek was correct about the h-k leakage on one of the tubes. Tube swapping lead to the bad tube and a swap left the amp quiet from about a foot distant. This is with AC heaters (I eleminated the Schottky rectified DC setup after a small cloud of noxious vapors reminded me of the folley of impatience).

I later ran it on a car battery and it was dead quiet. Mouser just delivered parts for a LD1085V based DC power supply and that is next. (BTW, John Broskie sells a complete kit for only $24US which I think includes a very nice looking circuit card worth at least that much. I think it includes the regulator, heat sink, resistors, diodes and caps for that price as well. What a deal. Too bad I didn't see that before I ordered from Mouser. Something to consider.)

The circuit checks out OK with all the voltages as they should be and I'm satisfied that there are no further build errors, but I'm not satisfied with the sound yet. It's very bright. I will hook up a signal generator and scope later to check the low frequency roll-off to confirm what my ears tell me is a problem. I'll probably need to make some coupling cap changes.

I'm currently using Sonic Cap .47uf coupling caps. I also have some VitaminQ of the same value that I can press into service. I'm thinking I may need more capacitance and I may try these in parallel as an experiment.

Any thoughts on the coupling cap issue? What values have worked best for you?