What difference one capacitor cam make!

[3dzapper]

Over the weekend I reassembled my 300B SET amp. I was going to redo it into a 2A3 SET but I kept spending the money for a new power tranny elsewhere. I was listening to it most of the day yesterday. It had the SET sound clean mids and shimmery highs but "polite", no real slam and the bass just so so. And it had just too much hum for me.

This morning I took it out to the bench. My original plan was to rectify the AC filimant supply. I went off in the cold and snow to RS and bought a few power resistors and a rectifier. When I got home I realized that I was supposed to buy 2 rectifiers. Senior moment! So I said why don't I try something else.

Like most "home brew" 300B amps this one uses a large motor run dual section capacitor in the power supply. I measured the ripple and found it unacceptably high. In another senior moment I forgot to write it down. this PS needed more capacitance. Using a 10W 100 ohm resistor for a current limiter and to drop the B+ a tad, I installed a 330uF 450VDC Electrolytic cap off the second section of the OE cap and tied just the B+ lines to the 300B's into it. I left the input tubes connected as originally built so as not to change the voltages on them. I am running that 450V cap about 20V high but it hasn't burned up (YET!!!) The mod also lowered the cathode to plate voltage to 385VDC. It had been running at around 405VDC which is at the extreme end of a 300B's specs.

Now the good part. I was able to adjust the hum down to under 10 mV for the first time ever. It is still just audible with the Klipschorns but not as objectional. The better part is the bass is now THERE! The best part is that the mids and HF still have that "SET" clarity but now with SLAM! The overall clarity and soundstage has improved dramatically. Amazing what one capacitor can do!!!

Rick

[mike stehr]

That's a stereo 300B?

A critical well-damped supply is big thing with a class A triode amp.

You gotta have some capacitance for bass, but you can go too far and slow down the rise time.

Is that 10 millivolts, or microvolts? I'm at around 2 or 3 millivolts with my 300B amp, though I use 2.5 volt AC filaments.

Is that a 5 volt AC supply with hum pots? Maybe reference a couple of resistors to ground off each leg of the heater supply? I've had luck with non center-tapped 6.3 volt heaters doing this(without humpots). The value is not critical, as long as it's not over like 10k. I used 100 ohms, I'm sure you can tweak the value for lowest noise.

Additional de-coupling can be done for the driver stage plate supply, that could help kill more noise.

Download that Duncan amplifier's PS designer software.

Place in all the parts and values of said PS, figure out what load resistance the power supply is coupled to, and just use that last load resistor in the simulations.

You can adjust the parts/values, and watch the load line in the simulations for rise-time, overshoot, ringing...

Yeah, nothing can beat real-time bench testing, but the program is handy and can get you close enough to throw it together and confirm on the bench.

[3dzapper]

That's milli volts Mike. Good idea about Duncan's page. I use the tube data program regularly.

I think that next I will lift the drivers off the motor run as I am sure that that is where the residual hum is to be found.

The heaters on the 300Bs are each powered by AC from a separate winding with a hum pot. The ripple at the cathode is so low that my Fluke VOM says "000" so that is not the source.

Maybe I'll pull the driver lead with a 5 Ohm current limiting resistor and a 68uF cap. I don't want to overload the 5U4G with too much capacitance. It is a stout tube and will take a lot more input capacitance than I have also. I think I have 5uF, the 5U4G will take up to 40 at the input of the filter. The trouble is, like a Dynaco Mark III, finding a capacitor with a high enough voltage rating. My little pea brain says that I can probably use an SDS cap board for a Mark III in there from Uncle Ned with good results. I'll order three of them after Christmas.

Rick

[mike stehr]

Does the power supply have a choke?

Mine's a CLC power supply, and everything taps off the last capacitor. The driver de-coupling caps are parked near the driver tubes, 100 uF 450 volts, with a 10k 2 watt resistor. I had serious 120 Hz noise until I de-coupled the driver stage.

[3dzapper]

Yes CLC. I just think that the motor run cap has to go eventually. I did manage to get the hum level down to 5mV on the 4 Ohm taps with the Khorns. Almost inaudible from my listening position on the couch. Listening to Rachel Z's "Everlasting" right now. Even my wife commented on how good it sounds.

If you like jazzy piano by an obviously classical trained pianist, you'll like Rachel Z. Her rendition of Johnny Cash's "Ring of Fire" is reason enough to buy the CD.

Rick

[mike stehr]

I always liked that song, I'll have to check it out.

[ben.]

How do you go about measuring hum without a scope? Can you use a VOM? When my house is quiet enough to adjust my hum pots, I usually don't feel like fiddling. It'd be nice if I could just look at a meter. Can I just measure AC mV across the outputs of the power amps?

[3dzapper]

That's how I do it Ben. With no signal going in adjust for the lowest voltage across the output terminals. Probably the best way is to remove the driver tubes but, I found no difference in mine. I think the residual hum that I am experiencing may be in the EH 300Bs I'm using. I have to find another pair of tubes that I can afford.

Peter,Paul amd Mary sound good though.

Rick

[Deang]

What in the Hell are you guys talking about?

[ben.]

Thanks, Rick. PP&M? You ever see "A Mighty Wind"? High-larious...

[3dzapper]

----------------

On 12/20/2004 8:35:21 PM DeanG wrote:

What in the Hell are you guys talking about?

----------------

It's a secret SET code thing! You don't have one of the special decoder rings that came in the Applejacks box do you?

Rick

[3dzapper]

Ryan, Older analog VOMs where mostly 10K Ohms/Volt. My Fluke 73 (it's old but I made a lot of money with it) is 1Meg Ohm/Volt. I've never had a problem with power tranny interference but the readings can be unstable when reading the DC resistance of a transormer. Then you are correct a LOW impedance meter is better. I have a couple of those older ones. A heathkit I built back in the early Seventies and a lab grade Db meter. They are relics! They also measure RMS unless you have a new digital with a "peak hold" feature.

Rick

[mike stehr]

"If you guys want a real sensitive meter for measuring hum, get yourself a used VTVM (vacuum-tube volt meter) off of ebay, EICO or Heathkits work well. These are very high impedance meters capable of measuring AC Millivolts very accurately (much more accurate than the ear)."

Hell, I forgot all about my Hewlett Packard 400H VTVM.

The VTVM is handy device to have.

[3dzapper]

I've got one of those around somewhere. A Heathkit I believe that a previous owner put in a wooden box. Looks to me that the DC input goes through a series of known resistors to ground and the drop is fed to the grid. You are correct in that the VA-7 had an AC peak to peak function. The input impedance is fixed at 11MegOhms irregardless of voltage including the probe.

Rick

[mike stehr]

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=25411&item=3861305623&rd=1&ssPageName=WDVW

She looks a little battered, but the seller says it powers up.

There is a online manual for the HP 400H with schematics and calibration.

20 bucks is worth it to me, hell I paid 25 for mine. But it looks better.....

[Klipschguy]

Rick,

Hum is indeed annoying. It is amazing how a litle more filtration will quiet things down. How much extra capacitance did you add? As you know, too much filtration of the power supply can be hard on the rectifier tube.

I restore antique radios occasionally and one of the first things I do is replace the caps in the power supply to get rid of (or at least minimize) any annoying hum. I usually increase capacitance by about 15%-20% to take it easy on the rectifier.

Sounds like a nice amp. I bet those big glass 300's look cool glowing.

Best regards,

Andy

[3dzapper]

Andy,

I put 330uF on the load side of the choke. There is only 15uF on the input of the choke. The 5U4G will take a 40uF input capacitor and I would change that except I don't have anything with the voltage required.

Yes the 300Bs are a pretty sight all blue and orange with the lights out.

What do I need for 500VDC? Is 370VAC enough? That would be 523V P/P. I know 440VAC would be better but,

Rick

[mike stehr]

"What do I need for 500VDC? Is 370VAC enough? That would be 523V P/P. I know 440VAC would be better but,"

For a oil AC capacitor?

I'm using Mallory caps rated at 370 VAC. The input cap is 35 uf. The secondary voltage on the PS transformer is close to 700 volts AC total. I'm using a pair of 6AU4 damper diodes for rectifiers with the cathodes connected. I haven't measured the voltage off the last cathode of the damper tube before the cap, but the output cap in the CLC supply is right at 400 volts DC.

I really have no idea what the damper tubes drop from the HV secondary of the transformer.

I would guess that first input cap is getting more than 400 volts DC. I should find out...

But the amp and it's power supply have been very reliable.

The choke get's a tad bit warm after a full day of use, the caps don't even get warm at all.

I know the boils down to a simple formula converting AC voltage to DC, sumthin' with Root Mean Square.

Yikes

I think you be ok. Buy a extra and ramp it up to the voltage you want, and see if it pops.

[3dzapper]

RMs is .707 of peak AC voltage, conversly to get peak to peak, divide the RMS reading of your VOM by .707. A quick in the head division by 7 is close enough. Like 400VAC would be 570V P/P.

Oh yeah, another story. I tried DC to the cathodes of the 300Bs and it was the same. So I went back to AC and I rerouted all the heater wires. I got the residual hum down to almost a straight line on the scope on the right channel and a little ripple on the left (under .010V P/P). On the Khorns I have it down to 3mV. Not bad but......... I still could hear that hum in my room!!!!

The hum was is in a location between the speakers in front of the TV. If I move to the left or the right, it lessens. HMMM? "Must be a node" I say to myself. But!, when I walk over to the Khorns and stick my head in the bass horn, NOTHING!!! ARGHH!!!!! Wait a minute. I went into the back room by the furnace and corresponding to the plane where I hear the hum is the thermostat transformer and, it's HUMMING!! The hum I hear in that one spot in my room is being transmitted by the joist of the floor upstairs and radiating there!That's the problem!

In another project, I'll put in a new transformer.

Rick

[mike stehr]

Thanks Rick, I knew it was a easy conversion. I always forget half the time.

What I hate is out of all the amplifiers I have, the cheapest one is the most quiet. The Magnavox single-ended 6BQ5 amp. It's gotta be in the microvolts.

You have to kneel down and stick your head in the woofer to hear a wee bit of line noise.

And it's a CRC supply, EZ-81, no choke. 6.3 volts AC heated, with the referenced resistors to ground off each leg of the heater winding.

Maybe it's the low voltages involved with the amp.

The only other one that is almost as quiet, is the Magnavox 6BQ5 PP.

At least the hum problem is from your furnace, it did give you a excuse to improve on the 300B amp.