A Very Compact Hybrid Amp - Page 86

Whole lotta testin' goin' on

With some certainty, from further testing, I've drawn these conclusions:

• Some tubes will hum to some degree, at least in CTH, regardless of heater supply implementation and regardless of their heater draw.
• The primary cause of hum/noise at issue here is backfeed/line-in A/C EMI by switcher to common A/C supply line.

Here's what I did:

1. Disabled on-board SM heater circuit in my CTH. Done by removal of D1H & D3H (LM2595 DC in) & removal of L1H (as well as its unraveling destruction
2. Ran external wire for CTH heater pins (#4 & #5). Re-cased CTH to move it away from assorted bench EMI (fluorescent lights, etc)
3. With 4 D cell heater supply 6922 has NO hum. 6BZ7 has faint hum.
4. Used P2P POC heater circuit to power tube heater (same components as on-board CTH heater)... When CTH A/C + outboard SM heater circuit are fed from same A/C supply get good deal of hum. Feeding CTH & outboard SM heater circuit from separate A/C supplies (adapters) - very, very faint hum... Not easily distinguishable from ever-present low-level hiss/white-noise amp exhibits (even w/battery heater supply, w/R18s shorted & D2000s)

Note: While "Feeding CTH & outboard SM heater circuit from separate A/C supplies (adapters) - very, very faint hum..." I carefully waved the outboard P2P switcher supply very close to the CTH PCB & very close to the tube & this did not elicit noise, so I no longer suspect general EMI radiation from switcher as cause of noise. Ground (plane) contamination is also unlikely the cause (based on this test & wiatrob's raised switcher coil results).


My goal now is to get the hum level at or below the ever-present hiss level, IOW not easily discernible from it, for tubes that have no hum w/battery heater supply - e.g. my Mullard 6922).
I've a few line-in EMI filtering components in my parts graveyard to experiment with (to inject between outboard switcher diode bridge & its CTH-shared A/C supply/adapter). And that's what I'll be doing as time allows.

I have to say, i dont think i cn live with mine with the level of hum i am experiencing. now i have noticed it...... i listen to it. Most people may not even notice it but for me i cant stop listening to it.

So, i may recase the amp, and build a 6.3v heater circuit for it, running from its own supply. I'll buy a small transformer, and spec it for an amp so i can run most tubes.

I am going to hold tight for a bit though, with the amount of brains on this problem, someone will find a solution i am sure.

I react this way to annoyances too, so know where you are coming from... E.g. when we had our 25yrold MBR bathroom renovated, it came out gorgeous but I couldn't stand it until the drip from the shower head was fixed.
Best plan of action I think. I, for one, am not near giving up on this one.

BTW, AFAIK w/o SM heater circuit of some kind the CTH could not have happened, at least not near its target form-factor. Think the (linear) 24V reg gets hot throwing off 8 volts or so in heat imagine trying to throw off 26 volts linearly to get 6V heater

for sure, we wont give up until something is sorted, and with the 12au7 its really not too bad.

Here's a wacky thing. The audible noise when I was doing scope tests last night was lower than I remembered, even with the 6H30...

I just got a 6CG7 clear top here an it sounds fine - I have my hands cupped hard over the cups of my RS-1s and I don't hear any hum...none. If I flick the tube I can hear the ping, but it's really silent - even though the 'scope indicated i was seeing similar AC spikes as adamus (lesser intensity as I have a bunch more capacitance on my heater supply).

I am going to curl up in a quite corner of the garage tonight and do some purely subjective tests...

wiatrob - Yes, you raised C3H as I recall. And as my testing indicates, primary noise/hum is backfeed from switcher, makes sense that raising that cap's value had an effect.

When I get a chance I'm going to try:

1. raising C3H as you did. Perhaps 220uF -> 470uf. Perhaps adding a low uF cap as a sort of snubber.
2. introducing resistance prior to switcher to lighten its duties.
3. investigate EMI filtering (chokes, etc) between switcher & common A/C supply.

In that order.

Just thinking out loud, would a ripple on B+ cause hum? Has anyone looked at the B+ waveform loaded/unloaded? Of course, make sure your oscilloscope can handle ~100V inputs (or AC couple it).

Is VG stable compared to SG?

And is it recommended that SG point be connected to anything? I've read that people haven't noticed a change when grounding the pot. Is it worth grounding the case to SG if you are using plastic ends? I just ask this as I get noticeable hum when I am using my y1 DAC with a DC plugpack, although I know this is a seperate issue, but not sure how to resolve it.

SG should be grounded to the case: http://www.cavalliaudio.com/cth/imag...ingDiagram.jpg

The y1 issue may be a ground loop?

Well, it ain't scientific, but I did some 'normal' listening tests this eve - with the Senn 600s, relaxing in the quietest part of the house.

All 6.3V:

6CG7 Clear top - as reported today, silent. Very nice tube too.

6922 BB Hum in right, low but noticeable, varied by tube.

6H6P - noisiest, low but noticeable. In LEFT

6H30 barely audible (have to press phones to head), not distracting

6N1P Sovtek - silent(?!)

6H23 Reflektor - Hum in right low but distracting

Could triode imbalance play a factor? Is everyone else hearing this in the left ch only?


Raising the value of C3H did lessen the ripple. I have a sideways 1000u in heatshrink wedged in there now - so my amp is not really like the others anymore. I did get a 470u to fit nicely with some bent leads.

holy crap, I just got a siemens E82CC in the mail. It's tiny and sounds spectacular, not even warmed up fully yet and it kicks the previous Sylvania toob in the nads

This amp just keeps getting better... I'm in love

Mmmm, bassy, sparkly, smooth, crisp and engaging all at once.

I just got a Siemens for my EF1, but I may be trading the EF! away. Was gonna sell the new shrink-wrapped tube, but now I wonder if I should save it for a rainy day.

I think you should sell it to me cheap very quickly before changing your mind... it's the right thing to do

But what if I get a CTH? I probably need something to use my Mullard long plate and shirt plate, although the pair of Amperex Bugle Boys might be staying with Blutarsky. I did post in the EF1 review thread that I would take $40 shipped, then started to think it might be a mistake. The plan was that my ALO Amphora was to replace my EF1 and other compact amps.

A true addict will buy an amp just to fit the tubes that he already has...

You don't want a CTH, rubbishy little thing..

Solution for noise/hum in CTH related to heater supply - 330uf C3H?

Nice to hear the love, which I know is out there, while focusing on the tiny imperfection
Right now its like I'm using a magnifying glass on Cindy Crawford's mole... Focusing on that & trying to "fix" it.. So its nice to be reminded of the big picture

Have to credit others for this but adding 330uf to C3H looks to be enough to virtually eliminate the noise/hum contributed by SM heater circuit.

Adding more than 330uf yields no perceivable benefit - adding less than 330uf does not get full benefit. For 6.3v 600ma load at least.

My bet is that *if* hum is noticeable for a tube with 330uf added to C3H, that hum would be noticeable w/CTH heater circuit disabled & use of battery supply for heater. And any such hum may just be due to A/C proximity in CTH.... Perhaps unavoidable without drastic measures (mechanical shielding of A/C area incl all diodes?)

Here's scoping of A/C supply:



Left: clean A/C as from adapter w/no load of any kind.
Middle: A/C supply with heater circuit attached supplying 6.3V 600ma load.
Right: A/C supply with heater circuit attached supplying 6.3V 600ma load with 330uf cap in parallel w/stock C3H

Edit: I don't think we'll derive any audible benefit by further "cleaning" switcher's backfeed beyond raising C3H to 330uf, at least for stock target CTH heater loads.

Hunch is that removing 47uf C3H & somehow fitting 330uf (35v) worth of capacitance there will satisfy when 100R or greater R18s are used. 100R->150R R18s are needed anyway to give vol pot usable range & to mask OB hiss/white noise.

Please use a low-impedance 330uF cap rated @ 35V or better & lead-length as short as possible.

Edit: I'll be boxing up my larger CTH w/onboard heater circuit re-enabled & 330uF 35V low impedance cap C3H to complete validation of this suggestion.