New Circuit, New Build, New Questions...

[xk49w]

Is there a risk with the paralleled diodes of some additional noise at the line freq?

[Rick]

I haven't investigated it that thoroughly since it just works for me, but since the current must flow through either one diode and then the other, the currents and voltages should be symmetrical unless I'm missing something, I haven't noticed any additional noise when using these. I haven't tried it on the ground center tap, it should work the same or better.

I've seen that some think the UF4007 has lower switching noise. Since we're just heating filaments I doubt that it will matter. They are glowing almost white hot and the transients probably aren't noticeable. Keeping tight twisting on the feeds and avoiding any looping around the socket will help with AC hum more than anything unless you run clean DC which will draw more current I suspect. Filament windings are usually run pretty close to demand so I stick to AC. I'd try the UF4007s, also I see that these are mentioned having as much as 1.1V drop (1N4007), so they might work out just right. You will have to try it and see but it should be better than running 6.8 or 7+VAC on the fils at any rate. Let me know how it works.

I'm having a hum/buzz problem with a liverpool based amp currently where I'm running a bridge recto on a 100W marshall style PT (running 6-EL84s to 100W Marshall OT). I've tried every trick I can find to cure it and I'm starting to think it's switching noise from the bridge. I'm using 1n5408s and plenty of filtering and good grounding I think. Maybe I have noisy preamp tubes or soimething else is whacked. I'm thinking of changing to 1N4007s and maybe putting some caps around the bridge as I've seen the ValveWizard shows this in his PS pages. It's the stubbornest noise problem I've encountered, usually grounding the cathode to the input jack ground cleans this up, but this bridge seems noisy for some reason...

[Structo]

I don't know that much but I would think you would be passing a lot of current through any diodes that are on the heater center tap.

On my little 6v6 amp I have a large 25w Zener on the HT center tap and that sucker gets hot!
Probably apples and oranges but I would investigate how much current is going to be flowing in that.

It might not be much because Fender used to have a 1/2w pot for hum balance.

[Structo]

I'm having a hum/buzz problem with a liverpool based amp currently where I'm running a bridge recto on a 100W marshall style PT (running 6-EL84s to 100W Marshall OT). I've tried every trick I can find to cure it and I'm starting to think it's switching noise from the bridge. I'm using 1n5408s and plenty of filtering and good grounding I think. Maybe I have noisy preamp tubes or soimething else is whacked. I'm thinking of changing to 1N4007s and maybe putting some caps around the bridge as I've seen the ValveWizard shows this in his PS pages. It's the stubbornest noise problem I've encountered, usually grounding the cathode to the input jack ground cleans this up, but this bridge seems noisy for some reason...

I see quite a few guys here use .01uf 1Kv ceramics across the diodes in the power supply to keep them quiet. I'm not 100% certain that is the value but I think it's right.
But that kind of switching noise is a lot different sounding than the 60 or 120Hz hum you get when there is too much ripple voltage.
More like a high pitched buzz or whine.

It could be a defective bridge recto or a grounding issue but it sounds like you have messed with that.
Sometimes these things can drive you crazy!

[Rick]

Structo, thanks for the input. I have put caps on the recto diodes to no avail. You are correct I think, it doesn't sound like switching noise as you describe it. It is 60 or 120 Hz hum sounding. Not meaning to hijack the thread so I'll keep it short or start a new one. I will try upping the V1 filtering and see where that takes me. This is a pretty novel design. I'm running 6-EL84s, Marshall 100W transformers and a 2nd channel with an EF86 input to a single tone control, then mixed with the unused triode of V2 with a common anode mixer scheme. The EF86 channel is very quiet which is where you'd usually expect the most problems, but it is entirely in the liverpool section where I hear the hum it seems, turning the EF86 channel up doesn't affect the hum at all, but the final preamp stages seem to be making hum, maybe V2. Reducing the 0.1 uF coupling cap to 0.02 uF to the phase inverter helped some, so it seems to be in V2 triodes area. I'll try upping V1 filter next just in case it's starting there.

[Structo]

Yes, from your description it does sound like power supply ripple.

It depends where your heater voltage is at but I have read about guys putting one or two 1N4007 diodes in series with a heater wire.

Is your amp fixed bias? I assume so since you mentioned Marshall.
If it was cathode bias then you could elevate the heater on top of that.

I can't state for a fact but I think there are other methods and tricks of lifting the heater voltage from zero potential.

I even read where a guy connected the shield on the input coax to the plate of V1.

So there are plenty of tricks but you kind of have to figure out where the hum is actually coming from.
Maybe chopstick around or start unplugging tubes to isolate where it is?

[Structo]

Not sure if you have read this Rick but this is a good paper discussing heater and hum.

http://www.freewebs.com/valvewizard/heater.html

[flood]

reading that resource does make me a bit worried that the currents flowing through the diodes might be a bit much for them to handle. UF4007 is rated at 1000V 1A only. the secondaries are rated at 3A. maybe it would be more prudent to put 5-6A diodes in there? the resource shows a 60s1, which has up to 6A current handling capacity.

please bear with me if i'm being a bit too paranoid my friend, the buyer,, lives in new delhi, so servicing later is tricky at the very least, and horribly expensive too.

rick, i'm assuming you have swapped tubes around? it could be just a bad tube. otherwise, i'd probably think that a heater wire is a bit too close to signal or HT carrying wires.

just my noobish insights - i've built 4 amps so far, all low wattage and low gain (except this one), and i'd say you're a bit more experienced and have probably checked for all the usual suspects already hope it gets sorted out.

[Rick]

reading that resource does make me a bit worried that the currents flowing through the diodes might be a bit much for them to handle. UF4007 is rated at 1000V 1A only. the secondaries are rated at 3A. maybe it would be more prudent to put 5-6A diodes in there? the resource shows a 60s1, which has up to 6A current handling capacity.

please bear with me if i'm being a bit too paranoid my friend, the buyer,, lives in new delhi, so servicing later is tricky at the very least, and horribly expensive too.

rick, i'm assuming you have swapped tubes around? it could be just a bad tube. otherwise, i'd probably think that a heater wire is a bit too close to signal or HT carrying wires.

just my noobish insights - i've built 4 amps so far, all low wattage and low gain (except this one), and i'd say you're a bit more experienced and have probably checked for all the usual suspects already hope it gets sorted out.

flood, if you're worried about current handling, then a 1N5408 should be plenty. The 1000V 1A rating is at DC, the 1N5408 is 3A @ 1000V DC and should be a good plenty at 7VAC. EL34s are filament current hogs stating 1.5A current draw, so even with an EL34 and 2 preamps running .3A per tube, you should be well with current requirements.

To go the resistor dropping method, add up total filament current demand. per Ohms law 1V drop/current draw will be the required resistance in ohms to drop 1V. e.g. 1V/ 2.1 amps current draw = approx. 0.476 ohms. 0.5 ohms should be close enough. 2- 1 ohm resistors in parallel would do it and give enough wattage rating if they are 1W resistors etc.

I have built 50 or so amplifiers in the last several years, I have never run into such a stubborn hum problem. The EL84s are all cathode biased just as in the liverpool running 3 separate pairs of cathode resistors/bypass caps. Resistors are all 130ohm. I'm using 200 uF of bypass, maybe that is the problem, I may try lowering the bypass caps to 100 uF or less, I'm not sure how much that affects lower frequency bypass regarding hum, but in general I do know that bigger bypass caps pass more lows, smaller ones will attenuate lows and boost highs, so this may be something to look at. I see that many designs use much less bypass capacitance and maybe this is the reason why. I have never had to elevate heater supplies so far and I'm hesistant to do it if it's avoidable. I will have to fire up the O'scope I guess and see where this hum is being generated or amplified. It may end up that I don't have any "low hum" tubes as my current stock of new spiral filament tubes is wiped out. None of my NOS AT7s or used AX7s help, but they may not be spiral fils, so I'm not sure about that. There is a steady low level hum with only the phase inverter in place, it gets much louder with V2 and finally V1 in place. The EF86 has no effect at all in or out, so it's almost certainly not to do with EF86 unless the common anode mixer circuit is somehow affecting it. One thing is sure, you learn a lot when you deviate from established designs and wing it with something quite different.

[Richie]

I'd say you have a ground problem.
I would suggest try moving the power supply to its own ground. When the amp is on, volumes down,is there a slight change in hum when you check the plate of the output tubes? If so you could have something added to the power supply gound.
Sometimes it only takes one wire,moved or seperated to a different ground location to either cure,or add hum. Might even try using a jumper on the grounds, clip it,and try moving to different locations,see if it changes.If it does change, something tied to that section may need to be seperated,and moved to a different ground spot.
That being a powder coated chassis, i'd make sure you have removed the coating to make good contact.

[flood]

One thing is sure, you learn a lot when you deviate from established designs and wing it with something quite different.

oh, most definitely.

FWIW, i have a 1500uF (!!!!!) 35V cap on my power tube cathode.

[flood]

guys, am i being unnecessarily paranoid about the EL34? my PT, as stated earlier, is 275-0-275 at 100 mA i'm worried that this is too close to max anode+Grid2 current.

my first resistor in the RC mains filtering is a 22 ohm 10 W (47u/47u). would it help if i increased this to 100 ohm / 10W to impede the current draw a bit and possibly bring down B+? a tube rectifier is no longer an option at this stage.

like i said, there are 5 cathode bias resistors that can be chosen from; 130 - 470 ohm, about 80 ohms apart on average, all rated at 5W. i was thinking 130 ohms 5W would be ok for the EL34, but am dreading the current draw.

i have some trannies rated for 18 watters - 275-0-275 150mA and 3.15-0-3.15 5A. i could possibly interchange them if i really need to, but would rather not...

so is the 100 mA sufficient? or would i need to up it? thanks again.

P.S. clips next week

[Rick]

"FWIW, i have a 1500uF (!!!!!) 35V cap on my power tube cathode."

Why such a huge cathode bypass cap? 220 uF is about the largest I ever see in vintage amp designs. From Valve Wizard Single Ended power section page:

"Cathode bypass capacitor: As with most single ended stages, fully decoupling the cathode will maximise gain by preventing internal feedback and so maximise output power and input sensitivity too (making the valve easier to overdrive).
For a low roll off of approximately 10Hz:
Ck = 1 / (2 * pi * f * Rk)
Ck = 1 / (2 * pi * 10 * 220)
= 72uF
So we would use 100uF. This is a pretty generic value for single ended amps and in many cases we wouldn't even bother to calculate it."

100 uF should be plenty. You can use 100 mA PT (what's the voltage?) but it will be a limiting factor for power output. Take a look at Merlin's page on SE design for some clues on all this. Draw some load lines and then you determine where you want to go with your voltages and bias point. You also don't want to be underbiased or run outside the class A center bias value excessively.

I usually start with a classic or established design that has all the basic parameters designated, then I just fiddle with the preamp stages mostly to add gain and tweak tone. It is very easy (although admittedly taking shortcuts) that way since I have established voltage requirements, component sizes, biasing, etc. already done for me and then I concentrate on the preamp design which is somewhat independent of all the power amp requirements and it's easy to get an amp working that way.

I think a lot of "amp builders, geeks, experts" or whatever, fail to recognize the importance of preamp tone shaping. They talk about power amp distortion a lot as if it's the main consideration in amp operation, IMO, the lion's share of tone shaping is done every where else but the final power section. There it is enhanced by power tube distortion and even seemingly little things like grid stoppers (which can actually have a dramatic effect) but a whole lot is or can be done in all the prior stages to power amplification.
http://www.freewebs.com/valvewizard1/se.html

[flood]

rick, thanks a lot for that link - don't know how i oversaw that, it's an excellent resource.... so it seems i'm safe using 100mA for now, since high power is not really a concenr - we wanted a low power amp for bedroom to small stage usage, the preamp is handling 90% of the tone - power amp saturation is not that big a concern right now, but i have left space for a VVR knob, in case we want to retrofit one sometime.

what i was actually worried about was the EL34 drawing more current than the PT could supply and causing a burnout or something of the sort.

so from what i understand, i should be safe with 100mA from he PT secondary. 2k2 5W on the screen grid and 130 ohms 5W on the cathode...

the 1500uF cap was used because i have quite a few of them lying around, and from what i know, huge bypass caps stiffen up the bias considerably, making the cathode biased amp approach fixed bias characteristics, increasing punch and doing away with the vintage characteristics a bit. so far, i'Ve been quite happy with the sounds and response, so i wonder if i'll change it back to a lower value...

[flood]

attempted the diode drop methods to no avail. initially, i tried antiparallel diodes from the center tap to ground, then two 3A diodes in series to ground but neither had any effect on the voltages - they're stuck firmly on 7.8V no load and 7.1V with full load (EL84).

i'm going to try the resistor drop method today and i hope that works. the buyer needs to leave for a posting abroad soon, so i have to finish this one ASAP...