How to make a dead quiet PS for tube ?!?

[fikri]

Many thanks i give to all of you guys with all your replies. For me, all of your replies is a precious info for me. And for JC, can you proof it ?  :?

I used to have a BIG caps on my PS (around 470-560 uF) for my first tube project in order to eliminate hums, when i realized i was doing way too much for just a simple preamp, it was too much low and began to sound farty. And my next question is, do we still need a choke when we have a big uF caps ?

[puretube]

maybe the initial question of this thread shouldn`t stress the dead quietness of the PS, but of the circuit/build as a whole?

(what is the "little demon" that can`t get quiet?)

[fikri]

Thanks Puretube my friend. I stressed out the PS word because it is the biggest source of hum. And thats why i call it "little demon"  
Well, at least now my project is far away quiet, it is beyond my expectation ! A big thanks to R.G !  :wink:

And Now This thread is done.

[TryingToDo613]

That was actually part of my point puretube.  There is so much conflicting information on the web that the only way to really see what works is to do it yourself. For the part time hobbyist like myself, with three kids under three, experimenting is nearly impossible. If you ever want to just play what you build you just have to come up with a plan and do it. Just in this thread you have many man opinions, and even someone completely refuting RG, who isn't wrong, ever.  He seems to think that the old brass plate that corrodes was a good idea instead of a failed experiment.  He hasn't even watched the weber videos we've all watched.

I've spent alot of time over at the hoffman board and he advocates only a grounding buss accross the back of the pots and says that most amps that came into his shop with grounding problems were star grounded. That is from pure experience, and he's no slouch either. So there you have something that *isn't wrong*, conflicting with a technical explanation of ideal grounding that *isn't wrong*.  

That's why I said that I'd like the hedge of the filter caps to bail out a less than perfect layout and lead dress. There is even more conflicting opinions about those issues, like totally opposed too. The too stiff issue is another opinion. There isn't any sag when you have big caps for sure, but I think that's not an issue with single ended amps, where the big caps are the most effective, because in class A the tubes run full power all the time, or so I understand. -ph

[Dai H.]

fwiw, I don't think R.G. recommends the "pure star ground" (everything to one point) idea if you look at his article, the starring is done in sections in the Vox example iirc, so maybe there was a bit of confusion. I don't thing of star grounding as a method but more just a way of hooking things up. You could see bussing, starring, use of a ground plane, or whatever other ways in one single amp. Things will be far less mysterious if you take some time to study what is going on in the ground line and different grounding methods will make sense more as different approaches rather than "right" and "wrong" if you do.

BTW, nice article jc! Hadn't ever seen that one.

http://www.lynx.net/~jc/hum.html

[Eb7+9]

And for JC, can you proof it ?  :?

well, the most modern example I can find of this is in the JCM900's ... the preamp ground was bounce/raised using a 22-ohm resistor (with cap and diodes) between chassis and preamp ground lines - which are star'd to that point ... this was done in several high gain Marshall amp models ...

http://www.drtube.com/schematics/marshall/jmp50c1.gif

the non-orthogonlaity of the wiring in point-to-point tubes amps makes it impossible to prove anything - back in the day things were done with plenty of heuristics ... if you go loooking for all out provability you'll miss things that are near impossible to measure with home gear anyway ...

... my favorite example of playing with the non-idealities of older designs is found in the tweed Champ and Princeton ... I discovered that you can raise the common-mode reference of the filament circuit in these amps to a positive voltage using a resistor divider from the supply line which has ripple on it (I use 820k and 15k to give a few volts above ground) and, this is what's interesting, you can choose to tie down the bottom leg of the divider at different ground points in the signal path yield a better overall hum cancellation by tying  in one spot over others ... unless you can prove to me that the small amount of ripple from the supply line, divided down and fed to the filament common-mode voltage,  was responsible for cancelling hum through filament-cathode interaction (which is always possible) then I'm forced to conclude that this exercise shows that the circuitry is cancelling shared impedance currents in the ground lines ... I'm still curious to know what the real principle behind this fix is, nonetheless it does works - and it makes these SE amps even more of a joy to play than they already are ...

I'm NOT suggesting tube amps be designed this way - I'm just showing that understanding how each hum mechanisms works can lead to an overall remedy - esp. in older amps ...

again, I'm not discounting the usefulness of Star-Grounding in general - I'm just saying that it shouldn't necessarily be used in all tube amp designs as a rule ... I've noticed many modern amp designs that use a complete star ground approach where they also typically DC rectify the filaments - so there's no cancelling possible and no need  if you then shield the PT really well ... my thing is in being able to cancel hum in older chassis designs without reverting to those methods - which I don't disagree with per-se ... I personally prefer the challenge of playing with the basic elements that give rise to hum in the first place ... and it's way more ellegant in a vintage amp than sticking DC on the filaments and actively regulating the supply lines - which typically stiffens the response ...

... it's only my opinion that star grounding has turned into a blind myth in tube DIY land - like I said my views are mainly  based on my experience cancelling hum in older gear, and I've only ever seen one literary reference mentioning inserting resistance in the preamp ground as a means of reducing overall hum in common-chassis amp/preamp combos [I will try to locate it again] ...- the rest is hearsay from old-school audio guys I've known ... I don't think these should be taken as hard rules either way and that WAS my main point ... it's just meant as alternative info to arule that shouldn't be so ubiquitous as it has been touted in the last decade or so ... esp in DIY circles ... if you still get hum after start grounding your amp, you now have more of an idea where to look ...

R.G. I appreciate your respectful tone for a change - I'm open to discussing these things ... like I said, my opinions are based on my direct experiences and that's all I'm sharing here ... so again, I'm not saying star grounding is bad, I'm just saying that alone it will not guarantee hum free operation in classic tube amp designs running on AC filaments and a standard PI filter ... that there are other things you can do short of going all out with DC filaments and regulated (ripple-free) supply lines to elliminate hum in classic designs ...

I'll admit this is a difficult topic to argue over because we're dealing with chassis resistance, EM induction and the like ... definitely a dark area of amp design/building and worth shedding light on it ...

~jc

[Eb7+9]

fwiw, I don't think R.G. recommends the "pure star ground" (everything to one point) idea if you look at his article ...

sorry I missed this ... my understanding was that R.G. was advocating a true Star Ground system as THE right way ...

Thx for pointing that out to me ...

~jc

[Dai H.]

no problem.   I think you are completely right about how the whole "pure star ground", "true star grounding", or whatever people want to call it is not particularly a great idea (despite the many advocates). To be blunt, I'd say a dumb, inefficient, overcomplicated approach (the big mess of wires) which could end up giving MORE hum depending on how it's done. Interesting point about the hum canceling too. Hopefully in future, I can understand it enough to put it to practical use.

[Doug_H]

A few things:

1. As cd mentioned, use a high voltage regulator, but...

2. You may not want to do that if part of the desired sound/behavior of the amp is due to the looseness of the power supply (e.g. in a  "vintage" amp).

3. Star grounding cures hum problems due to the grounding scheme, but there are other sources of hum, i.e. unbalanced push-pull output stages where it is assumed that the p-p stage is providing some hum-cancelling that it actually isn't, etc...

4. Make sure what you are hearing is 60hz "hum" and not 120hz "buzz". "Buzz" is usually hashing noise produced by the rectifier. It can be exacerbated by lead dress, layout issues, and so forth. Snubber caps, FREDs etc can be used to help quiet that.

5. Incidentally, DC heaters increase the probability of "buzz" problems, due to the additional rectifier. IMO, pragmatically speaking DC heaters have diminishing returns unless you are building some mega-gain preamp. Elevating the filament ground reference is usually sufficient to keep the heater circuit from producing more than an acceptable level of hum.

Doug

[R.G.]

sorry I missed this ... my understanding was that R.G. was advocating a true Star Ground system as THE right way ...

It sure helps to understand what you're criticizing before you launch a polemic. I highly recommend it.

R.G. I appreciate your respectful tone for a change - I'm open to discussing these things ...

You sure do know how to make even a retraction sound like a dig, don't you? Wouldn't a simple "oops" have done?

Begin rant:
I was trying to be nonconfrontational. Would you like to make a few retractions about textbook EEs? As a gesture?

There is a quotation that I remember in several forms that I find quite useful. Some of the renditions are:
- The first rule of knight-errantry is that sometimes the dragon wins.
- The race is not always to the swift nor the contest to the strong, but that's the way to bet.
- From a Roman military philosopher: The gods favor the armies with the big divisions.

Literature all the way from prehistory down to today's TV and movies have made a lot out of the incidents when the trained experts are wrong. It's become a worn cliche that the formally expected result will be wrong.

Unfortunately, those are just the most interesting incidents because they are unpredictable. They appeal to the human need to feel at least equal if not better than the rest. An expert making a mistake makes the untrained guy feel much better.

The reality is that the formally expected result does, in fact, happen the vast majority of the time. If the experts were always wrong, there would be no point in being an expert no training to become expert and we'd all be untrained.

Beyond that, training happens in many ways. One way is through textbook instruction. Another is by laboriously collecting an assortment of scars to remind us of things that didn't work. One of these ways is a lot quicker and less painful, and we all choose which path we take.

End rant:

I'm not saying star grounding is bad, I'm just saying that alone it will not guarantee hum free operation in classic tube amp designs running on AC filaments and a standard PI filter ... that there are other things you can do short of going all out with DC filaments and regulated (ripple-free) supply lines to elliminate hum in classic designs ...

Yes. I wrote a whole article about that. It's one possible element in a whole regimen of techniques. Hey - wait a minute. I submitted a whole list of things to do, of which star grounding was only one. Maybe you could read what I wrote.

You could see bussing, starring, use of a ground plane, or whatever other ways in one single amp. Things will be far less mysterious if you take some time to study what is going on in the ground line and different grounding methods will make sense more as different approaches rather than "right" and "wrong" if you do.

Frankly, it's not too mysterious to me now. But people who are building an amp need "do this" kinds of instructions, not suggestions that they try grounding at many different points in hopes of finding somewhere that makes the hum less. Suggesting easter egging is what you do when you don't know what else to suggest.

I realize that this is somewhat foreign to you , but you might want to hunt down and read a textbook. It's called "Electronic Noise Reduction Techniques", and it devotes a substantial portion of the book to the different ways you can ground one device or a whole installation, makes note of how grounds are best served by wire grounds at audio and the slow transition to ground planes being the most useful at RF. Moreover, it discusses why. It should help with that learning different approaches stuff.

You could also see my descriptions of grounds in my posts here, including the analog of ground lines to an electron sewer.

Just to be helpful, here's the problem with hum cancellation techniques. Hum cancellation is quite old, coming from the days when people didn't have textbooks, and being part of the heuristics legacy. It works two ways: one is the obvious one where any hum currents you find happen to cancel whatever hum  you're getting. Given the nature of the hums available in a tube amp (and there are several different KINDS) this one is of limited usefulness. Keeping ground returns off the chassis is very helpful in keeping chassis feedback and oscillation down, though, especially if you keep the output and power stage returns off the chassis.

The second and more prevalent way it works is by cancelling the hum induced in loops of conductor formed by the wiring. The transformers provide lots of magnetic grist for this mill. There are strong 60Hz fields available, and any loop of wire that's cut by a changing magnetic field has a voltage induced across it.  The voltage induced is proportional to the magnetic field strength and the area of the loop. (oops, there's that textbook stuff again!) The more loops of wire you have, the more different little sources of hum you get.

This is why easter-egging for ground points seems to work. It's not the chassis voltage drop of current returning to the power supply (unless you've made the mistake of using chassis as a heater return) that makes the hum. It's the loops of wire of various areas that are set up at different places and different phases. With a zillion places to modify, some ways of bending wire will have better cancellation than others.

You missed one simpler way of cancelling hum: Put a pot across the filament winding and ground the wiper. Now you can dial in a little hum in either direction to make it less.

The problem with all hum cancellation techniques is that sometimes they un-cancel. That's the primary reason the hum cancellation pot disappeared from guitar amps.

Cancellation is what you do when you can't do anything else. Making the sources of hum smaller is the rational first step if you have the ability. F'rinstance: use toroidal transformers if you can; they broadcast less hum to start with. OK, if you're working with existing transformers you can't do that. So you can use an aluminum chassis to not "conduct" the M-field to sensitive circuits. OK, if you're working on an existing chassis, you can't do that. So you make the layout better by getting some distance between the hum sources and the sensitive preamp stages by putting the power transformer and AC wiring as far from the inputs as you can get them. OK, if you're working on an existing amp, you can't do that. So you can rearrange wires to keep the hum pickup down. Finally, you do hum cancellation for what's left.

I'd be happy to discuss this at some length if you like.

[puretube]

some times I use these for quieting down things...

(esp. when "touch-" or "proximity-" sensitive)

[Dai H.]

You could see bussing, starring, use of a ground plane, or whatever other ways in one single amp. Things will be far less mysterious if you take some time to study what is going on in the ground line and different grounding methods will make sense more as different approaches rather than "right" and "wrong" if you do.

Frankly, it's not too mysterious to me now. But people who are building an amp need "do this" kinds of instructions, not suggestions that they try grounding at many different points in hopes of finding somewhere that makes the hum less. Suggesting easter egging is what you do when you don't know what else to suggest.

um, fwiw the above wasn't directed at you R.G. (I'm sure you know what you are doing!), but just a general statement. And I'm not quite sure where anyone recommended guessing at ground points but maybe I missed something.

[TryingToDo613]

Yup. I'd love it. And also where I can get torroid trannys in the US. Nobody seems to know. And also where you get them thar 47uf poly caps. Want to build small quiet reliable amps, like really small and really quiet and really reliable. This is a pedal forum I know, so I'll also mention that I put pedal circuits in my amps right in the circuit, with a true bypass switch, (and next time I'm trying that millenium instead of the 3pdt).  Pedal circuits in amps need it really quiet because they pick stuff up and amplify it easy. This is enormously helpful. -ph

[puretube]

http://diystompboxes.com/sboxforum/viewtopic.php?t=32861&postdays=0&postorder=asc&start=45
Peter?, Z, Gez?...

(I can imagine 2 or 3 people, who know...)

[TryingToDo613]

ZV answered that I can buy them from Amveco, by the thousand. Their website has just wholesale distributors. with all the toroid stuff I see on this board and the ax84 board there must be someone selling onesies and twosies. -ph

[R.G.]

It's no mystery. You can get toroids from Amveco, Plitron, and Toroids Inc., in ones. Theyr'e expensive that way, and you may not get exactly the voltages you want, but they have a pretty good selection of stock items.

Here's a secret that doesn't get around much. Modern toroid winding equipment can make toroids at very nearly the same cost as EI equipment. They don't HAVE to be more expensive, and they are getting cheaper all the time. The toroid makers are just enjoying the reputation for expensive as long as they can.

[TryingToDo613]

Great, thanks. I'll call tomorrow. -ph

[puretube]

yes - no mistery: all information is stashed in this forum - for me, I`m out of time doing a search...

the difficulty in manufacturing are the small ones, with their very thin wires with a lot of turns for the high voltage windings...

BTW: Harmony-Central didn`t have its "search"-function working last week: total chaos  :shock:

[Doug_H]

some times I use these for quieting down things...

(esp. when "touch-" or "proximity-" sensitive)

Ton,

I always thought those were supposed to be most useful for RF, instead of audio (?!?).

I have tried them before and so far they have not had any effect in my builds. (again ?!? )

Doug

[fikri]

Strange, but it wont work for me either !  :lol: