• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

DC compensated SE output transformer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Re: Ok what about parafeeding??

Gluca said:
quite a simple and experimented way to get the OPT free of the DC.

Ciao
gianluca


True. But, the inductor version of the patent omits the blocking cap. I don't pretend to know whether or not that's a big deal. But it seems to me that smaller differences are celebrated here frequently. Something a bit different to try. As Kuei pointed out, might be a good way to use pp opt's for a single ended project. Diy'ers gotta like that kind of thing.

Sheldon
 
quite a simple and experimented way to get the OPT free of the DC.

Sure, Gianluca

but now we have DC current in the choke and this may give us
saturation problems similar to the previos ones.

Sorry, like the compensating current winding the peramnent magnet is quite an old idea.

Yes, moreover the compensating DC current need some
constant current source and the problem is only shifted.
Maybe a SolidState CCS will do a fine Job but it is not very
elegant.

On the contrary I think that the "permanent magnet" idea,
old or not, it is a good one. If it will work, naturally.

There was a time in the first half of the 20th century when magnetic amplifiers were the vogue, these used a controlled saturation transformer leg with a variable DC winding to control ...

I am just now reading a manual on the subject
very intriguing.

Federico
 
OT: Hybrid push pull

Maybe a SolidState CCS will do a fine Job but it is not very
elegant.

Maybe it should have some interrest to feed the secondary winding tru that SS CCS. Less voltage but more current needed.

Just a thought !

Hey, waitaminit !
Why not a push pull with a valve feeding an hi Z winding and a SS feeding a lo Z one !

So silly !

But remember where you've heard about that for the first time :D

Yves.
 
Re: Re: OT: Hybrid push pull

Kuei Yang Wang said:
Konnichiwa,



Indeed. And how about sending this SS CCS'ed current through the heaters on it's way to return to the source....?

Sayonara

Mmmh, let primary Z be 4K, secondary being 8, and the DC in primary be 50mA, we need near 1 amp in the secondary to compensate.
At 5 watts the secondary AC voltage will be 18 vollts pp.
Allowing 2 volts headroom for the CCS, we need at least 20v DC.
That means 20W wasted :D

Of course, efficiency of tubed amps is not an issue !!

Yves.
 
Recap - what is the objective?

Single ended amplifiers have some interesting characteristics and of course some fundamental limitations. To some extent that which makes them sound good is closely related to these same limitations. i.e. if you fix their problem(s) you may lose what it is you like about how they sound.

The push pull amplifier is quite an elegant fix for most of the SET non-linearity and transformer saturation issues at the same time. True they do not have the same sound but at least part of the SET sweetness is an artifical addition of even harmonics. There are good solutions for the infamous phase splitter issues and the newer so called "super triode" push pull seems to offer many of the triode benefits with the addition of low output impedance, increased damping and higher power levels for a given triode.

So I ask what is it you want to accomplish?

1. Cheaper tranformer same watts:)?

2. Less distortion:angel:?

3. Experiments for the fun of it:smash:?

4. Don't know, just expanding knowledge base:whazzat: ?

5. Trying to make the simplest amplifier less simple:cannotbe:?

6. No you missed it completely, I'm trying to.....:D
 
Re: Recap - what is the objective?

Konnichiwa,

hermanv said:
True they do not have the same sound but at least part of the SET sweetness is an artifical addition of even harmonics.

That statement, as repeated illustrated by psychoacoustics AND by adding such harmonics to PP Amplifiers is plainly and completely false, prooven so.

Make an amplifier that is Class A1 push-pull and offers the same power as another SE amplifier using a pair of output valves of the same type and two SE transformers (secondaries in series, PSE is a truely terrible idea) of 1/2 primary impedance of the PP Transformer.

Compare the measurements and do suitable level matched listening tests. Invariably the PP Amplifier will sound subjectively more distorted, not less and that for good reasons. The same phenomena is also observable with agressive interstage cancellation of even harmonics in SE Amplifiers. The Amplifier measures mush lower distortion as THD level, but subjectively sounds much more distorted (and it MUST do so, invariably, based on the way huma hearing operates).

Sayonara
 
Re: try also this:

plovati said:
http://www.audiocostruzioni.com/forum/topic.asp?TOPIC_ID=1159

permanent magnet instead of addictional winding. It is an original (to my better knowledge) idea and should works.


it does work, the problem is the the thickness of the magnet adds a sizable airgap to the magnetic circuit and in the case i tested a much smaller airgap would net the same offset.

I have some .008 neodym sheet magnet and i placed it in an 80% nickel core, it offset about 15ma of current but the .016 gap gave the inductance a pretty hefty hit. With a .016 gap alone the core would handle 60ma before saturation.

so my conclusion was unless i can find some much thinner and stronger material than the neodym sheet i have, the airgap from the thickenss of the magnet is far more effective than the magnet itself.

this all assumes the perm of the magnet to be near that of air.

dave
 
SET distortion.

Sorry Kuei Yang Wang; Single ended triode amplifiers produce by nature high amounts of even order products, they do and since they do, it does affect what you hear.

At least one commercial pre-amp offered a knob to intenionally introduce these products to allow for adjustable sweetness. I think your psychoacoustics folks are the same ones that prove again and again that none of us can hear any of these differences as PROVED by double blind testing.

In one sentence you say that its been proved that we can't hear this distortion and a few sentences later you propose an example of two out of phase single ended amplifiers with the outputs in series. This to prove that human hearing can hear the difference and does in fact respond differently to distorion products from one push pull design to another and that instruments respond differently. Which is it, can we or can we not hear these things? You seem to want to have it both ways.

Now that is not to say that you shouldn't buy whatever it is that pleases you and it is also true that measurements are a poor way to predict how a given amplifier will sound but what is not in doubt is that distortions of 2%, 5% or substantially more are common in SET amps.

Our hobby has it's fair share of nonsense including the opposite extreme, the "all amplifiers sound the same" mantra. They don't and the reasons are not fully understood. Measurements have their place and many excellent first class amplifiers measure very well.

I have not heard all possible SET amplifiers but from what I read they have a "signature" sound. To what do these psychoacoustics folk attribute this phenominum? It is not sufficient to simply deny something, another hypothesis should be offered that is believed to be a superior answer.
 
Re: SET distortion.

Konnichiwa,

hermanv said:
Sorry Kuei Yang Wang; Single ended triode amplifiers produce by nature high amounts of even order products,

Absolutely agreed.

hermanv said:
they do and since they do, it does affect what you hear.

Equally agreed. HOWEVER, if you produce a LEVEL playing field for PP vs (P)SE you find that all else sufficiently equal the (P)SE Amplifier for ANY given level of power producess LESS AUDIBLE distortion than the PP variant.

We may still argue about the overall relatively high levels of distortions in either circuit, however no-one prevents you from applying tons of negative feedback to reduce this distortion to very low levels.

If you wanted you could build a driver using an Op-Amp with a suitable High Voltage Mosfet in tow and multiple nested feedback loops (Or the kind of circuit in the "Evolve" 300B SE Amp), with the final one from the Anode of the SE Valve stage. This way you could easily minimise any THD.

But the argument you presented was SE vas PP and you claimed that the "large amount of even order distortion" made the SE Amp sound "sweeter" than an otherwise identical (eg no NFB etc) identical PP counterpart.

That argument is fundamentally flawed as due to the masking phenomenae in the ear the distortion produced by a suitably equal PP Amplifier is more audible (and also more objectionable), DESPITE being of materially lower value. As remarked, the same can be demonstrated using SE amplifiers with agressive distortion cancellation.

hermanv said:
In one sentence you say that its been proved that we can't hear this distortion

I stated NOTHING of the like. I did state that your claim that the "SE Amp Sound" was a result almost entierly of the presence of high levels of even order distortion as such was false. And it IS false. I did comment on relative audibilty of distortion elsewhere however and yes, given sufficient SPL levels (> 90db appx.) even as much as 5% 2nd harmonics is easily FULLY masked by the human auditory system to the point of being INAUDIBLE

hermanv said:
and a few sentences later you propose an example of two out of phase single ended amplifiers with the outputs in series.

Nope, I propose to use the two SE Amplifiers IN PHASE in series. The reson is to avoid the problems incurred by paralleling devices with materially different characteristics (curves) without local degeneration, in other words this is about solving the "PSE" sonic problem.

hermanv said:
This to prove that human hearing can hear the difference and does in fact respond differently to distorion products from one push pull design to another and that instruments respond differently.

Actually, it prooves absolutely nothing, as you completely failed to uderstand what I proposed as test and why.

hermanv said:
Which is it, can we or can we not hear these things?

What is audible strictly in terms of HD and what nor, with music and other stimulae is a well researched subject.

hermanv said:
You seem to want to have it both ways.

Nope, I merely insist that your statement: "at least part of the SET sweetness is an artifical addition of even harmonics." is demonstrated false. Here is why:

1) It is proven beyond any resonable doubt that odd/higher harmonics of HD have a much greater audibility than even/lower order ones.

As far back as the 1950's a weighting system was proposed that accounted for the audibility and variably suggested from around 12 - 18db more audibility per step up in the order of the harmonic while others suggested a system where the linear value would be multiplied with the square of the order of HD, in other words 2nd*4 of it's value, 3rd*9 of it's value, 4th *16 of it's value and so on.

A more precise and modern is offered by the GedLee distortion metric.

2) If you construct a platform that allows a test that is sufficiently free from differences OTHER than methode of operation; that is PP vs. PSE (where PSE is operated thusly to avoid the devices "current hogging"); you will find that the PP platform prudces materlially LESS 2nd & 4th order HD components than the PSE platform, but materially HIGHER levels of 3rd & 5th order HD.

In other words, the type of distortion produced by the otherwise equivalent PP stage contains lager amounts of higher order products than the SE Stage which in turn are more audible and are heard as more unpleasant.

ERGO, the SE Amplifier (compared to an otherwise equal PP device) does not sound sweeter because it has MORE distortion, but because it's distortion, irregardless of the Amount is LESS AUDIBLE, than that of an equivalent PP device.

And that was my point.

If you wish to claim that compared to that fabled beast, the distortionless amplifier, a zero feedback valve amplifier (PP or SE) has large amounts of distiortion and that these depending on circumstances may lead to audible alterations in the sound I am completely in agreement. It is a matter of record.

Though there are levels below which certain distortion spectrae (which one may term ear-congruent) despite high actual THD become audible if the SPL is sufficiently high. That too is a matter of record.

Therefor the blanked claim "SE Amp's sound sweeter because there is an artifical addition of even harmonics", which is what your statement claims, is false.

hermanv said:
I have not heard all possible SET amplifiers but from what I read they have a "signature" sound.

Have you HEARD this signature sound?

(BTW, my experience is that there is no such thing - SE Amp's using drastically different topologies sound dramatically different, much more so than traditional PP Valve Amps and/or Solid State Amplifiers by a LARGE margin)

Have you practically researched the influence of circuit topology on harmonic spectrum and attempted correlation with percieved "good sound"?

Or are you theorising simply based on "reviews" in the usual rags?

hermanv said:
To what do these psychoacoustics folk attribute this phenominum?

Which pehnominaon? That of the dramtic differences in subjective percieved sound among SE Amplifiers having differing circuit topologies? Or the supposed and much claimed but AWOL "signature sound"?

hermanv said:
It is not sufficient to simply deny something, another hypothesis should be offered that is believed to be a superior answer.

The answer is actually quite simple. Just look at the armonic spectrums of different "high distortion" Amplifiers printed in Stereophile (a Print at 1KHz is preferable though). This tells you the majority what you need to know, reading a little Crowhurst, Olson and Shorter on the topics of distortion will help to interpret these curves. This is modern science after all and not voudoun.

Sayonara
 
Higher order distortion products

Although it is certainly possible to build one and two tube transformer coupled designs with different high order distortion products it is not intrinsically caused by the fact that one uses two tubes or a center tapped secondary. Pentodes can (not must) produce higher order harmonics more readily than SET amplifiers but this is not guaranteed. Your microwave oven uses a triode.

It's just not that simple. Many factors enter this equation: Older slower tubes vs newer faster ones can effect the harmonic orders. Unbalance in the transformer or the notorious phase splitter are good candidates for many of these problems. The fact that the push stage may require a different feedback waveform than the pull stage needs can contribute to the high order products and the fact that typically a push pull amp has a broader high power spectrum means that the high order distortions may be delivered to the speaker more effectively.

The "signature sound" that I have heard and seems to match the description that most reviewers give to SET amps in general is one of apparent reduced high frequency energy and reduced impulse response, also common in the reviews is a less taut bass response. High order products can and do mask high levels of detail, the two are distinct and different but not allways obvious on short listening sessions. On the other side SET amps are usually credited with a midrange sweetness and accuracy that other amps can only wish for. One danger is that many other pieces of electronics do increase energy in the critical 2 to 5 KHz range (especially badly done CDs) and that an amp that dips or at least doesn't add to the problem in this region may go a long way to compensate for this most painful sound.

I know that human hearing can mistakenly perceive more high frequency content, whether as more gain or as more distortion products as more detailed and I am familiar with being fooled (but not for long) by this phenominum. Careful listening reveals the distinction between roll-off in level and reduction in higher order distortion and the converse of boosted highs vs. too many high order products. I have heard systems with over hyped high ends that did not fatigue although most systems with this flaw do cause listening fatigue.

It gets much harder. Local feedback is generally regarded as superior to global feedback, less feedback is usually credited with sounding "better". Push pull amplifiers carefully designed to minimize high order products will need less feedback so they may have an advantage. SET amplifiers with good bass control may need to be 5 or 10 times more powerful that equavalent bass control in a PP design. So there are always tradeoffs. It is dificult to avoid an apples and oranges comparison. With speakers of 90dB SPL a good 100-200 watts is needed to even approach live levels of intimate music much less brass or piano or worse yet a full symphony orchestra.

The SET offers a simplicity that makes sense for home builders, it is harder to get it completely wrong and less equipment is needed to fix a flaw.

Most of the SET amps provide from 5 to 35 watts. To reproduce at "home listening levels" a symphony orchastra using speakers of conventional efficieny will need more power than this for reasonably sized rooms. Yes I know more efficient speakers exsist but hey too have their limitations. Driver manufacturers do not intentionally make their drivers un-efficient but they do strive for flat and low distortion. This usually results in a driver of 85 to 90 dBSPL.

Klipsch was fond of proving that with his fairly efficient speakers it took an amplifier rated at 1000 watts to make an audience in his living room unable to hear the differnece bewteen his grand piano and a recording of the same piano.

So accross the board comparisons are most dificult, I believe a given topology will have advantages at a given power level but even then some careful work can overcome these differences. If one topology was intrinsically superior all amps would probably be made that way. Certainly for high end systems, parts cost doesn't seem to be the limiting issue.

The point I have been striving for is that it is the details not the topography that eventually controls the limits of amplifier behavior and I am concerned by any implied reponse that says that a given topology can not be made as good. I note that the sources you cite do not mention over what power range their results were collected I doubt it was with a 500 watt SET. Conversly 10 watt PP amps are hard to find but one low power PP made by the Bottlehead people gets reviews of having similar to SET sound. (You do say that for any given power level but I suspect that a half kilowatt SET hasn't been built and if it were would it have that SET sound?)

I am out of time, it does seem that it is hard to conduct a discussion that covers this much ground in this way. If you wish to continue our disagreemnet, please be patient it may be days before I can respond.
 
Re: Higher order distortion products

Konnichiwa,

hermanv said:
Your microwave oven uses a triode.

Not quite. It uses a magnetron, a somewhat different animal to a Triode, probably a much better one, as it has fewer electrodes.

hermanv said:
It's just not that simple.

Hence my assertation that it was essential to keel the playing field "level".

That failing, I can make an Amplifier with a 300B SE Output stage that has a THD of 0.0025% THD @ 1W without problems. Now, what will THAT sound like (If I find the time I'll try).....

Evolve Power Amplifiers 300B

hermanv said:
Many factors enter this equation: Older slower tubes vs newer faster ones can effect the harmonic orders.

They can? You mean Valves that are only good into the Short Wave range change the harmonic spectrum in the AUDIBLE (and super audible) range compared to ones that are good for VHF if no negative is applied (or actually even with, FWIW)?

I don't think so. That is just pure bovine unspeakable.

hermanv said:
The "signature sound" that I have heard and seems to match the description that most reviewers give to SET amps in general is one of apparent reduced high frequency energy and reduced impulse response, also common in the reviews is a less taut bass response.

Interesting. I would not particulary associate either feature with SE per se. Of course an amplifier may be tuned to appear like this (even a class B solid state one at that).

The "less taut bass" is the result of combining incompetent speaker designs; meaning ones that require significant levels of electrical damping with amplifiers that refuse to supply such.

hermanv said:
On the other side SET amps are usually credited with a midrange sweetness and accuracy that other amps can only wish for.

Largely hearsay. I can actually achieve the same effects using a high feedback push-pull circuit operating in fairly "lean" class AB and EVEN in ultralinear or penthode mode (and if I may say, measuring quite fine below clipping in terms of THD, thank you very much, due to very large amounts of correctly applied negative feedback).

The trick is not as such in absolute THD figures at all.

hermanv said:
One danger is that many other pieces of electronics do increase energy in the critical 2 to 5 KHz range (especially badly done CDs) and that an amp that dips or at least doesn't add to the problem in this region may go a long way to compensate for this most painful sound.

Another solution is called "equaliser".

hermanv said:
It gets much harder. Local feedback is generally regarded as superior to global feedback, less feedback is usually credited with sounding "better".

Sorry, but there is no such thing as "local" feedback, outside the advertising copy. And less or more feedback (to be precise INVERSE LOOPED FEEDBACK often called Negative Feedback) as such again is not per se the issue, if we account for the fact that ILF acts as multiplier on the order of harmonics and the levels of upper harmonics (plus adding PIM for good measure) while reducing overall THD.

hermanv said:
Push pull amplifiers carefully designed to minimize high order products will need less feedback so they may have an advantage.

They will invariably show higher levels of higher order products (and more high order harmonics) than a (P)SE circuit executed to identical standards, bringing back my level playing field.

hermanv said:
SET amplifiers with good bass control may need to be 5 or 10 times more powerful that equavalent bass control in a PP design.

What you call "bass control" is better referred to as "damping" and also as "radiation pattern/room mode excitation".

Any transducer is a damped mechanical spring/weight oscillator near the lower boundary of response. I have extensively covered this subject including here at diya, as have others.

Further any transducer in an acustically small room will excite room mode resonances at low frequencies. Again, I have discussed this. Combine a transducer that maxmimally excites room modes and that offers poor mechanical damping and using an amplifier with a high source impedance may indeed lead to "poor bass control", simply because some mentally handicapped person choose to use a poorely designed loudspeaker, inapropriatly so.

hermanv said:
With speakers of 90dB SPL a good 100-200 watts is needed to even approach live levels of intimate music much less brass or piano or worse yet a full symphony orchestra.

I take you are attempting to say:

"With speakers of 90dB/1W/1m SPL a good 100-200 watts is needed to even approach live levels"

Not only does this require a qualification WRT room size, but further, if you actually apply this levelk of power to a conventional 2" or smaller voice coil diameter "High Fidelity" driver you have generated a most excellent compressor and distortion generator that makes the average "high distortion" SE Amp pale in comparison.

hermanv said:
Most of the SET amps provide from 5 to 35 watts.

Most speakers reach > 1db compression at 1W applied and exceed 1% THD (mostly 3rd harmonics) between 2 - 5 Watt. Any more power will merley increase distortion and copmpression to very high levels. A traditional moving coil speaker with a sensitivity that requires it to operate at levels significantly beyond 1 - 5 Watt for sensible near realistic levels is certainly not capable of any fidelity whatsoever, so why bother with such misdesigned speakers (short of fashion of course, but that concerns me not)?

hermanv said:
Driver manufacturers do not intentionally make their drivers un-efficient but they do strive for flat and low distortion. This usually results in a driver of 85 to 90 dBSPL.

Sorry, that is simply and demonstrably wrong.

Driver manufactuers that produce the kind of drivers you mention, by the very basic laws of physics make drivers subject to comparably high levels of distortion and compression. They continue to do so as it provides profit, quality is actually immaterial in the context.

hermanv said:
If one topology was intrinsically superior all amps would probably be made that way.

Well, this is actually the case. There is one type of amplifier that provides it's makers most reliably and predicatbly with the performance they seek, namely profit. It is called by Douglas Self the "typhical Class B/AB Amplifier" and is solid state. The sound quality of said device is relatively speaking sufficient that for most people with most commercially available speakers the very limited quality does not matter. Hence. 99.9% of all amplifiers made anywahere are like that. Of course, as Sturgons law suggests, 99.9% of anything is C.R.A.P.

hermanv said:
Certainly for high end systems, parts cost doesn't seem to be the limiting issue.

!!?? You mean in speaker drivers?

If I look at the usual unimaginative magnet structures, small driver sizes and general design it would seem that what matters most is to claim the latest greates "high Tech" cone material which is ever so much better than plain paper (NOT) while expending very little effort on the actual motor, which determines lienarity and compression. Sort of like making a fancy car from fancy materials and claiming great things for it and then fitting 1 liter economy engine....

hermanv said:
I note that the sources you cite do not mention over what power range their results were collected I doubt it was with a 500 watt SET.

Because, as it so happens, power is completly inconsequential, if the neccesary SPL is reached with low inherent compression and distortion.

hermanv said:
Conversly 10 watt PP amps are hard to find

They are? Funny, I made several of them so far. I usually come back to SE though.

hermanv said:
but one low power PP made by the Bottlehead people gets reviews of having similar to SET sound.

Well, funny enough you mention it, but actually, given there is no "set sound" (unless you mean the results from using incompetent designed speakers, which are unsuitable for use with such amplifiers).

But yes, I can make a high feedback PP Amplifier to sound in the essential qualities just like a good SE Amplifier, and that despite one reaching 8W with around 7% THD and the other 25...50W with < 1% THD. Which BTW again puts paid to the concept that "SET Sound" (such as it may be) is due to "added even harmonics".

hermanv said:
I am out of time, it does seem that it is hard to conduct a discussion that covers this much ground in this way.

Forgive me, but was it not you ho generalised and claimed that there was one main explanation?

hermanv said:
If you wish to continue our disagreemnet, please be patient it may be days before I can respond.

I am sorry, but in my view facts do not allow disagreement. The facts have been stated. You are free to like or dislike them, but they become nontheless factual as result.

Sayonara
 
DC cancellation and magnets

Wow, I'm going to clarify some concepts on magnetic DC compensation, and now this 3D becomes a PP/SE battlefield.

In my opinion Farber method of DC compensation is a good solution to use "junk-box" transformer to use as SE output tranny. Is it also possible to use some other DC current source and aux winding to balance the DC, for example usiyng the DC filament supply or the DC B+ supply (in this case with accurate balance can be wind over the output trafo the power supply choke to balance also the hum).

All the three method suggested by Farber works in achieve DC cancellation, Fig. 3 is not a parafeed so does not requires a cap.
Main advantages of this method over the parafeed are the lack of heavy costly inductance or power wasting solid state current source and the simpler frequency response adjustment.
The advantage of Farber method over push-pull is the lack of phase inverter stage, but this is a little advantages compared to simpler power supply requirement and increased power out of PP.

To design a transformer 'ad hoc' for DC flux cancellation is best to use the Fig. 1 arrangement, usiyng the aux winding made of thinner wire also as cathode resistance. In this case just one cap works as cathode bypass and ac short for aux winding.

The aux winding can be made very simple, not interleaved with PRI/SEC coils, then leave more space to active winding, decreasing leakage and loss. Using a high permeability core is possible also to have larger inductance and reduced dimension (low capacitance and DC wire resistance).
Winding the aux wire at the outermost layer is better also because helps the cooling.

In this case I think Farber arrangement has some not little advantages over a common SE transformer with gap. If one prefer to avoid the distortion at 0 flux, is always possible to made aux turn less than primary turns, polarizing the cores like the SE case but with much useful flux swing.

A further improvement along the road of DC flux compensation (not cancellation in this case) is to use a permanet magnetic flux superimposed to a standard no gap core. Maybe the idea is not new (nothing really new under the sun...) and I'm sure that in the tube era many people thinks about something like that, but the application not come. I think that with magnetic material common now and CAD available should be possible to find out a geometry that can works.
I'm not caiming any patent, I'm just wondering if somebody has tried also this solution.
Simply put a magnet in the same magnetic path of the ac flux does not work, because permaent magnet have mu like air, so it is like having a gap, with minor dc flux compensation that works only for low currents.

Is better to use a different arrangement (I'm investigating what) with magnet in a separathe path or use some special magnetic hard material (which?) that can be directly magnetized by some high current pulse thru the winding.

Piergiorgio
 
Re: DC cancellation and magnets

Konnichiwa,

plovati said:
Wow, I'm going to clarify some concepts on magnetic DC compensation, and now this 3D becomes a PP/SE battlefield.

Not at all. This is not about PP vs. SE, but about fact vs. fiction, the fiction being the bit where "SET sound sweet because of an artifical addition of even harmonics".

It is this highly inacccurate and fallacious argument regulary advanced by those who wish to appear extra pious, straining out the gnat (of the "high distortion SE Amp") and gulping down the camel (of high distortion, high compression low efficiency speakers) whole, including the bale of straw on the camels back and even to the last straw....

plovati said:
In my opinion Farber method of DC compensation is a good solution to use "junk-box" transformer to use as SE output tranny.

Agreed up to point. Thinking about the various methodes has gotten me to a design (which for now will remain outside the public domain as I will investigate patenting it) that offers a much superior methode.... ;-)

plovati said:
All the three method suggested by Farber works in achieve DC cancellation, Fig. 3 is not a parafeed so does not requires a cap.

I disagree, actually. Fig 3 is equivalent in many to Parallel feed and replaces the small value coupling cap with a large value PSU Capacitor.

plovati said:
Main advantages of this method over the parafeed are the lack of heavy costly inductance

In Figure 3 you need an inductor that is LARGELY of the same value as a parallel feed one, but it requires only 1/2 the current handling.

plovati said:
To design a transformer 'ad hoc' for DC flux cancellation is best to use the Fig. 1 arrangement, usiyng the aux winding made of thinner wire also as cathode resistance. In this case just one cap works as cathode bypass and ac short for aux winding.

I will see to try this eventually on a breadboard (I'm just getting swamped again with commercial work, just when I had gotten to the top of the previous pile!), but I still suspect that AC shortcircuit from the capacitor across the winding, when driven from a non-zero impedance will in effect kill the signal.

plovati said:
Simply put a magnet in the same magnetic path of the ac flux does not work, because permaent magnet have mu like air,

How about a Ferrite magnet?

Sayonara
 
Re: Re: DC cancellation and magnets

Kuei Yang Wang said:
Konnichiwa,
Not at all. This is not about PP vs. SE, but about fact vs. fiction, the fiction being the bit where "SET sound sweet because of an artifical addition of even harmonics".

It is this highly inacccurate and fallacious argument regulary advanced by those who wish to appear extra pious, straining out the gnat (of the "high distortion SE Amp") and gulping down the camel (of high distortion, high compression low efficiency speakers) whole, including the bale of straw on the camels back and even to the last straw....


I'm out of this fighting, better to open another 3D to avoid spread blood here!

]Originally posted by Kuei Yang Wang
Agreed up to point. Thinking about the various methodes has gotten me to a design (which for now will remain outside the public domain as I will investigate patenting it) that offers a much superior methode.... ;-)

In my experience, ad hoc designed Farber transformer can lead to some advantages in case of high DC current. Below 100mA the size reduction and electrical advantages of this solution does not compete with the linearization introduced by gap in conventional SE opt tranny. Maybe a combination of small gap and aux compensation winding will be the best.

Kuei Yang Wang said:
I disagree, actually. Fig 3 is equivalent in many to Parallel feed and replaces the small value coupling cap with a large value PSU Capacitor.
True, but is a cap that is always required, so no extra cap and less trouble to optimize the low frequency response.

Kuei Yang Wang said:
In Figure 3 you need an inductor that is LARGELY of the same value as a parallel feed one, but it requires only 1/2 the current handling.
I don't catch You. Are You thinking at a comparison at the same power level or the same tube inductance?

Kuei Yang Wang said:
I will see to try this eventually on a breadboard (I'm just getting swamped again with commercial work, just when I had gotten to the top of the previous pile!), but I still suspect that AC shortcircuit from the capacitor across the winding, when driven from a non-zero impedance will in effect kill the signal.
The same doubt has been raised in a previous post, but also the simulation confirms that there is no signal shorts. Remind that there is always the secondary winding and when You short the aux winding, the fem is balanced by the secondary.

Kuei Yang Wang said:
How about a Ferrite magnet?
Was my first design temptative, but when closed in a magnetic circuit the magnetic load line generally drops below the B-H curve knee, and the hysteretic loop with applied signal is large and distorted. The equivalent mu in this condition is very low, less tahn normal gap tranny. I'm not able till now to find out a reasonable combination of geometry and material that can give me 1Tesla induction on a square section of 1cmx1cm and permeability of at least 1000.
 
Re: Re: Re: DC cancellation and magnets

Konnichiwa,

plovati said:
I don't catch You. Are You thinking at a comparison at the same power level or the same tube inductance?

Compared to the same tube in traditional parafeed.

plovati said:
The same doubt has been raised in a previous post, but also the simulation confirms that there is no signal shorts.

I remain unconvinced.

I think from extreme points, let us assume an ideal, infinite value capaictor. We have a shorted turn.

If I took a normal transformer and applied a shorted turn, I'd loose any output.

If I wind the cathode winding using thin wire I would still find that the DCR of winding appears in parallel with the load.

At least that is all I learned about transformers during my EE courses and a lot of experience tell me. I'm willing to try though and be convinced of the opposite by empircial methodes....

Sayonara
 
Re: Re: Re: Re: DC cancellation and magnets

Kuei Yang Wang said:
Konnichiwa,



Compared to the same tube in traditional parafeed.

The DC current in case of Fig. 3 shares between the two primary and aux coils with the same ratio of the ohmic resistance of the two branches. Because the inductor has a DC resisiance too, the balance cannot be perfect. Proper design of aux winding turn and ohmic resistance of aux + inductor can be better than parafeed but I think Fig. 3 arrangement doesn't worth too much.

]Originally posted by Kuei Yang Wang
I remain unconvinced.

I think from extreme points, let us assume an ideal, infinite value capaictor. We have a shorted turn.

If I took a normal transformer and applied a shorted turn, I'd loose any output.

If I wind the cathode winding using thin wire I would still find that the DCR of winding appears in parallel with the load.

At least that is all I learned about transformers during my EE courses and a lot of experience tell me. I'm willing to try though and be convinced of the opposite by empircial methodes....

My view is simple: the conservation of power applied to a transformer with three windings requires:
V1*I1=V2*I2+V3*I3
if one coil, let's say winding 3, is shorted V3=0, so we have:
V1*I1=V2*I2
that is the equation of the usual transformer. The shorting of coil 3 affect the reflected impedance but no other, at least in the ideal case.
 
Re: Re: Re: Re: DC cancellation and magnets

I remain unconvinced.

Me too.

In your simulation, if you set your plate load DCR to 0, coupling to 1, and bypass cap to 1 (farad) you will not get any gain. (at least i don't)

Adding the DCR, assuming a K of less than 1 and a reasonable Bypass will get you gain in the sim, but is that the kind of load we want?

if you play with the coupling and bypass cap values you can get nice smooth response, but when you build the circuit you get quite different results.

While plaing in the sim land, its interesting what happens whan you remove the cap and play with the turns ratio between the windings :)

dave
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.