Trying Dave's odd ideas about transformer coupling methods

lead boatanchors are best.

Moderator: Paul Barker

Post Reply
Paul Barker
Posts: 59
Joined: Sun Feb 20, 2005 10:20 pm
Location: UK

Trying Dave's odd ideas about transformer coupling methods

Post by Paul Barker »

Presently the 212 is it coupled from a px25 using AE bifilar wound IT's. When listening at home on low level I take the output from secondary of these IT's to an opt. I have some parafeed 5k double c cores.

Integral to the resultys is that for this part of the amp I have resistorless capacitorless power.

I tried hooking up as parafeed instead of interstage. It sounded worse, that is to say more distortion and less depth. Quite a reasonable difference, not that slight.

Could robbed of the advantage of removing the power supply caps from the loop the truth about parafeed is revealed?

A useful means to an end in some circumstances but not a good design goal?
IslandPink
Posts: 56
Joined: Thu May 12, 2005 8:01 am
Location: Denbigh, North Wales

AC/DC

Post by IslandPink »

Maybe you end up with the same issues I've got with PP , Paul - once you get rid of the DC current in the output transformer ?

Mark
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

isn't the output impedance of the noR noC PS for the most part resistive thus removing another reactive element from the mix?

i still feel the parafeed cap just does wacky stuff with the inductances around it. If forced to cap couple, I would look up my friend gary pimm for a CCS to remove the reactiveness of the load on the tube, and then see how bad it gets with the remaining CL... i'd also make sure theat L was as linear as possible (even though it lacks and dc i'd gap it) then if need be i'd load the L to control its reactivenenss.

Then of course i'd wake up in a cold sweat and realize the thought of me using both a resistor and a capacitor was all just a horrible dream.

dave
sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

Just thought I'd chime in a bit... In a no R no C supply, the low impedance is primarily resistive, the inductor reactance being the high Z. This contrasts with a "normal" PS where the low impedance is primarily reactive (the capacitor) and the high impedance is either reactive or resisitive. Notice that with a LC filter, looking backwards from the load, the LC is a PARALLEL RESONANT circuit. (This is nasty in that at resonance, the supply impedance is effectively infinite, which is definitely un-good).

The no R no C circuit has no such resonance.

One, er, could, er, draw an analogy between a, er, parafeed resonant, er, circuit and LC filtered PS stage. (standing aside and covering up)

:shock:

-Steve
Paul Barker
Posts: 59
Joined: Sun Feb 20, 2005 10:20 pm
Location: UK

Post by Paul Barker »

My power supplies for the 212 amp have thus far limited (masked and dirtied) the better sound I know to be there. Fortunately thanks to your old experiments Steve I have an alternative direction to try than arranging a remortgage for sufficient Black Gates. Not that a r less c less for a 212 output stage would cost any less than the present UK stock of Black Gates.

Well I'm not trying to make parafeed sound any good, the ccs way would be right in these circumstances if you were inclined towards it, but then the r less c less supply would be impractical (as if it isn't now). I'd have to spend my mortgage on a parafeeed cap of any quality that would cope with the voltages.

I have heard parafeed sound better than series feed on the same pio loaded power supply and parafeed was better, so I thought maybe the only real benefit of parafeed (apart from making the transformer winders job a doddle) was power supply related. Once that issue is dealt with no taste for parafeed.

Of course in a 833 or gm100 amp due to the dc currents it may again be of benefit, but the 212 sounds OK at 130mA.

The resonance of the LC power supplies is clearly defeating good base performance, when you hear the base from the r less c less isn't it Steve? Not that that is the only benefit, all frequencies are improved greatly, I most notice the extra sparkle on the top.

Waiting some Silver Gold Mundorf's for the Ariel Tweeter caps, that should complete the high end saga. Better had do.

Paul
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

sbench wrote:
One, er, could, er, draw an analogy between a, er, parafeed resonant, er, circuit and LC filtered PS stage. (standing aside and covering up.

-Steve
sure... and maybe every time you eliminate an additional resonant circuit you are better off. (assuming you don;t shoot yerself in the foot doing it) some will argue the parafeed circuit replaces the PS circuit, but that assumes the two NEVER interact. I 'spose thats the case for a first approximation so its gotta be true.

Along those same lines, i do not see the parafeed cap as a replacement cap, i simply see it as an additional cap. (and don't care what the ad copy... ummm i mean first approximation says :-)
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Paul Barker wrote: I have heard parafeed sound better than series feed on the same pio loaded power supply and parafeed was better, so I thought maybe the only real benefit of parafeed (apart from making the transformer winders job a doddle) was power supply related. Once that issue is dealt with no taste for parafeed.
my biggest complaint about parafeed is an amp or stage can sound really good in one situation, but god awful in another. its very tweekable, but if i wanted a tone control i'd just build one that behaved predictably.

as for the PS isolation, there are two approaches.. use parafeed to cover up a crappy supply or just build a better supply... seems you have come to a similar conclusion as me.
The resonance of the LC power supplies is clearly defeating good base performance,
also remember the cathode bypass (if any) works into this mix.
Waiting some Silver Gold Mundorf's for the Ariel Tweeter caps, that should complete the high end saga. Better had do.
is there any tweeter attenuation for the ariels??? (can you see where i am going??)

dave

Paul[/quote]
Paul Barker
Posts: 59
Joined: Sun Feb 20, 2005 10:20 pm
Location: UK

Post by Paul Barker »

I must be thick, I don't see where you are coming from.

The silver golds arrived today, just put them in, a 1uF as coupler betwixt ml4 and px25, then a 4.7 and 0.47 series cap for the ariel tweeter. the improvement over vintage pio bipassed with 0.056 teflon is only very slight, but it is now better, hopefully as time goes on bigger enjoyment will follow. This was a very costly upgrade, I imagin e more exotic transformer core material would be a better direction to spend the money. However this is an instant impression, maybe after a few weeks the caps will sound a lot better.

Paul
jarek
Posts: 62
Joined: Wed Feb 02, 2005 10:02 am

Post by jarek »

Hi Steve,
The no R no C circuit has no such resonance.
But nothing comes without a price - LR would tend to have lower efficiency in ripple reduction due to 1st order filtering.

3 phase a la Thomas looks like a perfect match, so 3 phase thyratron supply anyone?

jarek
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

or i am thick :-)

i thought the caps ordered were for the speaker x-over which prompted me to ask if there was a resistive divider to match efficiencies of the drivers.

In my defense, it is tough to follow what exactly is going on, and my memory isn't what it used to be.

i could look back and try to figure what your current setup is for the speakers, i remember tweeter amp being mentioned so you may be biamping, i really don't know.

I was assuming that your ariels were running full range and the tweeter cap was a crossover one. since there was discussion of a tweeter output with the nickel cores, i had the idea to just place the "tweeter output" at the driver in the form of an autoformer, then you could use a smaller cap (yeah) and a single well tuned amp to drive it all.

the two people i have coaxed into trying this have had good results, but its very uncharted territory. Luckily LTspice makes the design easy and will be a powerful tool in setting this type of system up.

take a look at how steve has handled his space. every topic has a new thread so individual ideas are easy to track. I think this is very important since new people coming in can quickly pick up on and keep track of the history.

at some point i want to have a clearer picture of where this section should go, but putting it down in words is difficult so maybe some random thoughts will work.

a great website is well organized and has a ton of great information that is easy to navigate. (steve's site works great for this) the only problem is its a ton of work to do, and very one sided. (much like a lecture form a great professor)

a great forum has a ton of great members and a constant influx of new people and ideas, but the signal to noise ratio gets high and the same topics get covered over and over with little progress. (i have found i have made the same post on several occasions years apart.)

i envision something half way inbetween the two extremes.

It seems the key to a good website is editing and organization, and the key to a good forum is anrachy. Maybe the inbetween ground is having a well edited and organized forum?

this is where you come in... you are the editor and the organizer. Its your system and you are definately thinking outside the box! I honestly believe if you keep it up, you will attract the right people to make us all benefit (and you are attracting the "right" people :-) I honestly believe if you keep it concise and organized you will have peers from all over the globe helping you build your system. The key here is they will not be doing it for money or ego, but because they might actually learn something from your experiences. (yes, maybe you become the guinea pig but the tests are rarely fatal)

dave
Paul Barker
Posts: 59
Joined: Sun Feb 20, 2005 10:20 pm
Location: UK

Post by Paul Barker »

"yes, maybe you become the guinea pig but the tests are rarely fatal"

One touch and I'm a gonner, but one good thing about -r -c power supply, pull plug and breathe a sigh of relief, all HT gone in an instant.

I understand what you mean about tweeter now. Just at moment the caps are to go in the standard ariel crossover. Thus far they weren't worth their money, exhotic transformers would have been a better investment, but maybe they will become stunning. I was expecting stunning out of the box, I must still be naiive.
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

nevermind.

i was thinking that the tweeter needed attenuation to match the bass, but it appears as if the two LF drivers in parallel pretty closely match the tweeter.

the idea of winding two discrete autoformers and placing them on the drivers is an interesting one though. essentially you could design an impedance match into the speaker crossover by placing an appropriate autoformer on each driver.

I suppose sending a few hundred volts AC though speaker cables would make some nervous though :-)

just another dumb idea for the scrap heap.

dave
Post Reply