 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: Re: Experiments with "UltrPath" connection posted by Pär on March 06, 2001 at 02:15:39:
I thought that the idea of "ultra-path" (hyped up name as far as I'm concerned) was that the b+ to cathode capacitor was the value of the cathode cap divided by the gain of the stage. This way any anomaly of the b+ supply would be divided by the gain and fed into the input in anti-phase to cancel out any effect.It should (if done correcty) help reduce b+ defficiencies for example hum.
I don't think it would be very valid in a PP amp since these unwanted b+ caracteristics are already well cancelled if the OP stage matching and phase splitter are well sorted.
Chris
Follow Ups:
Hi Chris,
This explanation - sort of make the tube act both as a amplifying dev. and a (sort of) voltage regulator :-) - sounds like a nice idea to me.
Will have a closer look at it to furter improve B+ noice rejection in a riia amp... [A little calc-session ofthen helps me a lot to understand things :-) ] ...but I wonder if it makes "tube-spead" and ageing more of an issus.
I do love people who explains things theoretical, instead of just hailing it and giving a mythical explanation.
[N.B. Though this has not been the case in this thread. N.B.]/Pär
> > It should (if done correcty) help reduce b+ defficiencies for example hum.
I don't think that will happen under any cap choice. For cancellation to occur will require a negative feedback circuit or neutral, this is actually positive feedback. Any amount of capacitance will increase PS induced noise from what I can tell.Kurt
The explination for how the circuit works was described to me by thorston at the london show last year, I will try to relay my understanding of it.If the B+ rail falls, either under dynamic signal or ripple conditions, then an unwanted AC voltage will form across the OPT. To correct for this an error signal of amplitude reduced by the gain of the output stage and inverted in phase has to be fed into the grid of the output tube. A convinient way of removing the need to invert the signal is to add it into the cathode circuit so it is subtracted from the input. Since the AC conditions of the cathode are determined by the cathode by-pass capacitor, it is easy to inject this signal using a capacitive divider made with the cathode capacitor as the lower leg and an additional "ultra-path" capacitor connected between the cathode and the B+. To get the correct cancelation the AC voltage divider needs to have the correct ratio, the ultra path capacitor should therefore be the gain of the stage minus one times smaller than the cathode by-pass. Also the "ultra-path" capacitor should be of a similar construction type and characteristic to maintain this ratio over wide operating conditions.
If the value is severly wrong then positive feedback on the signal will occur, but since the dynamic impeadence characteristics of the PSU are frequency dependant, the amount of positive feed back will also be frequency dependant, causing a rise in the bass response.
HTH
Chris
That's a very good explanation. Thorsten is a smart man. Now I think I was wrong about the hum cancelling not being possible with "positive feedback". But I think there's positive feedback all the time, it's just that the gain of the positive feedback loop is supposed to be equal to one to get the proper cancellation, and for all frequencies. More than one and it's causing gain of the noise, less than one and there's less than perfect cancellation. It also explains why I got more hum on some amps than others and more buzz than hum on some - some frequency dependency. It also explains that "too much of a good thing" is possible and can cause some overshoot and instability.This is quite an interesting subject. I will see what else I can find out and get it optimized. I really don't know if the capacitor really needs to be perfectly matched, or rather if it's practical to hope it to match up real well across the band. I expect it to underdo or overdo and miss the mark. Like I said, this is tough to simulate and tough to measure what it does, and control it. I can "ballpark it", though, with a comparable cap of the best nominal value as Thorsten suggests.
Thanks for the valuable insight.
Kurt
- simplifying the AC current loop, not necessarily the hum cancellation effect? With UP (in a conventional SE "series feed" design) the AC loop closes back at the cathode, rather than going through the cathode R/bypass C and also the final PS cap. At least that is what I remember from one of Jack's pieces on the idea. Or perhaps I am remembering what Chris Beck said when he tried it, but I think Chris is probably paraphrasing Jack Elliano.-j
That's what I got as well - it skips the cathode cap, and PS cap. Any change in gain, hum, noise, etc. should be simple enough to measure, which would confirm / deny all the other effects.
BTW, there has been a little bit more coming out on vibration effects within electrolytic caps, esp. in PS duty. The internal plates can actually move in response to strong AC signals, causing nonlinear modulation effects on the power rail. Bypassing all this would seem to be the way to go.
nt
Apparently not everyone got the news...Not much I have seen written or discussed, with the exception of the Nichicon stuff...Have you ever done measurements?
I have a question, will this not increase loading of the cathode part of the cct, as by placing the cap to B+ you put the cathode to ac gnd. through the B+ filtering cap. Unless the final element in your B+ filtering is a very large choke or some sort of resistance(not including the parafeed choke.)If it is, then it looks like pos. feedback at work all right.
Heath
unless there is a choke or resistance as the last element in your b+ filtering.
Didn't have the schematic open as I typed the above post.
Heath
I don't know if it was Ultrapath, but I'm sure I read somewhere about a year ago about this method of hum cancelling in SE amps think it was VTV, and I seem to remember a rule of thumb, that you use a cap 1/3 the value of the cathode cap. Don't remember the article mentioning the extra dynamics you found, the emphesis was on hum cancelling.Regards
Paul Barker
Hi,
One third sounds right for a "ordinary" output triode?
Maybe someone could try a one to sixty ratio in a noicy preamp using 12ax7?
mvh /Pär
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
