|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
67.189.165.161
In Reply to: RE: You may have made a mistake with your modelling posted by Garg0yle on January 26, 2015 at 12:15:02
Can you post a screen grab of both circuits under test?
dave
Follow Ups:
Dave,
I love your subtle way of calling bullshit.
I guess that backfired huh?
hows the bass on the 2a3 amp? :)
Excellent. I assume you haven't followed the Garg0yle thread, but increasing the cathode bypass and ultrapath brought back most of the transient response I lost when switching amps. Warmed up the mid range a bit too.
604E in 16 cubic foot boxes with a 30Hz -3db, yielding 95db with 1 watt @30Hz with no EQ. Life is pretty good.
How about you? ;)
No I hadn't followed.. I'll check it out.
I'm still DOA after a lead out surgery incident killing a nice power transformer. :)
l-critical 6v6pp is fun though... enjoying that for now!
(Dave) Sure.I took the BBProof power supply and removed basically all of the inductance, series resistance and capacitance.
The sims were intended to be dramatic, to visually demonstrate the ill effects of such power supplies.
These sims use an ideal battery and have no speaker isolating the issue to the B+ configuration.
A more detailed model or real versions will have added issues as the effects of back EMF from the speakers will also interact with the B+ voltage as well as wild current swings in the choke(s).
Edits: 01/26/15
more resolution please.
dave
Are you asking to purchase the rights to my Harmonica power supply schematics?, because I already have interested parties.
On a serious note, you appear to be knowledgeable in spice, it appears that you host a spice forum.
Perhaps you could share a simulation of positive attributes of the LSES rather then trying to pick apart my amateur simming capabilities?
Perhaps I am being defensive, I'm just not sure what your angle is.
No angle.
you post a picture of ugly results. I want to see the circuit (with legible values) that produced those results. Even better you can simply post the .asc so everyone else can "play along". I have built several different "monkey" type amps and have never seen the behavior you show so I would like to dig into it a bit more before forming any opinions.
dave
OK no problem.I'm not sure how to attach the .asc but for expediency here is the screen shot.
The choke is 1H/10ohms.
For comparison I used a 270u ultrapath with a 330u cathode bypass in series and added 147 ohms to the battery. The choke is 7.5H 110ohms. The rest basically version 1 of the BBProof schematic that Paul Joppa posted.
Keep in mind this effect is exaggerated with the small caps for (50u/1H/5u) demonstrative purposes to show why less is not more.
I wouldn't think this is attributed to monkey style amplifiers alone.
Edits: 01/26/15
I think it is an ultrapath thing. Make the top 10U cap 100u or .1u and the squirrels go back into hiding.
dave
Indeed the .1u does fix up the waveform, however severe compression is still evident as it only swings ~125V as opposed to the ~176V in the higher induction/capacitance sim.
Thanks for going easy on me, that wasn't so bad. lol
How can you be sure there is 'severe compression' rather than variation of gain with frequency due to a partially bypassed cathode?
I suspect your results are that of a poorly applied ultrapath circuit and have nothing to do with power supply behavior.
dave
So why not take it out then?
The following is a 700mH/15u/1H/15u vs a 50u/7.5H/150u. The cathode bypass/ultrapath connection was removed. The 6J5 with 390ohm resistor is attached at the midpoint of the voltage dividing resistors used for the meter.
The compression is still there. To be more precise, the compression is formed relative to the ringing after a transient. As you can see below the when the Red B+ starts to wiggle it is starts to pull the 2A3 up and drop it in sync with the wiggle.
IMO this is not an ideal response to a benign sine wave, this will definitely colour other frequencies with a musical source that references that jiggly B+. Bass notes and transients will have a subdued attack with a bloated decay compared to a stiffer supply.
That is the dynamic compression I am talking about. To be dynamically correct, that first sine cycle needs to be full size instantly, no more, no less. You can't make drums pop by bloating the decay.
Now look at the second chart that has the stiffer supply. The sine has nice level peaks right from the first cycle. That is what a snappy transient response looks like.
To what extent this happens in practice, would depend on how extreme one went combined with other variables.
Impedance dips in speaker response make this compression much worse with the LSES, compared to a stiff supply that is hardly bothered at all by such dips.
I feel this effectively communicates my point in a illustrative manor for reasonable people.
I would still be genuinely interested in your simulations that show positive attributes of the LSES.
More resolution please.
I cannot make out the connections at the cathode of the 2A3 and the plate choke. (where is the AC ground?)
As transients go, I wouldn't rely on a 30hz sine to tell me anything about transient response. Try a square wave... or better yet a musical signal.
It is also helpful to run a DC operating point sim (.op) and an AC sweep to make sure the circuit is functioning as expected.
dave
"More resolution please.
I cannot make out the connections at the cathode of the 2A3 and the plate choke. (where is the AC ground?)"
I think the drawing and my description shows the connections clearly.
"As transients go, I wouldn't rely on a 30hz sine to tell me anything about transient response."
I disagree, if it can't handle a sine in high fidelity, what makes you think it will get better with a complex signal?
A clean sine lets you reference the attack of a monotone against following waves.
"Try a square wave... or better yet a musical signal."
I agree.
"It is also helpful to run a DC operating point sim (.op) and an AC sweep to make sure the circuit is functioning as expected."
At this point my circuit is so chopped up I'm not sure what is to be expected.
So I will ask, for a third time, for you to share a simulation that exudes the virtues of the LSES operating in the environment you deem suitable.
If not, I suppose I will have to write you off as just another huckster hawking $50 resistors. Can you sim that?
"We make no promises, but the 65% credit to future purchases applies if returned."
C'mon, show some conviction! No return policy because "Our resistors are so awesome no one would dream of returning one!"
Sims provide us with a lot of data. Interpretation of that data can be open to preconceived ideas. All sorts of conclusions can be arrived at. If the model isn't performing correctly, there could be a whole bunch of reasons why.
From the sidelines, I think Dave is trying to point you at getting the thing to run correctly in the first place, then you can alter things to see what effects your idea of an LSES supply will do or not do.
It is not up to Dave to provide you with an answer or a sim to prove something about LSES. He is however helping you to sim the thing properly and so guiding you to view a bunch of different, possible conclusions about the data you're looking at.
Slagging off his business approach doesn't seem to me to be good manners for someone helping you. You only have to look at his record to see where he's coming from.
cheers,
Stephen
"Sims provide us with a lot of data. Interpretation of that data can be open to preconceived ideas. All sorts of conclusions can be arrived at. If the model isn't performing correctly, there could be a whole bunch of reasons why."
Really? I did not know that. Thanks.
Despite the attempted revisions, the negatives of low inductance and capacitance are present in all the models.
"From the sidelines, I think Dave is trying to point you at getting the thing to run correctly in the first place, then you can alter things to see what effects your idea of an LSES supply will do or not do."
No, that is just what you think.
If he wanted to show the effects of the LSES in what he considered a good simulation, he would have provided that, the first, second or third time it was requested.
Instead he would rather insinuate this and that, hoping I would get tripped up in my amateurish ways.
"It is not up to Dave to provide you with an answer or a sim to prove something about LSES. He is however helping you to sim the thing properly and so guiding you to view a bunch of different, possible conclusions about the data you're looking at."
Yes it is up to him. In previous thread, simpler sims showed the effects are the same, this already has already ruled out the specific amplifier circuitry.
"Slagging off his business approach doesn't seem to me to be good manners for someone helping you. You only have to look at his record to see where he's coming from."
You are pretty naive if you think his intentions were to help me rather then protect the mystic nature of some particular businesses.
You are pretty naive if you think his intentions were to help me rather then protect the mystic nature of some particular businesses.
sorry if it makes me such a horrible person to want to see if the claimed cause and effect actually holds together. Plenty of pictures of the effect have been provided but I am still awaiting a good example directly linking it to the cause.
The only way I have been able to get simmed circuits to exhibit squirrelly behavior is to intentionally (or unintentionally) break them... feed a resonant circuit with its resonant frequency or apply some frequency based positive feedback.
Dave
No Dave, I feel your posts intentions were to instill doubt and contention into my little demostration.
The claimed cause is the low induction and capacitance in a power supply. Nothing new here.
Sorry if I am such a horrible person for wanting to see if the claimed cause and effect of the LSES actually holds together.
However you chose to stonewall me despite saying you wanted to play along. You took your simulated ball and went home.
The LSES power supply resonates in the audible region, which was demonstrated previously in a thread, which I think you were involved as well.
That thread also showed the same results with a with a different model, posted by a different member. It was inspiring and exciting to say the least.
The wiggly power supply is the frequency based feedback you are looking for, it's right there in front of you.
I was able to demonstrate, albeit a little exaggerated to make it more visible, something that I have heard over the years.
What's your problem with that?
While one might think they are good in the capacitance department, real world devices are not ideal, which is why one might benefit for a little more L and C then an ideal model shows.
I find the ear is quite sensitive to this compression, small amounts are audible. I find you can hear it at much lower levels then my simulation shows. The simulation was intended to zoom in and get some insight as to what mechanisms make inadequate supplies low fidelity.
Will more L C make all amps sound better? No, only the ones with inadequate power supplies.
Are you guaranteed to like the sound?
No, as has been mentioned before by many, some people may like the sound of a inter-modulation distortion and harmonica bass.
My personal preference or benchmark for a second capacitor in Pi filters is 450uF for ~300v B+, ideally comprised of 3 parallel 150uF caps to lower ESR.
This is suitable for amplifiers in the 1-5 watt range.
Somewhere along the line good engineering practices gives way to economics and now fashion and vanity.
Seems now some people would rather admit to using an undersized film cap but would never want to be seen using 150uF electrolytics. lol.
Function follows form for some.
They say measurements lie and to trust your ears, not to trust engineering.
I would love for them to try build an amplifier using unlabeled components and unknown voltages.
However you chose to stonewall me despite saying you wanted to play along. You took your simulated ball and went home.
stonewall? I simply asked for a legible drawing so a conversation could continue.
dave
"I would love for them to try build an amplifier using unlabeled components and unknown voltages."
I mean what else can you say when LowMu refers to a scope as an "osillyscope". I can back that up with a post if needed.
Now as a youth in the 1970s I too built many audio amps, both tube and solid state without a scope. Reasons were economic. So I've been there too! But since the early 80s I have never built any project without the aid of a scope. And the results are always well worth it.
You seem to have some misguided notion that someone owes you something. Good luck with that.
It's nice to see you sharing your "knowledge" of sims for everyone to see. Carry on.
Holy smokes Stephen R, that is a very delusional view of what actually happened isn't it?
lol
Well no actually.
Look from a previous thread, you seemed to be here for the right reason. Now you're being rude to someone trying to help. Now you're being rude to me. I really now have no idea why I bothered. You've clearly made you mind up already rather than being an enquiring mind; my mistake.
I'm not deluded thanks. But hey, you think what you want.
I'm sorry but from the above image I cannot see the little "square" that denotes a connection when lines cross which i why I asked for a better image.
Where does the plate choke connect and what is the AC path to ground for the cathode of the 2A3?
dave
I just think it's silly how some people here believe that potato farmer or what ever he is along with a retired commodities trader have discovered a revolutionary new technique in tube amp power supply design.Do we really think the masters of the golden HiFi era, Marantz, Hegman, Coderman, didn't tread here before?
Edits: 01/27/15
That seems like a rational objective response. Dennis and Jeff rarely provide any data other than trust your ears. I have no idea if two people hear exactly the same and data to support what you hear, is not a bad thing. The best part of stereophile is the measurement section which often corresponds to what the listenener reported.
.
Lets have a look at one of those impedance dips that one might get with some types of speakers speakers.It's pretty ugly, the attack is only 60% of the decay in the LSES. This will really subdue percussion etc which have all their energy at the start of the transient.
The stiff supply is shrugs it off, maintaining nice flat peaks.
Edits: 01/27/15
EDIT: You don't have to add the 147ohms to the "good" sim battery, as it handles zero ohms almost as well, just slightly more compressed on the first couple of waves.
Hey Dude,
You are using 15 uF as C2?
Thats funny, for 9 years now I have been showing 47 to 50 uF as being the C2 value on this Forum, never varies.
But the Fit Hits The Shan, when we get to A-B amps with different supplies, full range, on music !!
Jeff Medwin
.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: