john curl said:
What criticism? 😱
I only stated that unlike transistors tubes don't care which frequencies how to distort, be it in Russia, or on the Moon, or outsourced to Chinda/Inia.
Traderbam knows well what I was talking about.
Hi Glen,
I don't know if you are arguing semantics here, or the design idea. If you buffer the Vas, your THD will drop as the load impedance goes down for bipolar output stages. The buffer reduces impedance variations that the Vas "sees".
It's old knowledge, been proved in practice and simulated to death. Whether this particular design has had this additional buffer simulated or built is unknown to me, but the effect is factual.
A "Blameless" is simply a straight forward design with care taken to avoid common problems. Some are based in layout. Nothing magic about the design at all.
-Chris
In this case, ostripper is correct.
I don't know if you are arguing semantics here, or the design idea. If you buffer the Vas, your THD will drop as the load impedance goes down for bipolar output stages. The buffer reduces impedance variations that the Vas "sees".
It's old knowledge, been proved in practice and simulated to death. Whether this particular design has had this additional buffer simulated or built is unknown to me, but the effect is factual.
A "Blameless" is simply a straight forward design with care taken to avoid common problems. Some are based in layout. Nothing magic about the design at all.
-Chris
Stinius, R is real, Z is a combination of real and imaginary. You know, Bybee devices and such!
Seriously, there is either cap or inductive added impedance and phase shift with R to make Z.
Seriously, there is either cap or inductive added impedance and phase shift with R to make Z.
Hi John,
Or to put it another way, Z=R when the phase shift = 0°
When important, Z could be expressed as x ohms at y phase angle with + or - to denote leading or lagging against current.
Real or imaginary would be a "rectangular notation".
You made me think of those classes. I tried for years to forget them! 🙂 I'll probably have a dream about this, thanks to you guys.
-Chris
Or to put it another way, Z=R when the phase shift = 0°
When important, Z could be expressed as x ohms at y phase angle with + or - to denote leading or lagging against current.
Real or imaginary would be a "rectangular notation".
You made me think of those classes. I tried for years to forget them! 🙂 I'll probably have a dream about this, thanks to you guys.
-Chris
Thank you.
I didn't want to argue, just learn. I won't use "EF'ed" anymore. 😀
That doesn't sound right.
If you are objectively comparing 2 things (amps or foods 🙂 ),
you can't say "that" amp or "that" food. (you have to throw a noun in somewhere.)
I have built both
the apt and leach with and without a emitter follower , so
I was curious about Glen's opinion.
He stated my apt was flawed.I just wanted a specific opinion
as to why.
Strange , Mr. holman ran his old apt that came back from
garage parties and other abuses and it spec'ed out perfect
after 17 years... so it can't be that bad..
OS
Hi ostripper,
You know, I have looked at the schematic of the APT. I think I had serviced one in the past. However, with all I have learned in the past few years, I would love to get one on my bench at some point and go through it.
With any luck, someone in my area may want their amp freshened up. I have no plans to modify anything, I just want to see how far I can improve the performance through matching and replacing any ceramic caps I find with better.
-Chris
You know, I have looked at the schematic of the APT. I think I had serviced one in the past. However, with all I have learned in the past few years, I would love to get one on my bench at some point and go through it.
With any luck, someone in my area may want their amp freshened up. I have no plans to modify anything, I just want to see how far I can improve the performance through matching and replacing any ceramic caps I find with better.
-Chris
john curl said:
So what? So I supported your heroic struggle for sound quality with no tubes:
http://www.diyaudio.com/forums/showthread.php?postid=1793057#post1793057
Why you John are always looking for criticism even in posts that support your point of view?
Below I have posted an example of one of the circuits I simulated which demonstrates the VAS fighting problem. At 20 kHz and about 27 V peak sinewave, the VAS current just softly cuts off.
I am probably behind on reading this thread, as I have very limited time on the internet here at FSI. Montreal is truly a beautiful city. Had a great meal at Russian restaurant and then came back to listen to the latest Focal Grand Utopias.
I had not been a big fan of the Focals in the past, having usually preferred the Wilson Sophia (yes, over the big Wilsons). Focal finally got it right, as near as I could tell. They were effortless. Their portrail of female and male voice was superb. I can't afford the $170k price tag anytime soon. If you have to ask whether that is Canadian or US $, you can't afford them 🙂.
Cheers,
Bob
I am probably behind on reading this thread, as I have very limited time on the internet here at FSI. Montreal is truly a beautiful city. Had a great meal at Russian restaurant and then came back to listen to the latest Focal Grand Utopias.
I had not been a big fan of the Focals in the past, having usually preferred the Wilson Sophia (yes, over the big Wilsons). Focal finally got it right, as near as I could tell. They were effortless. Their portrail of female and male voice was superb. I can't afford the $170k price tag anytime soon. If you have to ask whether that is Canadian or US $, you can't afford them 🙂.
Cheers,
Bob
Attachments
Hi Bob,
Let me know if you are heading to the Toronto area. I live just west of there.
Some day I'd like to take in the Montreal show.
-Chris
Let me know if you are heading to the Toronto area. I live just west of there.
Some day I'd like to take in the Montreal show.
-Chris
Re: ratio
That would be my guess as well.
That goes a bit back to the question I raised; what is more important? The absolute value of 7th, or its ratio to, say, second. If the latter is the answer, it would seem to imply that the presence of second may somehow mask 7th.
Cheers,
Bob
Edmond Stuart said:
That would be my guess as well.
That goes a bit back to the question I raised; what is more important? The absolute value of 7th, or its ratio to, say, second. If the latter is the answer, it would seem to imply that the presence of second may somehow mask 7th.
Cheers,
Bob
john curl said:
Soryy, John, I have to disagree. Even if your premise was right about PIM, it could not exist without showing up in the spectrum.
Cheers,
Bob
Lucky you , I wish I could see the "real" APT . 😱
I actually have .. with a wilson cm and a 100r trimmer
on the PNP VAS. You trim it for small signal and it stays
nearly perfect with any signal (real amp - cheap picoscope).
the leach is more dependant on the P/N hfe matching as the
apt is not. (ran the real amp with BC546/ksa1845 LTP pair).
It is also known that complimentary differentials cancel out
even order harmonics and single ended ones (apt, DX)cancel
out odd orders (H5/7).
So I pursue the best of both worlds , the low Thd of the leach
without the odd order components. And matching is a breeze
(dual SMD one chip NPN pair - screw hand matching)
the wilson CM balances itself , so no issue there.
OS
Re: Re: VAS fighting
Hi Glen,
Take a look at the circuit I just posted. This one is quite stable on the VAS bias and relatively un-crippled, yet it exhibits the VAS fighting problem. I want to emphasize that not all complementary input pair architectures are prone to this, but it is something that one should be on the lookout for.
Cheers,
Bob
G.Kleinschmidt said:
Hi Glen,
Take a look at the circuit I just posted. This one is quite stable on the VAS bias and relatively un-crippled, yet it exhibits the VAS fighting problem. I want to emphasize that not all complementary input pair architectures are prone to this, but it is something that one should be on the lookout for.
Cheers,
Bob
That must be a worst case scenario.. all that gain (CM + EF).
GK will "raz" you for only getting .012% with a leach. 😀
OS
Hi ostripper,
Not wanting to be OT in John's thread too long, but it does concern complimentary differentials. The standard diff pair cancels even order harmonics as long as it's balanced. I find that many complimentary differentials sound "odd" until they are closely matched. Even then, I don't care for them for a reason that has just come up in a conversation I am having with a couple other designers.
It has been my long held belief that BJT diff pairs have a variable input impedance. So, unless your input signal is driven from a low impedance source, the signal is distorted to some degree before it even enters the amplifier. I had discussed this before somewhere. I feel that the complimentary differential input suffers more from this. I'll bet that this is why I strongly prefer J-Fet inputs. I doubt tube amplifiers suffer from this unless you cause grid current to flow. Complimentary J-FET differentials cause me to be very suspicious because they likely will not act in a matched complimentary manner. This would be the entire reason they are supposed to exist. Since John has far more experience with these, I'll defer to his judgment here.
There have been amplifiers designed that incorporated an input buffer while the main circuit using BJT differentials. I know of two examples. One dating from 1969 and the other in the 90s I think. A strong argument that this effect is real. I'll bet other products also used an input buffer for this very reason. It would also be a good place to limit your bandwidth if you wanted to. I would.
-Chris
Me confused.
Not wanting to be OT in John's thread too long, but it does concern complimentary differentials. The standard diff pair cancels even order harmonics as long as it's balanced. I find that many complimentary differentials sound "odd" until they are closely matched. Even then, I don't care for them for a reason that has just come up in a conversation I am having with a couple other designers.
It has been my long held belief that BJT diff pairs have a variable input impedance. So, unless your input signal is driven from a low impedance source, the signal is distorted to some degree before it even enters the amplifier. I had discussed this before somewhere. I feel that the complimentary differential input suffers more from this. I'll bet that this is why I strongly prefer J-Fet inputs. I doubt tube amplifiers suffer from this unless you cause grid current to flow. Complimentary J-FET differentials cause me to be very suspicious because they likely will not act in a matched complimentary manner. This would be the entire reason they are supposed to exist. Since John has far more experience with these, I'll defer to his judgment here.
There have been amplifiers designed that incorporated an input buffer while the main circuit using BJT differentials. I know of two examples. One dating from 1969 and the other in the 90s I think. A strong argument that this effect is real. I'll bet other products also used an input buffer for this very reason. It would also be a good place to limit your bandwidth if you wanted to. I would.
-Chris
ostripper said:
The sim of that circuit with the caps of equal value drops the THD-20 down to the neighborhood of 0.00005%.
Cheers,
Bob
That is a world of difference.. 😱
I must be doing something wrong. A real leach does not do
that good. (Ideal CCS's / voltage sources
)OS
OS - I think by "that circuit", Bob meant the circuit in the diagram he posted, not the Leach.
- Status
- Not open for further replies.