Canuck Audio Mart Hifi and Audio ForumCanuck Audio Mart Hifi and Audio Forum
It is currently Wed Sep 19, 2018 8:17 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 124 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8, 9  Next
Author Message
PostPosted: Thu Mar 08, 2018 1:50 pm 
Offline

Joined: Sat Feb 12, 2011 8:14 am
Posts: 214
Location: Windsor, ON, CA
Nakamichel wrote:
... As I said in a post above, the power supply (read DC) is the most important thing in an amplifier. I was like you before, trying night after night to find a way to get rid of that ripple...

I will not argue with that statement.
Agreed! If you do not have a good PS... You have absolutely, no chance of achieving a great, Power Amplifier.
Yes, Power Supplies are really, that important!

Nakamichel wrote:
... But experience's ended by pay. I found a marvelous way against the ripple that I can't tell here. But for the PSRR, the trick is simple (but costly) and efficient...

I am quite confident, that I already know how you did it. It is not a new idea. And yes, I agree that it is also very simple.
However, as you have already stated... It is not going to be 'inexpen$$$ive' to implement (*Especially for the Size of your PS - 3200VA). :D

Nakamichel wrote:
... The transformers need to have ultra low winding's resistance. My amp are 50w class A, 200w lost in heat. A friend of mine said "why the hell you use 3200va, it's not overkill, it's insane."
Ah... I don't need 3200va at all, I need ultra low serie resistance. That's all. As you can imagine, my power rails don't move at all. They are uber strong.
It also helps to have a high DF. It's also very beneficial for the PSRR (the ability of the amp to not move the rails on current demands).

Agreed. Therefore, what have you been able to achieve for your 'Nominal' V Power Transformer Outputs and 'Minimum' V Power Transformer Outputs (@ Full Load)?
Also, (and only if you should choose to share)... Who wound those Power Transformers -- For you?
Just curious... :)
--
That sounds like a very, interesting (albeit Large) Amplifier Project. Thank-you!
Best Regards, -Tim.
________


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 2:36 pm 
Offline

Joined: Mon Apr 26, 2004 2:33 pm
Posts: 3204
Location: Toronto, ON, CA
retiredelectrician wrote:
PS. i asked this question because my mere NAD 100 watt "THX" ? amp drives my Proac Response 3's quite well. I have never seen the clipping indicators come on (maybe they don't work? lol). But. I do notice what appears to be a little sloppiness in the bass.


Joe,

Hey, I was an audio genius in the 1970s, too! Funny how that brilliance has slipped away from us. . .

As for the "sloppiness in the bass," that could simply be a limitation of that particular speaker design and/or it could be an amplifier issue, but not so much an issue of damping factor, as one of power and current delivery.

Likely that old Bryston 4B—or any of the newer ones—would tighten up the bass in your Proacs. Although I don't find Brystons to be the "bass masters" that many reviewers call them.

In other words, a more powerful, higher current amp might well cure or improve the sloppiness in the LF. My wager would be that you'd notice real improvements with such an amp.

In terms of specs, ignore the damping factor (if supplied), but do take a look at power output at lower impedances. Again, no guarantees in audio, but an amp that doubles its power output from 8 ohms to 4 ohms is a good candidate.

My own experience is that, other things being equal—and I admit that they often are not!—the more power the better. And I am talking about performance at low listening levels as much as at high levels.

In addition to Bryston, there's lots of good choices in the 200 Wpc and up power category.

For example, those two other Canadian amplifier stalwarts: Classé and SimAudio. Older McCormacks such as the DNA-1, especially those upgraded by SMc Audio, are brilliant amps, especially for what they often go for on the used market.

D-Sonic in Texas is getting a big thumbs-up from me these days. Forget their amps are Class D—but Pascal not Ice Power—and have neither fans nor heatsinks. Forget that a stereo amp of 1500 Wpc cost you under US$2500 brand new with warranty. Pretend, rather, it was three times that price and enjoy the music—including bass that's as tight and agile as a ballerina's bottom.


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 3:12 pm 
Offline
User avatar

Joined: Mon Mar 31, 2008 9:17 am
Posts: 14897
Location: Toronto, ON, CA
b3733366 wrote:
The car suspension analogy is not only appropriate, it is literal, and that is exactly the point Toole is making ... that the electrical damping factor (specification) don't mean squat, rather the dampening that is going on is a function the physical system, it stops itself. It's not a standard spec., and certainly isn't the electrical damping factor.

I wouldn't go that far. Speaker T/S parameters differentiate the mechanical and electrical Q of the drivers. Though often ignored in calculations, the amp's DF does factor in the value of the driver's electrical Q. This is one reason why speakers frequently sound markedly different with different amps.

_________________
To a hammer, everything looks like a nail


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 3:39 pm 
Offline

Joined: Sat Feb 12, 2011 8:14 am
Posts: 214
Location: Windsor, ON, CA
dr.joe wrote:
... D-Sonic in Texas is getting a big thumbs-up from me these days. Forget their amps are Class D—but Pascal not Ice Power—and have neither fans nor heatsinks. Forget that a stereo amp of 1500 Wpc cost you under US$2500 brand new with warranty. Pretend, rather, it was three times that price and enjoy the music—including bass that's as tight and agile as a ballerina's bottom.

Also, in the Class D Category... Is the new Mivera Audio Icepower 1200AS Purepower and Purepower SE Amps.

http://www.audiocircle.com/index.php?topic=154913.0
https://www.miveraaudio.com/
https://www.miveraaudio.com/shop


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 4:02 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 649
Location: Montréal, QC, CA
Takira71A wrote:
Nakamichel wrote:
... But experience's ended by pay. I found a marvelous way against the ripple that I can't tell here. But for the PSRR, the trick is simple (but costly) and efficient...
I am quite confident, that I already know how you did it. It is not a new idea. And yes, I agree that it is also very simple.
However, as you have already stated... It is not going to be 'inexpen$$$ive' to implement (*Especially for the Size of your PS - 3200VA). :D
I didn't invent anything regarding this. However, I won't reveal how I did it. Then, we'll never know if you actually know unless you share
your thoughts here that I won't confirm or deny. Which doesn't have much interest for you since I'll be the only one to know if you're right.

Takira71A wrote:
Nakamichel wrote:
... The transformers need to have ultra low winding's resistance. My amp are 50w class A, 200w lost in heat. A friend of mine said "why the hell you use 3200va, it's not overkill, it's insane."
Ah... I don't need 3200va at all, I need ultra low serie resistance. That's all. As you can imagine, my power rails don't move at all. They are uber strong.
It also helps to have a high DF. It's also very beneficial for the PSRR (the ability of the amp to not move the rails on current demands).
Agreed. Therefore, what have you been able to achieve for your 'Nominal' V Power Transformer Outputs and 'Minimum' V Power Transformer Outputs (@ Full Load)?
I'm afraid that I don't understand what you mean here. But I can tell you that each symmetrical part (supply+outputs) can deal with 0.5Ω.
For a total of 1Ω. But the actual heatsinks can deal with only 4Ω+4Ω. 8Ω. (200w heat for 50w power)

Takira71A wrote:
Also, (and only if you should choose to share)... Who wound those Power Transformers -- For you?
Just curious... :)
The R&D was made with 4 Antek transformers (they're cheap) of 4x 12v AC 800va toroids that I removed some turns to get ~10v AC.
But now I use R-core transfos of the same ratings. Toroids are not good for high power. The core get saturated in the inner part more than the outer part. Not good at all.
Even EI transfos are better in this aspect. Although toroids are less lossy and have lower radiation and vibration. R-core are better in all those aspects.
They are less lossy, radiate and vibrate less and all the windings are in a linear region. I did wind them myself. Ultra easy to do with a bobbin.

Takira71A wrote:
That sounds like a very, interesting (albeit Large) Amplifier Project. Thank-you!
Best Regards, -Tim.
Agree !

- Mik

_________________
Keep things simple and even simpler
Nakamichi BX-300/Balance BW-1010 DAC, Balance MR-50R, Infinity RS 8 Kappa
Nakamichi Dragon/Onix Dac25b, Audio Research SP-9, Spectral DMA-100s/Nakamichi PA-7II, Martin-Logan CLSII/Balance Helicia


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 4:41 pm 
Offline

Joined: Sat Feb 12, 2011 8:14 am
Posts: 214
Location: Windsor, ON, CA
Nakamichel wrote:
The R&D was made with 4 Antek transformers (they're cheap) of 4x 12v AC 800va toroids that I removed some turns to get ~10v AC. But now I use R-core transfos of the same ratings. Toroids are not good for high power. The core get saturated in the inner part more than the outer part. Not good at all.

Even EI transfos are better in this aspect. Although toroids are less lossy and have lower radiation and vibration. R-core are better in all those aspects.
They are less lossy, radiate and vibrate less and all the windings are in a linear region. I did wind them myself.

Thank-you! That answers a lot of my questions.

Initially, I thought that you were 'perhaps' using a Custom ('One-off') Power Transformer from somebody like 'perhaps' Heyboer Transformers
(*As, they will pretty much wind up anything for you).
However, @ 3200VA... You probably, would need a 'Front-End Loader' to lift it. :lol:
--
Yes... 'R core' Transformers are very, good indeed. Excellent Choice!
And the fact that you wound them yourself... Makes it even better and will give you the very, best "Pride of Ownership".

Thanks for writing Mike!


Top
 Profile  
Reply with quote  
PostPosted: Thu Mar 08, 2018 6:49 pm 
Offline

Joined: Sat Dec 08, 2007 4:04 pm
Posts: 143
Location: Toronto, ON, CA
Dr. Joe

Recently I had a Bedini 100/100 and a 90 Watt Rogue, borrowed from friends in my system to try out.
Overall, though not an ABX comparison, there were some subtle differences overall in the upper bass, mids and treble, but what was bothering me prior from my NAD, bass that sounded a tad sloppy and almost out of control, cleaned up with both amps. It is for sure not a Proac Issue.
My long time friends in the stereo retail biz have their feelers out for another awesome deal on a used amp.

I told them just find me another arc welder like the KSA 250 they scored for me a few years ago.

We will see what emerges.


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 8:32 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4027
Location: Ottawa, ON, CA
ripblade wrote:
b3733366 wrote:
The car suspension analogy is not only appropriate, it is literal, and that is exactly the point Toole is making ... that the electrical damping factor (specification) don't mean squat, rather the dampening that is going on is a function the physical system, it stops itself. It's not a standard spec., and certainly isn't the electrical damping factor.

I wouldn't go that far. Speaker T/S parameters differentiate the mechanical and electrical Q of the drivers. Though often ignored in calculations, the amp's DF does factor in the value of the driver's electrical Q. This is one reason why speakers frequently sound markedly different with different amps.

I avoid analogies, since they usually break down.
Also, its been about 30 years since I looked at transfer functions
however, If I recall correctly, a car suspension develops a second order transfer function.
It provides a great example of damping requirements.
As I recall, an ideally damped system occurs when Tau = 0.707 (1/sqrt(2)).

As the article posted by Takira71A pointed out, when an amp manufacturer expresses damping factor
its assuming a real impedance 8 ohms - in order words, there is a resistive component only.
That's in the land of make believe.
Clearly, the transfer function needs to be developed using the output impedance of the amp
and a reasonable LCR model of the speaker's impedance being used with that amp.
Minimizing the impedance of the speaker cable, Amp Zout and Speaker Zin form a voltage divider.

The part I don't understand is that in an ideally dampened system, Tau = 0.707.
But, with an amplifier and speaker, Amp Zout should be minimized, in order to get
current from the power supply, through the actual amp, to control the speaker's voice coil.
As Nakamichal has indirectly said, the ideal power transformer [and power supply]
would have an output impedance of 0 ohms.

BTW: this little read on speaker parameters looks good.
http://audiojudgement.com/speaker-equivalent-circuit/


.


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 11:34 am 
Offline

Joined: Sat Feb 12, 2011 8:14 am
Posts: 214
Location: Windsor, ON, CA
Uunderhill wrote:
... As I recall, an ideally damped system occurs when Tau = 0.707 (1/sqrt(2)) ...
... The part I don't understand is that in an ideally dampened system, Tau = 0.707...

What is "Tau" the acronym for?
Thank-you!


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 12:33 pm 
Offline
User avatar

Joined: Mon Mar 31, 2008 9:17 am
Posts: 14897
Location: Toronto, ON, CA
Uunderhill wrote:
BTW: this little read on speaker parameters looks good.
http://audiojudgement.com/speaker-equivalent-circuit/

Thanks....good read.

If you look at it as a source/load relationship it follows that the source should ideally be 0 ohms. Most speaker designs are built to a Qt of about 0.9; if the source impedance is not 0 ohms the target Q will rise. Of course, the target Q is measured/adjusted with an amplifier of known Zout, so if the end user doesn't plug something completely different in their system the designer's target Q should be realized.

-- 09 Mar 2018 16:36 --

Getting back to that Gamut mosfet, is it possible that they paralleled several on the same substrate? Would account for the unusually high input capacitance, and maybe the ability to double current into 4ohms. Also, being on a single substrate, it might sound like a single device as tolerances would be identical, rendering matching unnecessary.

I know they do this for long-tailed pairs for the exact same reason. Why not output devices?

_________________
To a hammer, everything looks like a nail


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 12:49 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 649
Location: Montréal, QC, CA
Tau is a greek letter. Used here to represent a time constant or 3dB or .707 or 1/sqrt(2) or 20 log.

The car analogy is unfortunatly incorrect.
In a speaker system, the moving mass receives force in only two direction (bi-directional).
In a car suspension, the wheel up/down movement is also affected by lateral force(s). (when the car accelerate or decelerate)
The transfert function curve must be then calculated with 3 dimensions. (one more pole required)

And as Ripblade pointed out, the T/S parameters differentiate the mechanical and electrical Q of a driver.

Uunderhill wrote:
The part I don't understand is that in an ideally dampened system, Tau = 0.707.
But, with an amplifier and speaker, Amp Zout should be minimized, in order to get
current from the power supply, through the actual amp, to control the speaker's voice coil.
It's because in an ideally dampened system, the mass would (also) be considered null. Since in real life we have a mass to deal with, we have then to play with another parameter (lowering Z as most as possible).
It's why I use 'insane' va to lower Z.

ripblade wrote:
Getting back to that Gamut mosfet, is it possible that they paralleled several on the same substrate? Would account for the unusually high input capacitance, and maybe the ability to double current into 4ohms. Also, being on a single substrate, it might sound like a single device as tolerances would be identical, rendering matching unnecessary.

I know they do this for long-tailed pairs for the exact same reason. Why not output devices?
I knew you were good Ripblade, but now you're a genius because you have just read my mind. :mrgreen:

_________________
Keep things simple and even simpler
Nakamichi BX-300/Balance BW-1010 DAC, Balance MR-50R, Infinity RS 8 Kappa
Nakamichi Dragon/Onix Dac25b, Audio Research SP-9, Spectral DMA-100s/Nakamichi PA-7II, Martin-Logan CLSII/Balance Helicia


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 1:25 pm 
Offline
User avatar

Joined: Mon Mar 31, 2008 9:17 am
Posts: 14897
Location: Toronto, ON, CA
Nakamichel wrote:
ripblade wrote:
Getting back to that Gamut mosfet, is it possible that they paralleled several on the same substrate? Would account for the unusually high input capacitance, and maybe the ability to double current into 4ohms. Also, being on a single substrate, it might sound like a single device as tolerances would be identical, rendering matching unnecessary.

I know they do this for long-tailed pairs for the exact same reason. Why not output devices?
I knew you were good Ripblade, but now you're a genius because you have just read my mind. :mrgreen:

Genius? Not even close. But I do have a wild imagination. :lol:

_________________
To a hammer, everything looks like a nail


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 1:39 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4027
Location: Ottawa, ON, CA
ripblade wrote:
Getting back to that Gamut mosfet, is it possible that they paralleled several on the same substrate? Would account for the unusually high input capacitance, and maybe the ability to double current into 4ohms. Also, being on a single substrate, it might sound like a single device as tolerances would be identical, rendering matching unnecessary.

I know they do this for long-tailed pairs for the exact same reason. Why not output devices?

The IRF equivalent part number FB180SA10 - on the data sheet it says the HEXFETs are paralleled
Ciss = 10.7 nF

however, with the IXYS equivalent IXFN 180N10 - the data sheet says single MOSFET die
Ciss = 9.1 nF

Seriously - I have absolutely no idea how Gamut gets such a low output impedance
from their amps using 2 of these devices.
Especially, since a 0.33 or 0.47 ohm source resistor needs to be connected
to control thermal runaway.
Again, a switching MOSFET has a positive termperature coefficient
where as lateral MOSFET's have a negative temperature coeficient.
For an analog amp, a follower type output stage has the lowest Zout that I know of.
But this can be improved by paralleling up MOSFET 's.

The term long tail pair implies a twin triode is being used, with a large voltage drop across a resistor
trying its best to be a current source.
A differential pair is formed by replacing the resistor with a CCS which will improve the CMRR (common mode rejection ratio).
As you pointed out, its critical to match both gain devices in a different pair to maximize the CMRR.
This is not just a theoretical exercise.
When 2 jFET's are used to build a differential pair, and they are not matched, you can hear the sonics move further back.
That's why in some tube circuits, having matching twin triodes, in the same tube, is important.
... and you are right, Toshiba used to make a pair of matched jFET's in a single package
but of course that part is discontinued.

Actually, I think the long tail pair was developed by the British for RDF (Radar) in the late 1930's.

.


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 2:08 pm 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 756
Location: London, ON, CA
Some hints from the world of professional audio power systems. This graphical tool has been included, for many years, with all manuals and application info for CROWN pro audio amplifiers.

These folks, and their competitors, have the reputation of being able to accurately reproduce a "kick drum" at stunning power levels at live performances.

Interesting to me that their selection of speaker wire size (AWG) is made here strictly on the basis of maintaining the damping factor!

The example shown on this graph indicates that a 10 foot long speaker cable needs to be copper wire of 8 AWG to maintain a damping factor of 500.

Using a ruler to put a different line through the wire length and thickness scales to meet the same point on the amplifier source resistance, shows how much the speaker wire diameter can be reduced, if shortened.

This information matches my listening experience with active woofers and powered subwoofers. Their power amplifiers may be connected to the speaker terminals with only 1 foot of wire, with no crossover component between amp and driver.

I'd appreciate knowing if someone could do a little experimenting with shorter lengths of very thick speaker wire with something like a Bryston or Krell amp, that they feel does not seem to kick like it should. Maybe it's the speaker wire?

Attachment:
CROWN Power Ampilifier Reference Manual Wire Size Nomograph..jpg
CROWN Power Ampilifier Reference Manual Wire Size Nomograph..jpg [ 64.44 KiB | Viewed 312 times ]


Top
 Profile  
Reply with quote  
PostPosted: Fri Mar 09, 2018 2:29 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 649
Location: Montréal, QC, CA
You are right jst_canuck, I did test my amps with different wire length/gauge.
No doubt here, shorther and bigger is better...

That reminds my friend's system. He had a H/K receiver with big Cerwin vega (can't remember model).
One day I saw his speakers wires. Thin like a lamp cord. I was laughing. I told him to try some bigger wires to get some bass.

His answer was funny... but it really was true. He tried with bigger gauge, but the amp protection was always triggered on each kickdrum !
He said, I'm not able to drive the volume with big wires. Obviously, the amp was unable to drive those big CVs. The thin wires acted like hp filter.
The Z, DF and LF were really affected by those thin wires.
But he was happy with the result. :lol: :lol:

Uunderhill,
here's a good application note to read
https://assets.nexperia.com/documents/application-note/AN11599.pdf

_________________
Keep things simple and even simpler
Nakamichi BX-300/Balance BW-1010 DAC, Balance MR-50R, Infinity RS 8 Kappa
Nakamichi Dragon/Onix Dac25b, Audio Research SP-9, Spectral DMA-100s/Nakamichi PA-7II, Martin-Logan CLSII/Balance Helicia


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 124 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8, 9  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group