AMP A / B COMPARISONS

[stainless824]

Omg, a monoblock b22 exists? Shottie makin one when i get bored of my burson and rake up enough $$$

 

[milosz]

Quote:

 
Quote:

Originally Posted by milosz /img/forum/go_quote.gif
 
The M³ uses[size=small] AD843 op amps.[/size]
 

Did you try any other opamps?  I find the 843s a little thumpy and grainy relative to the 637/627s, which appear to be smoother and have greater extension, on my M^3s.
 
The last time I did any testing, I found my Woo3 (Cetron-Jan tubes)  sounded remarkably like my 637/627 M^3.  My GS-1 sounds different with noticeably more space around things and a wire with gain sound.
 
USG

No I never tried any other opamps, though I have  OPA627's sitting around that I want to try.  When I can find the time.  (Or is it OPA637's that I have?  I can't remember.) 
 
It is a good point re:  M3 - you can try ALL KINDS of opamps in there.  You could even use Audio-Gd discrete transistor opamp modules SUN, MOON or EARTH, though these may not fit in some M3 cases.  See http://www.audio-gd.com/Pro/diy/OPA/OPAEN.htm   they are not too expensive either.  And if you have friends who are snobby about monolithic opamps, using  EARTH modules will let you put them in their place....
 

 

[milosz]

Quote:

I know you didn't have a chance to A/B the beta before and after the upgrade but can you expand a litle about the diffrences you felt going from 2-ch to 3-ch?

It has one more board in it.  

 
I really can't say I heard any difference. There may be in fact a difference but, you know, the 2 channel Beta-22 sounded great and so does the 3-channel. The only difference in sound that I can clearly point to is that my amp had HUM after I added that 3rd channel.  Ultimately I had to enclose the power transformer in a steel sub-chassis box to cure this.  I can't work out why, exactly, but it seems the 3-channel version is more susceptible to picking up hum from magnetic fields, like the 60 Hz field from the transformer. At first I thought I had a ground loop or something, but in the end it was cured by wrapping the transformer in steel and not by changes in ground scheme or wire dressing. 
 
But as far as sound quality, no, I have no impression of how it sounded before /after so I just can't say.  In THEORY I would think that the most noticeable thing might be better bass- I think the active ground has a big effect on damping factor.
 
The Beta 22 specs list "Output impedance  less than 0.01Ω, 20Hz - 20KHz"  which is very low, this is what determines damping factor. This is for a 3-channel topology.  I'll ask AMB what it would be for 2-channel.  But you know, this is just theory. Maybe this makes a difference in sound, maybe not. 
 
Even though I found very little easily audible difference between Beta 22 and M3 on HD800 and LCD-2, I still prefer to use the Beta 22. This is purely psychological.  I just FEEL more appreciation for the Beta. Probably because it was a fair amount of work to build., I spent so much time on it, and because I admire the engineering in the design. It does sound GOOD - of that there is no doubt- the fact that it may not really sound much better (if at all) than the M3 doesn't mean I don't LIKE it more. If it sounded clearly worse than other amps, I would not like it all that much.  But the fact that it sounds as good or better than the other amps, and in addition I have certain emotional responses to it's design and my relationship with it during the build add up to me LIKING it more.  It's sound first and foremost, but in the end it's MORE than just sound.
 
Which reminds me, I must buy a thicker front panel for my Beta 22.  Par Metal makes a thicker one than the "stock" one I have.... I just love chunky aluminum audio jewelry.  I'd love to have a CNC machining center, I could make some really great chassis. I know the guy who owns Tormach, I wish I could buy one of their benchtop 3-axis milling machines for cheap but he won't do that for me.  He won't let me come over to his shop and use his, either.  It's just as well, I'd put my eye out.

 

[eliziz]

Thanks! In about 2 weeks a should have my own beta (2-ch). It is hard to find information about the sonic differences between 2-ch and 3-ch builds.

 

[ardgedee]

This makes me curious how difficult it would be to make the shared ground board switchable, so you could do A/B tests on the beta22 against itself...
 
Thanks for the advice about projects. As tempting as it is to dive right into one, I know it's better for me to do some simpler things first and learn the basics. If I ever get into building a beta22 (or whatever comes along by then), what I imagine doing is finding an old Sparcstation pizza box to put it in. They look handsome (as computer cases go) and the shielding's built in, and when one crops up, it's usually available for the cost of carrying it away. Only real problem is even when they look great in windowless server rooms, the plastic yellows in natural light.

 

[PianistOne111]

These would be so much better if they were blind. Can't you get someone to hook up the amps to the switch for you?

 

[milosz]

Quote:

These would be so much better if they were blind. Can't you get someone to hook up the amps to the switch for you?

Yeah they would be better tests if on a more scientific footing.  To that end, I am building a box that provides a way to let software do the switching, based on a table of random numbers. The software would also record results that I could analyze in Excel.  This would be a true random A/B/X  test.  Not done making this yet, though.  (It's actually pretty simple- just using USB to control some relays. Software is a little harder, I have written some code in my day but I'm rusty.)
 
If I get ambitious, I will trot this setup around to some meets and let some golden ears take a crack at it, see how good they are at telling these amps apart.  
 
I plan to get more involved with this A B testing thing. I have recently registered the domain ABTESTS.ORG where I will post my tests in the next few days, and plan to start a nonprofit organization to act as overall venue for these activities. (I currently do I.T. for a few nonprofits in my spare time and my business partner is an attorney who has a specialization in nonprofit organization law, so it would be relatively easy for me to start this.)

 

[micmacmo]

milosz, I got to say that your A/B testing is fantastic as it is and your level of ambition is really impressive. I think this is a great service to the community.  

 

[Prog Rock Man]

I agree milosz, this is brilliant. Your tests appear very scientific to me. I suspect that a blind test would result in the same as this one
 
http://www.bruce.coppola.name/audio/Amp_Sound.pdf
 
which found level matched amps (Pioneer, NAD, Mark Levinson and some very high end tube amps) do sound indistinguishable.
 
I am also not surprised that the Yamaha integrated amp you tried sounded surprisingly good as I auditioned some headphones from an updated version and was surprised at the sound quality.

 

[BebopMcJiggy]

Quote:

The M3 has lower random noise, THD, and IMD than the B22 into any load. Just FoodForThought.

Aren't they measured with different gains and such?

 

[milosz]

I just bought a used Tim Rawson-built First Watt F3 clone on Audiogon. The First Watt F3 is a Nelson Pass design using power VFETs, and while it is a low power amplifier, it is supposed to sound just great. You can read the 6moons review here: http://www.6moons.com/audioreviews/firstwatt4/f3.html
 
 
I am anxious to see what this sounds like on the Audeze LCD-2 and the HiFiMan HE-6.  And I wonder if I am risking disaster if I use it with the Sennheiser HD800.  Let me see if I can find my Ohm's Law Thinking Cap here.... It's 4 AM here in Chicago (I am an insomniac-) and at this hour, for me, it's a little hard to work this out. Just fiddling around with guesses made from looking at the distortion-vs-power curves for 4, 8 and 16 ohm loads from the 6moons audio review of the F3, I'm going to guess that this amplifier is capable of delivering about 12 volts into a 300 ohm load.  That's just under the 0.5 watt "Max. nominal long-term input power" spec from Sennheiser for the HD800. I'm not so concerned about blowing out the HD800's from listening, but from some inadvertent preamplifier pop or loud 60 Hz hum from an RCA cable falling out or something.  But it doesn't sound like the F3 can produce enough power in such a scenario to open the HD800 voice coils. The F3 is a pretty low-gain amplifier, so anything like a halfway-out RCA cable on it's inputs ought not to generate an output strong enough to vaporize my HD800's.  (I hope)
 
Of course I could always add a series resistor. This amplifier has a low damping factor of 8.  So how much sound quality would I lose adding some series resistance?  The amp doesn't really have a "tremendous grip" on the voice coils with a damping factor of 8, so maybe reducing that damping factor by adding a series resistor won't make a difference.  But maybe I will want to keep all the damping factor I can, considering I am starting with pretty low damping to begin with.
 
The Beta 22 has a damping factor greater than 30,000 (!) with the Sennheiser HD800's, and greater than 5,000 with the 50 ohm planar headphones (if you use superconducting headphone cables.)  (The Beta 22 lists output impedance as "less than .01 ohms from 20-20,000 Hz," which is remarkably low.  Damping Factor = load impedance divided by amplifier output impedance.)
 
Anyway, a Nelson Pass design will be an interesting addition to the A/B tests.

 

[Rdr. Seraphim]

I had a similar notion to use one of the Pass' lower power amp designs for a reference headphone system. Just add the Placette Remote Volume Control (128 steps), and a good DAC (fed anyway you'd like, USB, CDP, etc.) and this "should" be a great system. 
 
Quote:

I just bought a used Tim Rawson-built First Watt F3 clone on Audiogon. The First Watt F3 is a Nelson Pass design using power VFETs, and while it is a low power amplifier, it is supposed to sound just great. You can read the 6moons review here: http://www.6moons.com/audioreviews/firstwatt4/f3.html
 
 
I am anxious to see what this sounds like on the Audeze LCD-2 and the HiFiMan HE-6.  And I wonder if I am risking disaster if I use it with the Sennheiser HD800.  Let me see if I can find my Ohm's Law Thinking Cap here.... It's 4 AM here in Chicago (I am an insomniac-) and at this hour, for me, it's a little hard to work this out. Just fiddling around with guesses made from looking at the distortion-vs-power curves for 4, 8 and 16 ohm loads from the 6moons audio review of the F3, I'm going to guess that this amplifier is capable of delivering about 12 volts into a 300 ohm load.  That's just under the 0.5 watt "Max. nominal long-term input power" spec from Sennheiser for the HD800. I'm not so concerned about blowing out the HD800's from listening, but from some inadvertent preamplifier pop or loud 60 Hz hum from an RCA cable falling out or something.  But it doesn't sound like the F3 can produce enough power in such a scenario to open the HD800 voice coils. The F3 is a pretty low-gain amplifier, so anything like a halfway-out RCA cable on it's inputs ought not to generate an output strong enough to vaporize my HD800's.  (I hope)
 
Of course I could always add a series resistor. This amplifier has a low damping factor of 8.  So how much sound quality would I lose adding some series resistance?  The amp doesn't really have a "tremendous grip" on the voice coils with a damping factor of 8, so maybe reducing that damping factor by adding a series resistor won't make a difference.  But maybe I will want to keep all the damping factor I can, considering I am starting with pretty low damping to begin with.
 
The Beta 22 has a damping factor greater than 30,000 (!) with the Sennheiser HD800's, and greater than 5,000 with the 50 ohm planar headphones (if you use superconducting headphone cables.)  (The Beta 22 lists output impedance as "less than .01 ohms from 20-20,000 Hz," which is remarkably low.  Damping Factor = load impedance divided by amplifier output impedance.)
 
Anyway, a Nelson Pass design will be an interesting addition to the A/B tests.

 

[milosz]

Quote:

This makes me curious how difficult it would be to make the shared ground board switchable, so you could do A/B tests on the beta22 against itself...
 
Thanks for the advice about projects. As tempting as it is to dive right into one, I know it's better for me to do some simpler things first and learn the basics. If I ever get into building a beta22 (or whatever comes along by then), what I imagine doing is finding an old Sparcstation pizza box to put it in. They look handsome (as computer cases go) and the shielding's built in, and when one crops up, it's usually available for the cost of carrying it away. Only real problem is even when they look great in windowless server rooms, the plastic yellows in natural light.

 
Ha ha  yeah a SPARC box would be a nice Beta 22 case.  We've got to find someone to cast little shiny BETA 22 badges out of bronze.
 
You COULD do a switchable 2/3 channel setup, you'd need relays.  Too many things need to be switched to use a toggle. You can get lots of dandy 5 volt relays on eBay for peanuts. I'd use a separate little transformer and diodes to make a 5 volt DC supply, keep it off the amp's supplies.  Again. you can get small 110v-5 volt transformers on ebay for pennies.  You'd only need 50 ma.
 

 

[stainless824]

I reckon fabricate your own case, what could be more exciting? Building your own B22 just to put it into a generic case makes me depressed.

 

[GreyArea]

I'm seriously considering flying over to the next Chicago meet to witness some of these tests. I'd be interested to see how my Lehmann BCL stacks up.
 
Also, looking forward to the DACMagic/Audio-GD test. 

 
 
 
Quote:

I reckon fabricate your own case, what could be more exciting? Building your own B22 just to put it into a generic case makes me depressed.

Generic case -> Depression? 
 
Maybe this is what you need your your B22 build.