John Curl's Blowtorch preamplifier part II

[john curl]

Joshua, I'm serious, you should do the math on amp loading, however, re-reading your post implies that you did not mean CONTINUOUS loads, like Steve implied. OF COURSE this little amp of mine will drive sometime low impedance loads, that is one of my trademarks.

[stinius]

Originally Posted by Joshua_G

Thanks. Where can I find those designs?
.

I didn’t think it was possible to not see Mr Ovidiu's contributions to this forum.

Anyway this is his latest design, read the thread and see what the response is. here

You will also find a lot of other designs on his website, and they are free of charge.

BTW: Mr Ovidiu is also known as syn08.

Cheers

[Joshua_G]

Originally Posted by stinius

BTW: Mr Ovidiu is also known as syn08.

Thanks.
I'd rather skip his designs.
I'd never try designs by some one who doesn't hear differences between cables, because it means to me that he wouldn't hear differences in design the affect the sound quality the way I hear it.

[simon7000]

Originally Posted by syn08

Hi Simon,

Could you expand on those numbers? What exactly are they when it comes to noise? I see there's a frequency dependency, so they're probably not V/rtHz. They seem to be Volts, but then 10nV RMS noise (-160dB) in the audio band, in a power supply output, is out of this world, even for JC

I worked once to a project involving a low noise satellite receiver, liquid nitrogen cooled, etc... having the noise performance about 5x better than what we are talking here and trust me, the power supply (+18V) had way more relaxed requirements.

Is there something specific to high end audio that requires better performance?

BTW, DIYers have only one customer, but that's already another discussion.

Been away sorry for the delay.

A good RC filter with vibration isolation will make 10nv per sqrt Hz RMS. A good film capacitor is required. An electrolytic will get close.

HOWEVER if you have a typical 60db PSSR the I would say look carefully at both positive and negative PSSR, then relax to -120db, allowing 20db of margin. That is why many find the microvolt level to be more than adequate.

I used to use -140db as a design value and have found every so often the noise would begin to crop up, so I changed my design target.

I have a decent sheet metal machine shop. Aluminum has been as low as $1.10 a pound and as high as $3.50 for 5052 Alloy. I assume Blowtorch cases are 6061 which is a bit more. A slab 17 x 12 x 3 would weigh about 60 pounds. You would need two of these and another for the covers. Markup for small runs is 2 to 5 times cost of material. Anodizing is 3.5 to 5 cents per square inch. Crazy customer doubles all prices... really.

[syn08]

Originally Posted by Joshua_G

I'd never try designs by some one who doesn't hear differences between cables,.

You certainly made my day, Joshua That's a keeper, can I quote you on my web site?

[simon7000]

I have an article in This month's AudioXpress. It shows how to measure cable distortion. Also has some hard to read AP screen shots of different cables and forward/reverse differences.

No surprises here, silver is better than copper, deoxidizer is important and load impedance matters.

[PMA]

You guys are happy in this thread, as you may post off-topics here

[stinius]

Originally Posted by Joshua_G

Thanks.
I'd rather skip his designs.
I'd never try designs by some one who doesn't hear differences between cables, because it means to me that he wouldn't hear differences in design the affect the sound quality the way I hear it.

I'm not sure what to say without ending up in the sin bin, so I’ll just let it be.

[Steve Dunlap]

Originally Posted by john curl

Well, we leave it to you, Steve, to build the cheap stuff. What can you do with Autos?
Right now my 14 year old Acura with 26,000 mi is worth about $3,000. It runs almost perfect, gets 34 mpg at 70mph. Can you get me a better deal? ;-)

Here is some of my "cheap stuff". 100W mono block 200W at 4 ohm stable at 300W into 2 ohms. The bid to manufacture these at 1000 pieces was under $100 per pair, packed in shipping boxes. They were powder coated. There were no cheap sheet metal screws. I used press in captive nuts and button head Allen screws. Transformers were EI core made in house by the same manufacturer. I don't know what they cost because the bid was for the completed units.

http://www.diyaudio.com/forums/attac...k-internal.jpg

I built some not so cheap amps also because a customer wanted them. They were 0.195 Al cases with 1" face plate and back panel. Material for the case was just under $250 with an additional $125 each ($500 for 4 cases) for machining. The finished amp weighed 110 lb and didn't look cheep at all.

I don't do cars, sorry.

[syn08]

Originally Posted by simon7000

HOWEVER if you have a typical 60db PSSR the I would say look carefully at both positive and negative PSSR, then relax to -120db, allowing 20db of margin. That is why many find the microvolt level to be more than adequate.

Now it makes some sense, these numbers are also matching my experience. The HPS 3.1 Jung superregulators have around 100nV/rtHz (plus about the same amount of 1/f excess noise under 100Hz) and this, combined with a cap multiplier and the gain stage PSRR, is bringing the PS contribution to the total noise to almost nil.

All I am saying here is that it is much easier to build some decent PSSR in the gain stage (at least a cascode) rather than pushing unrealistic requirements (and allow for compromises, like the output impedance) in the power supply.