John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
syn08

syn08

Banned
Joined 2005
synaesthesia.ca
john curl said:
Actually, for working audio engineers, Bob, I have been very helpful.
The situation is this:
My ST analyzer, has been modified by Scott Wurcer and me to perform better than its stated specifications. This has been due to using Scott's AD797 IC's as the input to the analyzer, and Scott's 744 IC in the oscillator section. I could probably replace a few more IC's, but so far, I can get what I want, once I analyze the residual with a HP 3563 FFT based analyzer with a 256K clock, so that I can signal average, and screen out the distortion products
Of course there is a little 2'nd and 3'rd residual, but they are at .001% and below, so I ignore them and concentrate on the 5th, 7th, and 9th harmonics. Ideally, I can have fairly little higher order harmonics, and I can test this with a bypass test using the best measuring cables in my possession and noting the residual. I can always subtract this residual from the actual measurement, if I chose to do so, but I have not found it that important to do up to this time.
Now, what I am having trouble is a nagging residual 7th harmonic that I would like to reduce or remove, so that I don't have to do differential subtraction, either in the test equipment or in my head. Then when I spot some 7th harmonic on the screen, I will be able to generally attribute it to the DUT, rather than the test equipment. It is more a matter of convenience to reduce the 7th harmonic, but it might be a useful improvement.
One approach that I have NOT tried at this point, is to force the output stages to run Class A, by adding a fet current source on the output to the - supply, and hoping for an improvement. Of course the 2'd harmonic might rise slightly, but MAYBE the 7th will drop.
As Scott Wurcer has, at least, past experience with his own IC output stages, I thought that he might give me some useful input in doing this.
Repeatedly, I have asked Scott for the current of peak beta of the npn output device, but so far no luck. I would make a difference to me, if it is .5 ma , 5 ma, or 50 ma. From my perspective, all 3 are possible from my experience with discrete devices.
Now, have I made myself more clear to both you and Scott?
Click to expand...

I'm using a very similar equipment (3562A instead of 3563A and Amber 5500 instead of SoundTech, mine has digital display, IEE488 so I can automatize the whole process, and also IMD features). It is very difficult to measure 1ppm, unless you use the second channel in the 3563A as a reference and measure the input signal residuals. You then have to substract this from the DUT response (also using the DUT gain). You would have to do this manually (pretty tedious) I'm doing it by processing the data in a PC. I went down to -130dB this way... What is really critical is the grounding of the equpment and DUT, this can make a huge difference at these levels. Noise is also critical, you can't afford any neons, TV sets, fluorescent light bulbs, etc...

If you can reproducible measure -120dB with your equipment, that's pretty good.

I still fail to understand how knowing the peak beta of the npn output devices in AD797 will help you in making a better design.

BTW, care to share how much 7th has the JC-1 power amp?
 
john curl

john curl

Member
Joined 2003
SYN08, thank you for your useful correspondence.
Peak beta (might) help me optimize the fet current that I might select. True it would be a second order effect, but we are measuring pretty deep and it might be important. I MUST ignore all external sources that show up on the screen as well, so I don't usually try to make absolute measurements, but just note differences, and I can IGNORE with my own eyes, what is NOT on the 'line' so to speak. You should be able to do that as well, with your HP 3562 in linear sweep mode.
Also, I originally bought the 3563 in order to measure 7th harmonic in power amps, and reduce it to minimum. The JC-1 in high current mode is pretty darn good, and in low current mode, is measurable.
 
Steve Dunlap

Steve Dunlap

Member
Joined 2008
Bob Cordell said:



Hi Steve,

Thanks for this explanation; it seems to make sense. I guess you are mainly referring to the capacitive coupling from the hot-side part of the primary windings to the core.

This got me thinking, though. Although neutral on a mains circuit can be quite dirty, we usually do tend to think that hot is worse. If the issue is capacitive coupling from the hot parts of the winding, we have a choice of whether to allow that coupling to go to the core, which will be typically at earth ground, or to the secondary, which will typically be related to the circuit's star ground.

Perhaps I read you wrong, but I kind of got the impression that you were saying that it is better to have the hot part of the primary windings away from the core, resulting in less coupling of garbage into the core. But this then means more coupling into the secondary. Neither choice seems good, but if the issue is capacitive coupling as you have described, which type of coupling do we think is less bad?

Thanks,
Bob
Click to expand...


Hi Bob,

Although I have (thought) I heard the difference in audio equipment, I was speaking from my experience working in the engineering department of a transformer manufacturer. We made transformers up to 15KV. Winding geometry was considered a safety and reliability issue. It was company policy to polarise the primary and secondary so the hot end of the winding was as far from the core and the outside as possible. For the higher voltages this required extra insulation between the primary and secondary. This was considered an acceptable cost trade off if it kept customers safe.

At the higher voltages, the leakage (capacitive) can be significant to lethal. At voltages used in most audio equipment leakage should not be a safety issue. Here I prefer to wire the primary so the neutral is next to the core. I think it sounds better. I have no proof of that.
 
B

Bob Cordell

Member
Joined 2006
cordellaudio.com
john curl said:
Actually, for working audio engineers, Bob, I have been very helpful.
The situation is this:
My ST analyzer, has been modified by Scott Wurcer and me to perform better than its stated specifications. This has been due to using Scott's AD797 IC's as the input to the analyzer, and Scott's 744 IC in the oscillator section. I could probably replace a few more IC's, but so far, I can get what I want, once I analyze the residual with a HP 3563 FFT based analyzer with a 256K clock, so that I can signal average, and screen out the distortion products
Of course there is a little 2'nd and 3'rd residual, but they are at .001% and below, so I ignore them and concentrate on the 5th, 7th, and 9th harmonics. Ideally, I can have fairly little higher order harmonics, and I can test this with a bypass test using the best measuring cables in my possession and noting the residual. I can always subtract this residual from the actual measurement, if I chose to do so, but I have not found it that important to do up to this time.
Now, what I am having trouble is a nagging residual 7th harmonic that I would like to reduce or remove, so that I don't have to do differential subtraction, either in the test equipment or in my head. Then when I spot some 7th harmonic on the screen, I will be able to generally attribute it to the DUT, rather than the test equipment. It is more a matter of convenience to reduce the 7th harmonic, but it might be a useful improvement.
One approach that I have NOT tried at this point, is to force the output stages to run Class A, by adding a fet current source on the output to the - supply, and hoping for an improvement. Of course the 2'd harmonic might rise slightly, but MAYBE the 7th will drop.
As Scott Wurcer has, at least, past experience with his own IC output stages, I thought that he might give me some useful input in doing this.
Repeatedly, I have asked Scott for the current of peak beta of the npn output device, but so far no luck. I would make a difference to me, if it is .5 ma , 5 ma, or 50 ma. From my perspective, all 3 are possible from my experience with discrete devices.
Now, have I made myself more clear to both you and Scott?
Click to expand...


Hi John,

Thanks for explaining. The HP3563 looks like a nice piece of equipment. So you take the residual from the Sound Technolgy into the 3563? That's a good approach. I do that as well, but I only have an HP3580A that only goes up to 50 kHz. The 3563 goes up to 100 kHz, so I gather you can see the seventh harmonic of fundamentals up to about 14.5 kHz. Is that correct?

How big is the 7th that is coming out of the residual output of your SoundTech when you run the SounTech osicillator into the SoundTech input?

Does the SoundTech force an 80 kHz residual filter on you, or can you feed a larger bandwidth into the 3563?

Thanks,
Bob
 
Johnloudb

Johnloudb

Member
Joined 2008
john curl said:

I think that lead problems are HIGHLY overrated, and while there may be a sincere need to remove it from the environment, lead solder just isn't that much of a big deal.
In any case, we are generally in your camp.
Click to expand...

Lead paint can be very hazardous to kids, especially young kids who eat paint chips. I've heard about kids getting brain damage, because of paint dust caused by opening and closing of painted doors and wood framed windows. Land fills get loaded with lead and are often turned into housing development projects down the road. So, I think it is a valid concern. In fact the elementary school I went to was built on a landfill. Yeah, I turned out fine ..... hmm .... I had real hard time with elementary school, remedial math and reading ... it must of been the lead.
 
B

Bob Cordell

Member
Joined 2006
cordellaudio.com
john curl said:
Bob, for my tests, I have settled on 5KHz for general evaluation. It works very well, AND it is a real frequency that can be fairly high in level in real program material, on occasion. The main thing is to be able to measure DIFFERENCES in designs and settings.
Click to expand...


Hi John, that's fine. Being able to see down to -120 dB out to the 7th is quite good. It also makes it easier for me to make comparisons with your results, since my spectrum analyzer only goes out to 50 kHz, which would be the 10th harmonic. I'll check my setup under those conditions.

Bob
 
Johnloudb

Johnloudb

Member
Joined 2008
Joshua_G said:
Lead paint IS hazardous. John Curl referred to using lead solder.
Click to expand...

I know. And, most of it ends up in landfills and later housing projects. Kids go play in the dirt - you get the idea.

john curl said:
The dangers of lead paint have been known for at least 60 years. It was banned many decades ago in the USA.
Click to expand...

Yes, and it's still causing problems in some older homes. Where it's been painted over but still there and gets rubbed off from wood window frames and doors. Not a frequent problem now days, just an example.
 
anatech

anatech

Administrator
Joined 2004
Paid Member
Hi Bob,
I often look at the residuals from my distortion analyzer. The HP 339A has switchable filters at 30 KHz and 80 KHz. Left out, the bandwidth is very wide. After clobbering the fundamental, a spectrum analyzer's range is extended down somewhat. The spec for the 339A is 0.0018 % residual. Not the best, but not too shabby either.

I use an HP 3585A and I have a (broken) 3580A. The 3580A is nice because it has the adaptive sweep. Cool feature that allows you to ignore the noise floor (by setting a higher limit) and so sweeps take far less time to do. Some day I might get lucky and find a 3562A, like syn08 has.

Some audio sound cards have very good response and a high sampling rate. You may be able to get up to 96 KHz with a 24 bit card. That may work better than a 3580A once you have the interface worked out. The Sound Blaster (or Creative) X-Fi Music shows promise as it isn't bandwidth limited. The EMU 1212 should also be affordable and also has balanced in and out. I almost had mine worked out, but a Microsoft update cured me of that effort. One day I may be inspired to try again.

I hope there is an idea in there for you.

-Chris
 
john curl

john curl

Member
Joined 2003
Back in the old days, like 40 years ago, we used to think about transistor design in terms of beta, almost exclusively. Transistors were CURRENT magnifying devices, not really voltage magnifying devices, back then, in the minds of most design engineers.
The little amp that I showed you Scott, operated at the BETA PEAK of the output devices. Now I know that you saw the amp, because you and I are in a picture with the amp.
Of course, Ft is also important, but it usually is close to the beta peak, so if you know one, you usually know the other. Usually beta peak is an easier number to find than Ft peak, so I asked for it. Of course VOLTAGE drive reduces the beta nonlinearities' effect on the total distortion, but since you are current driving the output stage, MAYBE, JUST MAYBE, peak beta could be important.
For example, I once VOLTAGE DROVE the HA911 IC op amp and REDUCED THE output stage distortion at least 20 dB. I'll have to tell you how I did it, sometime.
 
syn08

syn08

Banned
Joined 2005
synaesthesia.ca
scott wurcer said:
Me three. This number is almost meaningless, peak Ft might be of interest but in this context it matters almost not at all.
Click to expand...

If I understand correctly, John wants to add a JFET current source at the output of a AD797, to move the output stage to class A. Knowing the beta peak will help him in calculating the current source.

Which opens another can of worms, about the benefits of moving half of the output stage to class A. If you ask me, zero, zip, nada, jack ****. Does more damage than benefits.
 
anatech

anatech

Administrator
Joined 2004
Paid Member
Hi syn08,
Which opens another can of worms, about the benefits of moving half of the output stage to class A.
Click to expand...
Since this method has been pushed in audio lore for a while now, I'd like to see if John does get his desired results. I have never tried this either, so I'm interested to see what happens. My personal feelings are that the chip designers would have done everything they could to provide the best performance. They are paid to do that after all. Who knows? John may be on to something here.

Hi John,
Any comment to my question about whether a BJT or J Fet CCS may work best? Which one has the highest output impedance?

My questions on this are relevant I think. You are measuring with a very low noise floor. I see the CCS type should directly affect your PSRR, and hence your instrument's residual noise (or amp, whatever device it's for).

-Chris
 
Status
Not open for further replies.
Top Bottom