John Curl's Blowtorch preamplifier part II

[PMA]

Your customers would not complain about RFI - they know nothing about it. Too late when they have demodulation, and they have not I am sure.

The case is, that this is a Blowtorch thread, i.e. the best possible solution. You could improve RFI suppression considerably, and maybe get even better sonic result then you have ever obtained.

I like your circuits and ideas, but hate boasting and patronizing and simplifying.

 

[AndrewT]

Would it be appropriate to place an R-C Zobel with a corner of some very high frequency on the output of a pre-amps line stage to help reduce any RF that might be coming in the output? If so what would be a good F3?

Also how about a small value air core inductor in the output?

I think yes.
Some pre-amps have this in the design.
I am building one just now (Diamante) that has the L//R on the output.

 

[john curl]

Sorry, PMA, I like to boast, it is all that I have left. I have to simplify in order to include everyone else, not YOU. I just got off the phone with Dick Sequerra and I just told him that I was so frustrated with this thread, because I have to simplify everything, that I get in the bad habit of doing it when I am talking to either him or Mitch Cotter, and they don't need simple explanations.

Scott, Dick and Mitch say HI, and would like for you to call them, sometime. I asked them both to get you off my ankles, by clueing you in on what I am TRYING to say.

 

[KSTR]

For example, I don't think that people should put their power amps on a shelf over their salt water swimming pool, only to have it pulled off by their pet monkey into the pool and perhaps hurting somebody. However, in this litigious society, the designer would be blamed anyway, so let's obey the rules, everybody.

http://www.inside-blog.de/uploads/2008/01/pool1.jpg
Stuff like this is *not* very funny, indeed.

 

[KSTR]

Sorry, PMA, I like to boast, it is all that I have left. I have to simplify in order to include everyone else, not YOU

Then maybe the Blog idea brought up recently is a good one, there you could express yourself freely at whatever level you choose, without getting interrupted by questions of less knowledgable people. I truly mean this in a non-offending manner. A lot of people would appreciate such a J.C. blog, me included.

- Klaus

 

[SY]

"Think of it as evolution in action."

 

[PMA]

I just got off the phone with Dick Sequerra and I just told him that I was so frustrated with this thread,

Same everywhere, even in our local forums (have had a similar call with Pavel Dudek /Upupa Epops/ today). These are public forum rules, we have to live with it, once we decide to join.

 

[Bob Cordell]

This is one place where a 'lead' cap in the feedback might be questionable. Think about it, everyone.

Hi John,

I agree. I have always avoided the temptation of using the lead capacitor in the feedback path for this very reason. I seem to recall someone else pointed this concern out a long time ago. It might have been Neville Theil.

Cheers,
Bob

 

[Bob Cordell]

Hi John,
Most people live in areas where there is a fair amount of RFI. Just think of all those cell phones for example. Switching power supplies, florescent lamps (in a neighbor's house for example). This is a concern for almost everyone in today's world.

RFI can cause any design to become non-linear. They can overdrive some stages well above your hearing, but you will hear the effects on your audio signal. Even with some systems with no active music input, you may be able to hear the effects if you listen close to your speaker. It varies.

Where my business was, there was a location in our area where a strong AM station was operating, and still is to this day. The RFI was so strong that many CD players would not operate, they couldn't track at all. Phono preamps were constantly distorted due to this, and some power amplifiers were sent to amp heaven. So, the effects of RFI can certainly be destructive, depending on the field strength. A couple recording studios I did work in were also strongly affected. The Metalworks in Mississauga had to create a Faraday cage around Studio A. It was wrapped in sheet copper, not expanded metal or perforated metal. A solid sheet of copper that was grounded and connected to the room ground. Technical grounds do not help in these situations.

I think too, you have to consider both the material and value of inductors and capacitors used to bypass RF in the audio range. A ferrite bead, as commonly used (not really large at all) has very little effect at audio frequencies. It's also not going to be saturated unless you have very real big RFI problems. Also, the use of NP0 / C0G ceramics at small values will not present a problem to your audio signals. You can always use Teflon® or mica capacitors without too much fear. If you are really concerned, use a Johnson (or equivalent) screw adjustable capacitor and run your lead through that instead. It will have external metalization already, solder to that. I'm talking about something like this, or this and similar products. You can see that you can run a relatively large conductor through these with minimal effort.

Once you look at the idea and consider the values of components under consideration, I'm pretty sure you will not think you are 'throwing the baby out to change the bath water'. Off the cuff, that view is pretty closed. Always remember John, "Condemnation without Examination is Prejudice". I think that applies here. The ideas presented here are valuable ones from where I sit. Many of your customers may not have the advantage afforded by living in an RFI free zone, not in this day of cell phones and (horror) Wi-Fi signals. Even AM radio presents a very real threat to musical enjoyment.

-Chris

Edit:

The idea is correct, but you are assuming that the output is low impedance at RF frequencies. That is where your idea may fall down.

Hi Chris,

These are all good points. One of the things I have often wondered about with RFI ingress problems is the following thought question: If the amount of RFI ingress is sufficient to upset circuits and cause intermodulation with an audio signal, wouldn't most forms of RFI ingress also result in an audible disturbance in the absence of the audio signal? Obviously, if the RFI ingress is merely an unmodulated carrier, this will not be the case, but that is the exception rather than the rule.

Its interesting to note that the old reliable cell phone test produces very audible artefacts.

Cheers,
Bob

 

[Netlist]

Sorry, PMA, I like to boast, it is all that I have left.

Noblesse oblige.

/Hugo

 

[Bob Cordell]

Rick, the best that we have found is high speed, high frequency linear open loop, low or no global feedback designs. Then, a little RF in the loop is not too bad. Kind of like designing a video amplifier.
IF it is an entirely open loop design, like the Ayre, it is relatively easy, because the high speed low open loop Z output stage, absorbs the RF coming in, just like an attenuator. If you use global feedback, then a very linear, high speed design is necessary. This is one place where a 'lead' cap in the feedback might be questionable. Think about it, everyone.

Hi John,

It is usually not going to be the case that the high-frequency open-loop output impedance of a non-feedback amplifier will be lower than that of a feedback amplifier.

While it may be true that the open-loop output impedance of a no FB amplifier is lower at audio frequencies, the usual Miller compensation will typically make the open loop output impedance of the feedback amplifier at RF just as low as the amplifier without NFB (maybe even arguably lower).

Cheers,
Bob

 

[pooge]

Hi John,

I agree. I have always avoided the temptation of using the lead capacitor in the feedback path for this very reason. I seem to recall someone else pointed this concern out a long time ago. It might have been Neville Theil.

Cheers,
Bob

Not only that. Leach states that lead compensation may be unreliable.

"[Lead compensation] is normally applied in the feedback network to correct for phase lag in the output stage. Because this lag is a function of the load impedance, lead compensation can be unreliable if the load impedance changes."

 

[Joshua_G]

Originally posted by PMA
The case is, that this is a Blowtorch thread, i.e. the best possible solution. You could improve RFI suppression considerably, and maybe get even better sonic result then you have ever obtained.

Any ideas as to how it may be done?

 

[Wavebourn]

Not only that. Leach states that lead compensation may be unreliable.

"[Lead compensation] is normally applied in the feedback network to correct for phase lag in the output stage. Because this lag is a function of the load impedance, lead compensation can be unreliable if the load impedance changes."

...and it changes every time the gear is moved from the laboratory for it's ordinary duty.

 

[PMA]

Any ideas as to how it may be done?

Not only ideas.

 

[john curl]

Good question, Joshua. I have some idea, in hindsight how, IF I made the BLOWTORCH again, how I could reduce the RFI input. However, I doubt that I would add either the expense (metal vs Plastic power connector) or extensive internal shielding, because I have never found the need for it.
You know, when I added shielding, I was criticized for being wasteful. When I avoided shielding (to protect the audio quality, or to lower cost) I was scoffed at for not adding extra shielding.

 

[Joshua_G]

Not only ideas.

I'd love see your suggested solutions.

 

[PMA]

You might see it later, in another thread. As you probably know, I not only theorize.

 

[john curl]

Oh, going to criticize and run? Sounds like what they accuse me of.

 

[PMA]

Similarities?