Quite high- the contact resistance is a few milliohms. And no, no more so than a spade lug and a binding post. You do understand that you already have that kind of stuff in your signal path?
If the particulate nature worries you, you could always use thermally sprayed zinc.
Earl, my wife prefers platinum.
No. I was unaware of that. Thank you for the link. I now have much to consider...
Cheers, John
Hi John,
I think we both agree that a 1 uH inductor is inaudible. There is a huge difference between 1 uH and 0 uH. With good design practices, an amplifier can usually be made to be stable into virtually any load. With 0 uH, it is virtually impossible to do that, especially without design changes that may seriously compromise the amplifier in other respects. Using overly-big base stopper resistors to hope to make an amp stable without a coil is a good example of such a bad compromize.
What many people do not realize is that stability is not just a matter of global stability concerning the global feedback loop. It is also a matter of local stability of the output stage. Emitter followers love to oscillate when loaded capacitively. The situation can sometimes be even worse when a very fast output stage is loaded with an unterminated transmission line. Indeed those who make no-feedback amplifiers are just as susceptible to this second local form of parasitic oscillation if they do not use a coil.
Another thing that many people miss is the way in which the collectors of an output Triple are connected together. Many people just connect the collectors of the pre-driver, driver and output transistors together to the same rail. This is asking for trouble. Even a tiny amount of common inductance can allow strong HF currents from the output transistors to cause HF voltage to get back to the earlier collectors and create oscillator topologies. The solution for this is simple: put a little bit of R-C filter in the rail between the main part of the rail and the part that supplies the driver and pre-driver. One ohm shunted by 0.1 uF is all that is needed. Note that sucha na arrangement does two nice things. First, it prevents HF from the output transistors from making it back to the collectors of the pre-driver and driver; secondly, it acts as a Zobel network on the main rail to damp out any resonances.
It is also not unreasonable to put such an RC filter in the rail between the driver and pre-driver for the same reasons.
I disagree that using very fast output transistors in the output stage allows one to get away without an output coil. Such transistors are wonderful and I love them, and they tend to reduce the chances of global oscillation, but they enhance the chances of local instability. A transistor that is ten times as fast only needs 1/10 the amount of stray inductance in the right place to become unstable.
One more thing that is important if all you are using to assess stability is a square wave into a capacitive load: perform the test at a very large variety of peak-to-peak squarewave levels. This takes into account the fact that transistor parameters vary strongly with voltage. For the same reason, the testing should be performed with a variety of resistive loads in parallel with the load capacitance, from infinity down to about 2 ohms, so that the current swing is also exercised.
I believe that it is a well-intentioned but unnecessarily risky design practice to not use any coil at all in the output of a power amplifier. A 1 uH coil with, say 2 ohms across it, will be inadible IF it is implemented properly and carefully.
Cheers,
Bob
Touche!! Good point. At 20 nH/inch, 5 inches to the output yields 0.1 uH.
Cheers,
Bob
A good move for consumers and perhaps something that will pull back the snake oil (cable) merchants:
New consumer laws have come in to play in Australia. If producers make claims about products in advertising, then they now have to be able to back them up. So rather than the ACCC (Australia's consumer rights regulator) having to prove that the claim is false or misleading, the party making the claim has to show that they have the evidence to back up the claim. That is – don’t make a claim unless it is true, and you can show it is true.
Of course the argument will then turn on what is sufficient "evidence". But it would appear that hand waving claims and other non-objective data would be on shaky ground. We can but hope.
New consumer laws have come in to play in Australia. If producers make claims about products in advertising, then they now have to be able to back them up. So rather than the ACCC (Australia's consumer rights regulator) having to prove that the claim is false or misleading, the party making the claim has to show that they have the evidence to back up the claim. That is – don’t make a claim unless it is true, and you can show it is true.
Of course the argument will then turn on what is sufficient "evidence". But it would appear that hand waving claims and other non-objective data would be on shaky ground. We can but hope.
you could always file back the solder/nickel coating on the lugs, then use solid cable and braise the coppers together......IF the paxolin/fibre board can take the heat....
Bob Cordell, please make something real for the marketplace. Have thousands of them out there, then we will talk further. To imply that I have not been careful is to impugn my reputation. Now prove it or cease the inference.
I used to use tungsten speaker wires, but found them inconsistent and annoying. At low powers they had a pleasing warm feel to them, but at higher powers they got much too "bright."
And they kept trying to coil up.....
Hi John,
I just disagreed with some of what you said, and gave technical reasons why. I also opined that not using an output coil is not a good design tradeoff. You are not the only audio designer who doesn't use an output coil.
Every engineer makes choices that involve tradeoffs of risk and benefit. I just don't agree with some of the ones you have made. However, I salute you for making your amplifier rugged enough and protected enough that it can go into oscillation without destroying itself. The JC-1 is a very good amplifier. No amplifier is perfect. Just because I don't agree with your design choice does not mean the JC-1 is not a great amplifier. Let's just say the it would be even greater if it had a 1 uH - 2-ohm R-L network in its output.
Cheers,
Bob
Can we make it in an eBook format?
That would be so convenient, considering I have a new NOOK eReader.
(just do not say it out loud - you will regret it.)
I love my new NOOK eReader!!
🙂
~kidding!~
Bob, I experimented with 1uH and even 1/2 uH inductors with 1 ohm series resistors, more than 20 years ago. Dr. Hawksford, remember him? He did the same thing. However, today, the coils, in my design, seem to not be necessary so long as some series output wiring is present. You are only bringing up, as you have done repeatedly over the years, to impugn my design reputation. IF you have something important to warn me about, why not E-mail me, or tell me in PERSON? We had a beer together last October, why didn't you give me some hard evidence then, or send it on to me later? Why go on the subject here, without clueing me in, on this website?
I should have stated: A resistor in series with the output, but parallel with the coil. Sorry for any confusion.
John,
This renewed discussion started with someone linking to info about instabilities with some JC-1 (I think, maybe JC-3) with specific types of speaker cables. (Very high capacity).
That's unfortunate, but it can happen. Some speaker cable creators go through incredible efforts to make cables that somehow sound different, even when they cause instabilities!
I believe it is very difficult to make your amp totally immune for everything fools throw at it, if you don't include some inductance, and the piece of wire to the speaker jack apparently is not always sufficient.
It's difficult to make something foolprove because fools are so ingenious 😉
jd
This renewed discussion started with someone linking to info about instabilities with some JC-1 (I think, maybe JC-3) with specific types of speaker cables. (Very high capacity).
That's unfortunate, but it can happen. Some speaker cable creators go through incredible efforts to make cables that somehow sound different, even when they cause instabilities!
I believe it is very difficult to make your amp totally immune for everything fools throw at it, if you don't include some inductance, and the piece of wire to the speaker jack apparently is not always sufficient.
It's difficult to make something foolprove because fools are so ingenious 😉
jd
One way to solder aluminum wire is to tin it in an ultrasound-activated solder bath. That's what Wharfedale was doing 50 years ago, anyway. Perhaps there's been some progress since then?
John,
Can't you just leave it at "You design for your market, I for mine and engineers can disagree."
Why don't you do something shocking and offer an .... (I can't use the actual word here it might shock some flamers into a heart attack!)
I use a lot of professional amplifiers. Do you have any idea what happens when you run 100 amplifier's outputs into the same conduit for 1400 feet!
Different strokes for different folks.
ES
Last edited:
Soldering or painting aluminum is a bit of a challenge. It oxidizes so fast it is scary. When we paint aluminum we remove the oxide with a lye solution. It then has to be dried and painted with a special acid wash primer within 15 minutes or else the paint doesn't really stick.
There are special fluxes for soldering aluminum, I have tried them and they have never worked for me. Again you can sand a clean spot or use lye, then you have a few minutes. The best I ever get is a weak joint compared to copper.
However the suggestion to use aluminum was a joke referring to the fact that it is whiter (less color actual) than copper.
For those that wish to try coaxial cable, Commscope makes a special RG59 with a silver plated center conductor. It is intended for use at the cable companies RF headend.
Last edited:
I agree 🙂 Engineers make fact based design decisions (that's my hope at least) and these need to be communicated to the end user. Engineering is not some form of practiced magic although in audio it is often sold as such ("Designed by the legendary ...", "With the legendary sound of ...", etc.).
Best, Markus
- Status
- Not open for further replies.