John Curl's Blowtorch preamplifier part II

[simon7000]

Originally Posted by gpapag

Scott, you missed the photos of the gaincard’s internals.

I have asked again in the past but got no answer. What would be a good safety factor (multiplier) for the power rating of a feedback resistor?



Which one? AN-348 ?

George

George

The best resistors at 1/4 rated power still have some artifacts. So I would use those at 1/10 rating. Inexpensive metal film resistors as low as 2% of power rating. Unless of course you want to do a fashion designer amplifier then a carbon composite at full rated power! You'll get great reviews.

Of course you will want to look at the resistor divider thermal noise to be sure that you aren't going too far. Now the best trick is to use multiple small resistors with JN's trick to reduce inductance. The best cheap resistors are Xicon. For a bit more Dale CMF. The Dale's in the mil spec 1/2 W size have the lowest distortion.

ES

[dimitri]

I have asked again in the past but got no answer. What would be a good safety factor (multiplier) for the power rating of a feedback resistor?

George, take a look on this application note Once you know resistor temperature coefficient you can estimate power rating

10 times increase in test power gives 30 dB higher distortion level.

[mikelm]

Iv'e started reading this thread from the beginning . . . . 100 pages read only about 5000 to go !

It's a pity that all of richard perez's links are inactive because much of the discussion is about circuits that we now cannot see.

Anyway I have a question about FR4 PCB's and I'm sorry if it has been covered already.

I wonder if anyone here has experimented with using different track sizes to see if wider tracks or or narrower tracks sound better in line level circuits. I'm not talking about 1oz or 2oz copper thickness - I mean did anyone experiment to establish if wider or narrower tracks sound best.

I'm sure that some people might have views about this topic but I was hoping to hear from someone who has actually tried different width tracks and drawn any conclusions - even if it was that there was no audible difference.

many thanks

mike

[gk7]

Originally Posted by Esperado

...
Gold, Silver, Diamonds (specially in buffers), anything aircraft or spatial, Teflon and Kevlar, carbon fibers.
Of course, the add of some species of rare wood and cotton fibers help to get a more natural sound.
All that need to be 'voiced' by audio gurus....[/URL]

The worst thing are comedians that get boring.

[john curl]

When it comes to bulk aluminum, I only repeated what I was told by my deceased colleague, Bob Crump, a number of years ago.
A simple Wiki search will give a number of designations of AIRCRAFT GRADE ALUMINUM.
Some are very strong, some are more weldable and workable. All cost money, but I suspect that some polish and finish better than others. That is why we took whatever the machinists recommended. They had the experience, we didn't.
Like a fine automobile, our fit and finish had to be outstanding, if we were to get some outside customers to buy the product, so that we did not have to assume the entire burden and cost of making the units just for ourselves. We didn't like it either when the cost of aluminum started going up, in that time period.

[qusp]

if for a high quality product of any type, audio or not, you want it to machine well and last well. 6061 or 7075 grade (aircraft designations) are those alloys you want to use, as do I use them. they machine well, are more likely to be consistent from piece to piece or even across one large piece, as well as less prone to corrosion and stronger. aluminium corrodes too, but the lower grades need anodising to avoid it, 6061 and 7075 dont. (edit: they do eventually corrode, but its MUCH more resistant)

some will still anodise it, the cheaper grades may also have 'grain' which is picked up by the process and is all the way through the piece, it shows through the anodising too because it has slightly different electrical properties

John would be silly to not use it, there may be extra finishing processes that would add up to possibly cost the same anyway.

besides its sexy as hell!! =)

[simon7000]

Originally Posted by dimitri

George, take a look on this application note Once you know resistor temperature coefficient you can estimate power rating

Dimitri,

That paper raises the issues of the different types of resistor distortion.

Second harmonic is mostly from the temperature coefficient. Tempco obviously showed up early as a problem with vacuum tubes. The testing was simple measure the value heat it up and measure again. The assumption was that the change was linear with temperature. That mostly holds true.

Third harmonic distortion is these days mostly from impurities in the resistance material or mechanical strain. Mechanical strain indicates a possible failure mechanism.

Fourth harmonic distortion is often a measurement or circuit error. It is most often caused in resistive dividers that use two resistors of different thermal mass.

Ninth order distortion is by the same mechanism but is much worse. That is because music has more energy as you go down in frequency to about 50-150 Hz. The ear is most sensitive at 3kHz or so. If you apply weighting 9th harmonic get a 50 dB headstart!

Any other distortions are usually material defects.

The other issue is of course thermo electric. Be sure your resistor heats both end connections uniformly.

There is almost no vibration introduced changes at normal levels.

There are capacitance and inductance issues at normally low levels except of course for wirewound resistors. These show up in my pictures as less suppression of the fundamental. (Although in really bad cases it is actual value drift.)

The final issue is self noise. Some resistors when DC is applied create more noise than others. This shows up on my measurements as a spreading at the FFT fundamental trace base.

ES

[albin]

I've come to think carbon get's bad press here.
and that solid ali boxes have no dificult joints.
regards albin

[Esperado]

Originally Posted by simon7000

Second harmonic is mostly from the temperature coefficient.

Did-you mean that temperature of a resistance if following a 1000Hz signal, despite thermal inertia ?

[john curl]

Thanks qusp for your enlightened and experienced input on aluminum. I read up on it, this morning, myself.
The very first aluminum box we made was welded together, by a specialist. We just could not take a chance that it would be consistent, so we went to 'hogging out' the chassis.
Jack Bybee now owns the prototype CTC Blowtorch that was welded together, with some oversized knobs that I nicknamed 'Dolly Knobs'. '-) I insisted that the knobs get less 'impressive' looking so that it didn't look like the preamp was falling over. Now we just call them 'Marilyn Knobs'. '-)