Neither do I, but is rather funny to see how things evolved from heavily copper plated component leads (as being bad for audio, according to Mr. Marsh) to magnetic PIM effects in GHz range microwave circuitry
.For the rest, you can see the effect of just a .1 OHm resistance of a ferrous metal on distortion here at DIY audio by another. SS forum under Distortion in Low Resistance ww resistors. Its small but its there.... 10 times higher than without ferrous alloy wire used in ww. Could just as well be same .1 Ohm in another form.
If one wants to get really low THD, other than circuit design in SIM is needed. That includes measuring plated parts on ferrous metals, contact materials and the like. Each should be looked at for their contribution... or just dont use them.
[Note: My frame for reference is performance -.001% and less.]
THx-RNMarsh
your turn.
If one wants to get really low THD, other than circuit design in SIM is needed. That includes measuring plated parts on ferrous metals, contact materials and the like. Each should be looked at for their contribution... or just dont use them.
[Note: My frame for reference is performance -.001% and less.]
THx-RNMarsh
your turn.
Last edited:
I use the Dale th resistors as they have the lowest distortion (and do not have copper weld leads.) but even smd resistors -- leadless -- have to be measured and not assumed. Because of the metals used and mfr processes etc etc.
www.davehilldesigns.com/smt_resistor_distortion_rev1.pdf or type it in on your web browser.
Thx-RNMarsh
www.davehilldesigns.com/smt_resistor_distortion_rev1.pdf or type it in on your web browser.
Thx-RNMarsh
Last edited:
How about this, Waly?
Analog Devices: Analog Dialogue: Ask The Applications Engineer - 22
I consider it an industrial definition of CFA x VFA. Plus many Analog Devices papers published in nineties.
I still do not understand this s....d semantic battle.
Analog Devices: Analog Dialogue: Ask The Applications Engineer - 22
I consider it an industrial definition of CFA x VFA. Plus many Analog Devices papers published in nineties.
I still do not understand this s....d semantic battle.
Links not working -- dont work for me either from DIYAUDIO ... who knows why. But it does work if you open it elsewhere -- like at your web browser line at the top of your window display. If I figure out why I'll correct.
-RM
-RM
go to David Hill Designs.com and then to SMT resistor distortion under 'tech' subject.
-RM
-RM
Last edited:
HTML:
http://www.analog.com/library/analogdialogue/anniversary/22.htmlPlease try ctrl+c and ctrl+v
VC?
Thank you-- I tried to remove it several times but DIY keeps adding it on for me.[noted that 'resistor' is spelled incorrectly but only works that way]
Many film resistors use an alloy which has ferrous metal ingredients. There are many reasons for doping them... altering tempco, is one. But, raising the voltage also raises the current thru the R --- The greater the current density, the greater the distortion produced.....
Thx-RNMarsh
Thank you-- I tried to remove it several times but DIY keeps adding it on for me.[noted that 'resistor' is spelled incorrectly but only works that way]
Many film resistors use an alloy which has ferrous metal ingredients. There are many reasons for doping them... altering tempco, is one. But, raising the voltage also raises the current thru the R --- The greater the current density, the greater the distortion produced.....
Thx-RNMarsh
Last edited:
You really need to remove thermal issues from SMT components before considering anything else. Remember we make <10 ppm instrumentation amplifiers with tiny thin film resistors 400 atoms thick. IMHO passive component issues in audio are a non-issue, the few cases that might apply are known and can easily be avoided without resorting to any bespoke components.
What thermal issues? Why are the resistors on Marsh's document distorting and what can be done to fix it?
What about using a resistor string? 10 in series would divide that distortion by 10. For high value resistors paralleling may be a bad idea. For instance if each resistor has 0.33pF capacitance, and you use 10 560k resistors for a 56k resistor, you'd have 3.3p parallel capacitance at least, which makes for an 800KHz RC which can be significant for the compensation.
What about using a resistor string? 10 in series would divide that distortion by 10. For high value resistors paralleling may be a bad idea. For instance if each resistor has 0.33pF capacitance, and you use 10 560k resistors for a 56k resistor, you'd have 3.3p parallel capacitance at least, which makes for an 800KHz RC which can be significant for the compensation.
Last edited:
To name a few:
- Thermal coefficient
- Voltage coefficient
- Excess noise (when a voltage is applied), interpreted as distortions.
But definitely nothing to do with the magnetics of any ferrous content in the materials. Boutique resistors ("naked foil", etc...) are as much a toxic legend as boutique gold foil capacitors for audio.
This doesn't preclude using good quality components, abundantly available at European and US distributors, for very cheap.
In the literature, the thermals have been sorted out/normalized and left with similar results. I have read these papers. IIRC, the thermals were not the answer to distortion cause. It makes for precise delineations of what is contributing what and why.
I know the low signal levels/currents used in opamps reduce distortion into the dirt. Power amps are another story. People measure at much higher Volt/Amp levels and with different parts, they can get significantly higher levels than you quote.
IMHO, also, it isnt a big deal for 99% of the designs. Many of the power amp designs in SIM here will have significantly higher currents (and voltage) than an instrumentation opamp and these SIM's indicate ppm levels possible with the circuitry. I am not saying its the end-all for everything. Just for constructors to be aware of it when both high currents and extremely low distortion is attempted.
THx-RNMarsh
Last edited:
There seems to be a lot of noise coming from you for your great distain for marketers and sales people in High-End audio. I agree with that, to a large extent. However, it just seems misplaced here as no one has promoted gold foil caps or naked foil resistors HERE. I'm not.
THx-RNMarsh
I wasnt a member of this DIYAudio site more than a year and a half ago. Dont blame me for what damage others did. I think the comments so far have been reflecting what so many people have said about the characteristic sound - if there is one, being different. So, we have strived to understand the CFA topology and its pro-con. It has yet to be listened to in many cases of the designs here in the past few months.... still in build phase. I/we only wonder IF THERE IS an explanation for such common listening results of CFA vs VFA and what it might be. I dont think this thread has promoted mysticism. Just SIM and measured numbers. So, I think you are barking up the wrong tree over here. Help us design CFA. Supergood CFA. And, make them work equally as good as the SIM.
Thx-RNMarsh
Last edited:
Hi, Richard,
I think you are right (as usual) with both of your inputs about iron and bridged amps.
I don't know if iron chassis has a great influence on perceived sound in an power amp, but, anyway, when both measurements and theory lead-us to avoid-it, why don't prefer coper wires and aluminum chassis and try our best ?
Good results are often a sum of little steps and cares.
It will not change a lot the price of our DIY baby and both will behave better with age and oxidation.
About bridged amps, anybody who try to listen the same amp in both configuration will prefer, (i believe) the bridged one. As the only negative issue is the damping factor, yes it is something to look at (a bridged CFA).
Somebody previously asked a question about the slew-rate: let's take an example. Each mono amp have 500V/µs of slew rate, the result will be double (not half), and the speaker will see 500Vx2 in this µs. One of the benefits (if any;-).
I think you are right (as usual) with both of your inputs about iron and bridged amps.
I don't know if iron chassis has a great influence on perceived sound in an power amp, but, anyway, when both measurements and theory lead-us to avoid-it, why don't prefer coper wires and aluminum chassis and try our best ?
Good results are often a sum of little steps and cares.
It will not change a lot the price of our DIY baby and both will behave better with age and oxidation.
About bridged amps, anybody who try to listen the same amp in both configuration will prefer, (i believe) the bridged one. As the only negative issue is the damping factor, yes it is something to look at (a bridged CFA).
Somebody previously asked a question about the slew-rate: let's take an example. Each mono amp have 500V/µs of slew rate, the result will be double (not half), and the speaker will see 500Vx2 in this µs. One of the benefits (if any;-).
Last edited:
Is it possible to have an amp where the gain is variable by placing a pot in the feedback shunt position, and being stable with good step response across the gain range? And not just stable, but relatively consistent step response and phase margin? I think this should be possible for both VFA's and CFA's.
BTW, why would anyone call it a CFA and risk all that drama on the name? Why not just call it emitter-feedback or source-feedback? I see far less potential contention in that.
BTW, why would anyone call it a CFA and risk all that drama on the name? Why not just call it emitter-feedback or source-feedback? I see far less potential contention in that.