I just measured the residual distortion of a 1/4 W 1K metal film resistor using the CLT-1. With 10V rms @ 10 KHz excitation I got about 600 nV of 30 KHz out. (around -145 dB or so). The system residual is about 15 dB lower. The spectrum is essentially just 3rd harmonic, very little 4th +.
Per John's note as the level increases the distortion increases very quickly with drive and decreases very quickly with drive.
I'll check some more different resistors tomorrow or the next day and try to provide more detailed info.
Per John's note as the level increases the distortion increases very quickly with drive and decreases very quickly with drive.
I'll check some more different resistors tomorrow or the next day and try to provide more detailed info.
John
Thanks for the book reference, first they set up the transfer equation for a follower amplifier and then show by that system's gain equation that the 3rd harmonic distortion increases by the 2nd power of the amplitude of the signal. Also the freakin book is now over $100!
I have attached two more plots, One is of a 1% MF 1000 ohm resistor at 7.5V vs 15V (RMS) it is at 2 frequencies 100hz and 1000hz. Note the distortion increases with decreasing frequency. This does note seem to come out of the analysis technique as shown in the book.
The second is at 1000 Hertz 16 Volts RMS
Both of these plots were done using all the same type and make of resistors. Ten resistors were used. Two groups of four were series paralleled to make the value stay the same but the wattage increase. These were then matched with plain 1/4W MF 1% 1K resistors. The classic Wheatstone bridge circuit was used to get the maximum change shown.
It is OK for someone to re-post this image as a pdf back to this web site only. (Thanks)
Scott
Note that the 5th and higher harmonics do not really show on the cleaner plot. The AP2 maxes out at 32K point FFT and 4096 averages and takes forever.
I also assume you got snowed in also!
Last edited:
No, I was having drinks with Kirkland and the guys from ESS in Los Alos
.The distortion will increase at lower frequency because of less thermal averaging. I repeat a good thermal model and separation of real voltage coefficient from the thermal effects would be very useful.
Aren't you on a PC? If so, just go to Microsoft's website and download the free Microsoft Word Viewer.
Otherwise, here.
Simon7000-01.pdf
se
Last edited:
Thanks.
Sorry. It saved out with a capital S on Simon and when I put up the URL here, I used lower case.
It's fixed so it can be grabbed from either URL.
se
No, I was having drinks with Kirkland and the guys from ESS in Los Alos
The distortion will increase at lower frequency because of less thermal averaging. I repeat a good thermal model and separation of real voltage coefficient from the thermal effects would be very useful.
Do you have a good thermal model! ? Right now is just play time. My feel says tinker around until something pops out at you.
I suspect that I am not finding any unexpected voltage discontinuity effects.
Pavel,
The 145db decreases as you go down in frequency. I am now using 333 hz as my test frequency.
The hearing threshold is around -10db at 3khz. Un-equalized music has about 10db more energy at 300 compared to 3k so there should be at least a 20db compensation factor for a 9th harmonic under those conditions. Some folks even put a 10-20 db boost on the low end!
My design goal as mentioned before is 160 db for the entire system. I would like individual components to be better than that.
My design goal as mentioned before is 160 db for the entire system. I would like individual components to be better than that.[/QUOTE]
Hello Simon,
What happens when you put a couple of resistors in series, the THD must obviously be reduced. This surely must be a solution to this problem.
I respect your high standards for THD performance.
Regards
Arthur
Hello Simon,
What happens when you put a couple of resistors in series, the THD must obviously be reduced. This surely must be a solution to this problem.
I respect your high standards for THD performance.
Regards
Arthur
- Status
- Not open for further replies.