Richard:
You are using the PCB as a coupling cap? Try putting it in parallel with the load with a 600 Ohm source and see how it affects the distortion. This is more realistic, and I suspect you will see something. If you do I'll unearth the CLT-1 and see what I can learn.
You are using the PCB as a coupling cap? Try putting it in parallel with the load with a 600 Ohm source and see how it affects the distortion. This is more realistic, and I suspect you will see something. If you do I'll unearth the CLT-1 and see what I can learn.
Depends on the circuitry between trace and ground plane....... or multiple (embedded) ground planes.
But go ahead and test it your way. It isnt a great capacitance dielectric. I am sure other materials are better. Would it help an ULD generator? It might.
.... and, the thd increases with increased frequency.
THx-RNMarsh
Last edited:
Demian,
I would not assume all nodes within a circuit will be low z (600 Ohms). Nor that no current will flow thru dielectric at audio freqs (series path). For a small C across 600 ohms, I probably cant measure it as the freq will have to be too high to even be of interest.
THx-RNMarsh
I would not assume all nodes within a circuit will be low z (600 Ohms). Nor that no current will flow thru dielectric at audio freqs (series path). For a small C across 600 ohms, I probably cant measure it as the freq will have to be too high to even be of interest.
THx-RNMarsh
Last edited:
Can you repeat the measurement with an equal-valued "known good" capacitor? This would verify the residual of the test method.
Thanks,
Samuel
/OT
>fancy night
In the meantime someone has approached me to do just that and my GF, not surprisingly, likes the idea. 😉
Cheers, E.
Hi Jan,
>fancy night
In the meantime someone has approached me to do just that and my GF, not surprisingly, likes the idea. 😉
Cheers, E.
Yes, of course. Here is a 0.1mfd ultra low esr and esl polystyrene capacitor at 5KHz.
Any how... I think pcb material deserves some further investigation for ULD apps.
THx-RNMarsh
<Irony filter>
Its obvious, its the paper cap you are forming with the clip lead and the post-it note, "everyone" knows that while euphonic, paper caps have higher distortion. For the fF value this must have it will surely be significant!
</Irony filter>
Cheers
Alan
Its obvious, its the paper cap you are forming with the clip lead and the post-it note, "everyone" knows that while euphonic, paper caps have higher distortion. For the fF value this must have it will surely be significant!
</Irony filter>
Cheers
Alan
Hi Richard,
I've never realized that PCB material could have such a significant effect. So high impedance nodes should be guarded by means of a low impedance copy of the signal at these sensitive nodes, right?
Cheers, E.
The capacitor needs to have the same value as the PCB capacitance (say within 10%). Otherwise the comparison is meaningless...
Samuel
Hi David,
Thanks. However, to the best of my knowledge, most of my questions reamin unanswered and not displayed on the front panel.
For example:
1. What is the capacitance of the PCB?
2. What is the 3-dB frequency of the HPF formed by use of that capacitor in combination with the circuit resistance?
3. What is the attenuation of the 5kHz fundamental going through the HPF that is formed?
4. Recognize that, if attenuation is large at 5kHz, the HPF will boost the harmonics of the generator signal by 6dB/octave, increasing the THD reading.
These kinds of issues are critical to evaluating the significance of this experimental data.
Cheers,
Bob
Richard can you test other PCB materials.
Maybe one of the board houses wouldn't mind sending you some scraps for samples.
Guards around inputs for inverting amplifiers is ground.
Maybe one of the board houses wouldn't mind sending you some scraps for samples.
Guards around inputs for inverting amplifiers is ground.
All good points and questions, Bob. But you and others have THD instrumentation to do this yourselves. I dont intend to go into great detail...
But here are some other comparisons... same ps capacitor and pcb tested for DF using Standford Research Model 720;
...... cap -- pcb
.
100hz .00012 -- .05
1KHz .00005 -- .032
10KHz .00039 -- .026
100KHz .0014 -- .028
The DF is in the range of an X7R type cap. But I havent checked pcb for voltage sensitivity.
Here, the quality of the pcb capacitance doesnt look so good.
THx-RNMarsh
But here are some other comparisons... same ps capacitor and pcb tested for DF using Standford Research Model 720;
...... cap -- pcb
.
100hz .00012 -- .05
1KHz .00005 -- .032
10KHz .00039 -- .026
100KHz .0014 -- .028
The DF is in the range of an X7R type cap. But I havent checked pcb for voltage sensitivity.
Here, the quality of the pcb capacitance doesnt look so good.
THx-RNMarsh
Last edited:
The smaller pcb is attenuated by only 4dB at 5KHz. The larger pcb is atten much less, of course; 0.3dB. The larger pcb is also different dielectric and thinner.
THx-RNMarsh
THx-RNMarsh
Last edited:
I'll pass on trying to replicate your results.
It is disappointing that you can't at least tell us the capacitance of the PCB and the impedance surrounding it. There may be other, smarter people here that can make confident inferences from your results absent such fundamental information, but I certainly cannot. While I do believe that ordinary pcb material can be nonlinear and cause distortion, I wish I could put your experimental results in context.
Anyway, thanks for doing the experiment and causing us to at least think about the phenomenon. I suspect it is what Tektronix called "hook" in the 1960's.
Cheers,
Bob
You must be living in the middle of Matto Grosso. Otherwise I couldn't explain a clean measurement of -125dB with those lose wires.
We have some pretty sensitive equipment here, and at -126dB (that's half of a microvolt) we'd better not dare to breath around, not to mention killing all CFLs, the coffee machine, the water cooler, all our cell phones, and any other CRT, LCD or SMPS based equipment in the lab. Thanks god there is no RF transmitter around, less than 10 miles away.
PCB "hook," similar if not identical to DA, is characterized by variation in effective
circuit-board capacitance with frequency (see Reference 1). In general, it affects high
impedance circuit transient response where board capacitance is an appreciable portion of
the total in the circuit. Circuits operating at frequencies below 10 kHz are the most
susceptible. As in circuit board DA, the board's chemical makeup very much influences
its effects.
http://www.analog.com/library/analogDialogue/archives/43-09/EDch 12 pc issues.pdf
circuit-board capacitance with frequency (see Reference 1). In general, it affects high
impedance circuit transient response where board capacitance is an appreciable portion of
the total in the circuit. Circuits operating at frequencies below 10 kHz are the most
susceptible. As in circuit board DA, the board's chemical makeup very much influences
its effects.
http://www.analog.com/library/analogDialogue/archives/43-09/EDch 12 pc issues.pdf
Bob, I would expect the board capacitance of the smaller board that Richard is measuring to be around 0.001uf. Yes, it is less that 0.1uf, but it is important, nevertheless.
Last edited: