Stupid question about resonant frequency test.

[CheopisIV]

I'll add those to the list; would make sense to have the ability to hard code the scale!

I've also done some testing and have paid attention to the scope display. I'm willing to bet my 'all the same' results from earlier testing was due to clipping. Here's a shot of the exact same pickup where the original plot had clipping and the red plot shows with the scope turned down for no clipping.


And this is the same Tele neck/bridge from a few posts back but with my own setup all tuned up! (No clipping, probes on right settings, phase chart all together)


I've also found my plots are nearly identical from 1.01 to 1.09 and change quite a bit above the 1.10 so I tend to go in the .03-.06 range.

[kayakerca]

I'll add those to the list; would make sense to have the ability to hard code the scale!

I've also done some testing and have paid attention to the scope display. I'm willing to bet my 'all the same' results from earlier testing was due to clipping. Here's a shot of the exact same pickup where the original plot had clipping and the red plot shows with the scope turned down for no clipping.
[ATTACH=CONFIG]29292[/ATTACH]

And this is the same Tele neck/bridge from a few posts back but with my own setup all tuned up! (No clipping, probes on right settings, phase chart all together)
[ATTACH=CONFIG]29293[/ATTACH]

I've also found my plots are nearly identical from 1.01 to 1.09 and change quite a bit above the 1.10 so I tend to go in the .03-.06 range.

The resonant peak measurement can be close for wide frequency steps at times and other times not. Depends where the step lands relative to the peak I guess.

Here is what I was talking about with the "stretched" view.





They are the exact same plot run at 2 different screen perspective dragouts. they don't even look like the same plots. Personally I prefer the perspective from the app launch.

[CheopisIV]

Well, just when things were clearing up, it's back to the drawing board I just received an email after asking about the probe setup from the Syscomp folk.

The basic idea of the x1 and x10 settings is to change the input resistance of the scope as seen by the circuit. On the x1 setting, it's 1 Megohm. On the x10 setting, it's 10 Meg ohm, which is much less likely to upset the operation of the circuit. The tradeoff is that the signal is reduced by x10 on the x10 setting. So..if the signal is sufficient then x10 is a better choice.

There should never be any occasion when you'd need to use X10 for the Channel A input. You can use the x1 setting. The impedance at that point is very low because it's connected to the output of the generator, which is about 150 ohms. So 1 Meg is very large in comparison and loading is not an issue.

A resonant circuit can be quite high impedance, especially around the resonant frequency, so a 1 Meg load might well have an effect on the circuit. So x10 is probably preferable for the Channel B input. The acid test is whether switching from x1 to x10 makes any difference. If it does make a difference, use x10. It also makes a difference what series resistance you use: a lower series resistance might make x1 acceptable. But that also affects the shape of the resonant curve.

In the ideal case, one should really drive the coil with a constant current source (which is effectively the same as an infinite series resistance) and measure the response with something that is an infinite resistance (open circuit). The series resistance we've been recommending is just an approximation to a constant current source, and the input resistance is 10Meg ohms max with the x10 probe. If you're comparing pickups and always use the same measurement setup, then it doesn't matter. But if you want an absolutely 'correct' measurement then you need constant current drive and infinite load resistance. If there was enough demand, we could design an accessory circuit that would provide constant current drive and open-circuit measurement. It could be done with a couple of battery-powered op-amps. My sense has been that absolute measurement is not a requirement in this case, but you folks can tell me.

This talk of a constant current drive as an expansion is intriguing. It might be a valid way to standardize the way we measure pickups. I know there's not a lot of standardization in the field but if people are interested, I'd suggest we flood the Syscomp people with emails demanding they hook us up

[kayakerca]

Well, just when things were clearing up, it's back to the drawing board I just received an email after asking about the probe setup from the Syscomp folk.



This talk of a constant current drive as an expansion is intriguing. It might be a valid way to standardize the way we measure pickups. I know there's not a lot of standardization in the field but if people are interested, I'd suggest we flood the Syscomp people with emails demanding they hook us up

Interesting. . .

Go back to the oscilloscope mode and set the to XY mode and locate the resonant frequency in both 1X and 10X settings. You really can't really do it in 10X setting as the wave from (blue Channel B Input B) and the XY mode ellipse signals are so erratic that you can't determine the resonant frequency. In 1X setting the signal represent as very smooth wave forms and can easily be tuned in in XY mode to determine the resonant frequency the same as in 1X setting on Input B in the Network Analysis Mode.

I'm a little more confused now for sure. But I'm guessing that comes as little surprise to most. . .

[CheopisIV]

Lol, I've been confused from the get go so at least we're in good company (maybe!)

I'll try what you're saying and see what I find if I can figure it out. I guess the biggest concern with this is getting accurate results for our applications that we can quantify for ourselves.

[kayakerca]

Lol, I've been confused from the get go so at least we're in good company (maybe!)

I'll try what you're saying and see what I find if I can figure it out. I guess the biggest concern with this is getting accurate results for our applications that we can quantify for ourselves.

That is exactly what I'm am looking for from all of this, an accurate and reproducible result I can use to do comparisons on my pickups.

Can you PM me the email of the contact you are corresponding with at Syscomp? I wouldn't mind following up on this 1X/10X thing a bit more.

[CheopisIV]

After playing with this for a day, I'm not entirely sure the results are at all accurate on either setting. Maybe the curve is good but the peaks are still way out of whack according to the data in Lemme's book. I'm going to have to go through his suggestions again (Lemme) and see if I can shift them into the right area or at least to where they'll make more sense. I measured the capacitance of my guitar circuit from tip to ground with a 20' cord and tried a cap in parallel and series but it didn't work as anticipated. Still learning for sure, but at least it's fun

Oh, and for the record, I'm finding the readings on the 1x setting a bit more believable.

[kayakerca]

After playing with this for a day, I'm not entirely sure the results are at all accurate on either setting. Maybe the curve is good but the peaks are still way out of whack according to the data in Lemme's book. I'm going to have to go through his suggestions again (Lemme) and see if I can shift them into the right area or at least to where they'll make more sense. I measured the capacitance of my guitar circuit from tip to ground with a 20' cord and tried a cap in parallel and series but it didn't work as anticipated. Still learning for sure, but at least it's fun

Oh, and for the record, I'm finding the readings on the 1x setting a bit more believable.

I finally figured out how to get the 10X graph in oscilloscope to generate smoother sine waves. So I'm OK with that now. Between the comments from Syscomp and the Lemme book, so long as you are using the same setting, you have comparability, but that's where it ends it seems. And that's even before all the external impacts between the pickup is self and the sound coming out the speaker.

Interestingly though, if I take the resonant peak that's listed on the SD site for something like an SSL-1 for which I have a very similar pickup (DCR and A5 wise anyway), the resonant peak on mine is very comparable when Input B set to 10X, but significantly lower when set at 1X.

There is just no end to things you can learn.

[CheopisIV]

I had another email form Peter today indicating that running the scope in 1x will add capacitance to the circuit which would explain the downshift. I asked how much capacitance and am awaiting the answer... Measuring my loaded Les Paul (2 buckers, vintage cable throughout and 20' cable) I get ~400nF capacitance in either single position and ~700nF in the middle position. I'm guessing the actual load from the scope is less than this which is why my curves are still too high on the scale. Regardless, it's mostly just a shift of the same (or very similar) plot lines so I think the numbers are wrong, but the curve is still valid.

[Jazz P Bass]

'Most' scope probes add 50pf of capacitance.

[bbsailor]

'Most' scope probes add 50pf of capacitance.

See this web link: http://circuitslab.case.edu/manuals/..._Tektronix.pdf
All probes will load the measured circuit somewhat. A 1X probe loads most and when used on low impedance circuits, has minimal effect on the measurement accuracy. However, on high impedance circuits when the pickup resonance is being measured, the 1X 1M ohm load and 20 pf to 50 pf (depending on the probe wire length and wire type) will lower the resonance peak slightly and the capacitance load will lower the resonant frequency. The 10X setting will offer a 10M load but will still have some capacitance loading on the circuit being measured.

Most accurate measurements should be made at the end of the length of the coax cable you are using to connect the guitar to the amp. This way, you are seeing the pickup just as the amp hears the pickup.

See Scope Probe Fundamentals for a complete explanation.

Joseph Rogowski

[Joe Gwinn]

I had another email form Peter today indicating that running the scope in 1x will add capacitance to the circuit which would explain the downshift. I asked how much capacitance and am awaiting the answer... Measuring my loaded Les Paul (2 buckers, vintage cable throughout and 20' cable) I get ~400nF capacitance in either single position and ~700nF in the middle position. I'm guessing the actual load from the scope is less than this which is why my curves are still too high on the scale. Regardless, it's mostly just a shift of the same (or very similar) plot lines so I think the numbers are wrong, but the curve is still valid.

One can measure the scope input capacitance by hooking an Extech LCR meter to the input, with each instrument measuring the other. The Extech test signal is about one volt, at 120 Hz or 1 KHz.

[CheopisIV]

I was going to try that the next time I had it set up. Thanks for the confirmation!

[kayakerca]

I was going to try that the next time I had it set up. Thanks for the confirmation!

I gave it a try. I set the LCR to 1 kHz with the follow results.

Set probe to 1X:
Rp = .9996 MΩ (~ matches probe 1X setting)
Cp = 180.8 pF
Cs = 320.2 pF

Set probe to 10X:
Rp = 9.751 MΩ (~ matches probe 10X setting)
Cp = 41.2 pF
Cs = 47.6 pF

The scope showed 1v as the output from the Extech LCR200 meter.

For what it's worth. . .

No technical background leaves me with no idea which is the more appropriate setting to use of the 2 other than which ever one is chosen, use it for all measurements for relative comparability of my own winds.

[Joe Gwinn]

I gave it a try. I set the LCR to 1 kHz with the follow results.

Set probe to 1X:
Rp = .9996 MΩ (~ matches probe 1X setting)
Cp = 180.8 pF
Cs = 320.2 pF

Set probe to 10X:
Rp = 9.751 MΩ (~ matches probe 10X setting)
Cp = 41.2 pF
Cs = 47.6 pF

Scope inputs are rated as a resistance and parallel capacitance to ground, so only Cp is correct, unless I greatly misunderstand the setup.

The specs for my ten year old Tek TDS3012B scope are 1 Mohm with 13 pF in parallel direct, and 10 Mohm with 13.3 pF in parallel with the model P3010 10X probe, so the above input capacitances seem a bit out of line, at least for a 100 MHz scope (the upper end of this scope series is 500 MHz, but same 13 pF.). The Syscomp is intended for

Note that a P3010 probe won't solve the problem if the direct input capacitance is really 320 pF - scope probes are designed to compensate for a specific small scope input capacitance range.

Hmm. That 320 pF seems far too high, even for a 200 KHz scope input (CGM-101). Go to website. Datasheet says 20 pF in parallel with 1 Mohm. The Syscomp CGR-101 is 2 Mhz (versus 200 KHz), but is still 20 pF in parallel with 1 Mohm. Probably uses the same input circuitry.

That 20 pF is quite plausible, and is most likely correct, so the mystery is how one gets to 320 pF. Is there a coaxial cable in there? There should not be (this is not a guitar).

I'm also suspicious of the 48 pF for the 10x probe, on the assumption that Syscomp would not choose such a probe given their direct 20 pF input capacitance.

More datasheet research. The model ACC-1 scope probes are selectable between 1x and 10x. In 1x mode, the input capacitance is 85 to 135 pF (depends on scope input capacitance) in parallel with 1 Mohm, while in 10x mode, the capacitance is 18.5 to 22.5 pF (again depends on scope input) in parallel with 10 Mohm.

I also did the Extech 380193 and Scope mutual measurement: The 10x P3010 scope probe measures as 20.0 pF with a Q of 1.231 (Rdc is 10 Mohm). Now I believe Tek when they say it's 13 pf, so the Extech is at the limit of its abilities. A LCR meter that can do 10 or 100 KHz will likely do better.

The reactance of 13 pF at 1000 Hz is 12.24 Mohms.

The 320 pF is still a mystery. Time for a circuit diagram of the setup, with all components shown, including cables.