Correct probe to avoid detuning

[petert]

Hi,

I'd like to make several measurements as described in the "Elenco AM/FM Radio Kit". In the manual on page 51 (pdf page 52) it is highlighted that:

Your scope probe must have an input capacitance of 12pF or less, otherwise the probe will detune T1 resulting in an incorrect measurement.

Other parts of the manual have the same requirements to avoid detuning.

Now my scope is a Rigol 1054Z (hacked to 100MHz bandwidth) with an input impedance of 1MOhm and 13 pf at each scope input. The probe itself is a PVP2150 with a compensation range of "5 pF - 29 pF" and 10 MOhm impedance (at 10x setting).

Would this work out fine, or do I need another probe? How would I need to adjust the probe compensation to get below 12pF? And if I do so, wouldn't that affect the rendering quality of the 1kHz calibration square wave (the one provided by the scope itself to set the probe compensation properly)? And in that case, wouldn't it also affect the quality of measurements in general?

[petert]

The reason I am asking is that if I need another probe, I'd like to order it with other instruments I will buy shortly.

I mainly wonder if a 100x probe with less capacitance would be less susceptible to detuning the circuit under test.

[Mechatrommer]

the attached pictures are input impedance (more or less) measurement at the DS1052E and DS1054Z tip of the standard bundled probe. if you want to say measure 100 ohm or higher impedance node at few tens MHz freq, then maybe the reading is a little bit off (or detuned and make it disfunctional). maybe active probe is the solution, as i guess most if not all cheap passive Hi-Z probe has similar characteristic.. ymmv.

[RoGeorge]

The reason I am asking is that if I need another probe

I wouldn't buy a 100x probe to measure 40mV/10.7MHz.  100x probes are usually for measuring high voltage signals.  Even if you'll buy one, then you'll need another oscilloscope, too, because the signal fed to the oscilloscope input will become too small.
 

To measure 40mV/10.7MHz stretching over 4 divisions with an x100 probe (as in the assembly manual you linked), you'll need an oscilloscope with 100 uV/division.  Rigol DS1054Z fully unlocked can do at best 500 uV/division, and the internal noise of the oscilloscope will make that range almost unusable unless you use the averaging mode.

As CJay said, you are overthinking it.

[petert]

the attached pictures are input impedance (more or less) measurement at the DS1052E and DS1054Z tip of the standard bundled probe. if you want to say measure 100 ohm or higher impedance node at few tens MHz freq, then maybe the reading is a little bit off (or detuned and make it disfunctional). maybe active probe is the solution, as i guess most if not all cheap passive Hi-Z probe has similar characteristic.. ymmv.

Thanks for the measurement, but I am not entirely sure how to interpret it, nor how you measured exactly (added a 100 Ohm resistor in series?)? Looking up the model it seems to be a network analyzer (testing equipment I am not familiar with ). How would I determine the effective input capacitance from it?

[Mechatrommer]

maybe you can work out from https://en.wikipedia.org/wiki/Electrical_impedance assuming inductance effect is negligible.. 1/10X passive probe can be quite complicated network, esp the coax cable used. it will be difficult to model correctly in SIM to get close to the actual measurement. below is my simplest attempt on my diy probe, the coax cable i used is only modelled as simple 75pF capacitor (measured with DMM), actually there should be evenly distributed of R, L and C over there. the SIM network analysis result is nowhere near the actual result. (KC901V VNA was used on the above result)

[MikeP]

 I think this question has a simple and cheap solution - a homemade active probe. I have one that works well. Input capacitance - 1.18pF. But later I will make a probe with OA, it will be much better. I recommend starting with this and ending with this.

[thinkfat]

Huh? A typical 10x passive probe has much less than 12pF. More in the area of 2pF, if I'm not mistaken. You shouldn't really have a problem with that.

The trimming capacitor in the probe is not to ground but in parallel to the 9M resistor.

[Mechatrommer]

Huh? A typical 10x passive probe has much less than 12pF. More in the area of 2pF, if I'm not mistaken. You shouldn't really have a problem with that.
The trimming capacitor in the probe is not to ground but in parallel to the 9M resistor.

there are two ways of trimming it, one at the head, one at the end connected to dso. now i dont know how much effective capacitance of it, but it is globally accepted by everyone with experience that it will load a circuit of more than few tens and worsen in hundreds of MHz BW.

[ejeffrey]

Huh? A typical 10x passive probe has much less than 12pF. More in the area of 2pF, if I'm not mistaken. You shouldn't really have a problem with that.

The trimming capacitor in the probe is not to ground but in parallel to the 9M resistor.

I'm pretty sure the probe input capacitance is 10-15 pF.  It forms a 10:1 divider with the capacitance of the cable plus scope and compensation caps which total a bit over 100 pF.  The tuning capacitor is to ground but is only a small fraction of the total ground capacitance.

[macboy]

Huh? A typical 10x passive probe has much less than 12pF. More in the area of 2pF, if I'm not mistaken. You shouldn't really have a problem with that.

The trimming capacitor in the probe is not to ground but in parallel to the 9M resistor.

Definitely not 2 pF.

Typical 10x 10 MOhm input passive probes have input capacitance of 8 to 15 pF, depending on brand, quality, bandwidth, etc. Higher bandwidth higher priced ones generally have lower capacitance by necessity.  E.g. Tek P6139A 500 MHz has 8 pF, the P6138 400 MHz probe has 10 pF.  Some high impedance probes go down to around 4 to 5 pF, but often need to compromise by having 20x attenuation in order to get there. E.g. Tek 6563A, 500 MHz, 20x, 9.5 MOhm, 4.7 pF input capacitance.  Tek's best currently-made passive probes have around 4 pF input capacitance and 10x attenuation and are certainly an exception to the rule.  They will set you back around $500 each.

Typical low-Z passive probes have an input capacitance of ~1 pF (0.5 to 2 pF) again depending on BW/price.  Their input impedance (resistive) is typically only 500 to 1000 Ohms, which is the necessary trade-off to get the low capacitance and high bandwidth. E.g. Tek P6158 at 1.5 pF.

To go much below 1 pF, you need to go to an active FET probe, and probably an expensive one at that.

[Mechatrommer]

the 500MHz 4pF 10X Tek TPP0500 impedance past ~60MHz is less than 1K, more than 100MHz, its around 100-300 ohm... (there's currently ~$30 in ebay (new) if you want to fight) but its TekConnect interface. i'm thinking to mod to normal BNC interface or some 3d printed adapter if mine arrived, its mainly for R&D $90 incl shipping down the drain so still.. if you really want to avoid detuning ie Hi-Z past 100MHz, active probe is still the answer imho...

[Mechatrommer]

btw TPP0500B is interchangeable tip (pogo and rigid) does anyone know if the 10MOhm + compensation capacitance reside in the removable tip? (4) see attached, or in the head body connected to the coax cable? where we plug the tip in. has anyone make measurement?

[BravoV]

To go much below 1 pF, you need to go to an active FET probe, and probably an expensive one at that.

Indeed, even after 3 decades, used one still relatively expensive  , example of FET probe with 0.6 pF input capacitance.

[0culus]

Standard 10X probes will work find for any measurement you need to make in that kit. Source: I built one.