JFE150: Capacitance added to gate by the input protection diodes

Hello Fabio,

I agree with Kai's assessment. We do not have data on the diode capacitance, although assuming 1-2pF is reasonable. Power supply noise will be a big contributor as well. Both of Kai's suggestions I agree with. 

Best Regards, 

Chris Featherstone

Hi Kai, thanks for your reply.

I agree on what you say.

To better understand whats is going on I've made some mesure of capacitance on BAV199 and JFE150 connecting only the protection diodes.

I've biased both with +/- 2V, 6V, 8V, 10V, 12V, masuring the capacitance between the central node (the gate for JFE150) and the reference ground..

It is correct?

On BAV199 I've mesured respectively 1.5-1.6 pF, 1.2pF, 1.0 pF, 0.8-0.9 pF, 0.8-0.9 pF, the capacitance decreases when the voltage increases.

On JFE150, 21.8, 21.9-22.0 pF, 21.9-22.0 pF, 21.9-22.0 pF, 21.9 pF, the capacitance doesn't seem to change with bias voltage.

Describing the effect on the charge amplifier with a detector capacitance of 25 pF, if I connect the JFE150 internal protection diodes, the noise rises from 2.76 mV RMS to 3.61 mV RMS, the risetime for a injected charge of 13.4 pC in 10 ns goes from 200 ns to 230 ns.

Definitively, the protection diodes inside the JFE150 have the effect of a increased detector capacitance: more noise, less bandwidth (higher risetime).

I'll try to put a BAV199 on amplifier input and will inform you about the results.

Fabio

I've re-done the mesures for JFE150 putting a 1G resistor on gate toward V- and a 22 nF isolation capacitor to avoid some bias from the meter.

The values are 22.1 pF, 22.0 pF, 22.0 pF, 21.9 pF, 21.3 pF, so it is possible to see some decreasing as the bias voltage increases.

Fabio

Charge amplifier with one JFE150, working +/- 10V, Rf=340M, charge inject of 13.4 fC in 10 ns.

It has 0.110 fC ENC (689 e-) with no

No protection diodes on input:

    detector capacitance 25 pF, noise=3.22 mV RMS, risetime=94 ns

    detector capacitance 54 pF, noise=3.85 mV RMS, risetime 142 ns

BAV199 between +/- 10V:

    detector capacitance 25 pF, noise=3.33 mV RMS, risetime=99 ns

    detector capacitance 54 pF, noise=4.06 mV RMS, risetime=151 ns

JFE150 protection diodes in place of BAV199, between +/- 10V:

    the amplifier cannot reach the operational point, the loop tries to reduce the gate voltage but reaching the minimum voltage, maybe the protection diodes current leakage on gate is competitive with Rf.

Tomorrow I'll use the +/- 6V amplifier that seems do not have this problem despite it has Rf=3G but the protection diodes are connected to +/- 6V.

---

Hi Fabio,

I haven't had enough time to go through all your posts but I only want to say that due to the fact that the protection diodes and the JFET are sitting on the same die you will probably not be able to measure the capacitance of the protection diodes separately from the JFET's capacitances.

Kai

Yes, this is true, what I can see are the effects on noise and bandwidth (risetime).

Here the mesurements of today morning.

Without protection diodes: VoutDC = -0.646V (through a 3G resistor it biases the gate)

     Detector cap = 25 pF, Noise RMS = 2.71 mV, Risetime = 202 ns

     Detector cap = 54 pF, Noise RMS = 3.46 mV, Risetime = 268 ns

With BA199: VoutDC = -0.655 V

     Detector cap = 25 pF, Noise RMS = 2.81 mV, Risetime = 206 ns

     Detector cap = 54 pF, Noise RMS = 3.57 mV, Risetime = 273 ns

With internal JFE150 diodes: VoutDC = -0.653 V

     Detector cap = 25 pF, Noise RMS = 3.65 mV, Risetime = 236 ns

     Detector cap = 54 pF, Noise RMS = 4.19 mV, Risetime = 304 ns

What we can see: using the internal diodes I see an increment in noise and risetime as the detector cap (on amplifier input) is higher, more than 40 pF added.

This is the effect, so in my application I cannot use the internal diodes.

Anyway, it could be interesting to understand what are the protection diodes characteristics.

F.

Hi Fabio,

thanks for the elaborate measurements. Your results are very interesting

The JFE150 contains six protection diodes in total. Have you replaced all the internal protection diodes of JFE150 by external BAV199 or only the input protection diodes?

Also, is the potential at the pin "S" of JFE150 sufficiently higher than the potantial at the pin "VCL"?

Kai

Hi Kai.

The external BAV199 is placed on the input only.

VCH = +6V, VCL = -6V, S = 0V (GND)

F.