Replacing GOhm resistors with diodes, the theory behind it

[MicUlli]

Hello All,
after finishing some experiments successfully i want to share the results. I replaced a 1GOhm resistor by a double diode and found the noise floor getting down considerably. Find attached a document i created to explain how it works..

 

[Khron]

Works well enough for parts like 2SK596 (with built-in diodes between gate and source)...

 

[molke]

As the resistors are usually mounted ‘hovering’ above the PCB, does SMD (and possibly the very high effective resistance) have practical implications when connecting the diodes?
I've always wondered about similar problems with all the SMD OPA164x solutions replacing a JFET.

 

[user 133392]

You mention GΩ resistors are hard to obtain.
Has anyone measured the noise of these 1GΩ? https://www.mouser.com/c/?q=RGP0207CHK1G0
They are "metal glaze" and in stock.

 

[rogs]

As the resistors are usually mounted ‘hovering’ above the PCB, does SMD (and possibly the very high effective resistance) have practical implications when connecting the diodes?
I've always wondered about similar problems with all the SMD OPA164x solutions replacing a JFET.

Fitting the Hi Z input of an SMD OPA1641 op-amp onto stripboard posed an interesting challenge! ...
I kept the connection between the 1G resistor and the input to the op-amp 'offboard' by not fitting relevant adaptor pin, and mounting the op-amp and resistor as in the attached photo. Seems to work pretty well....

 

[kingkorg]

Hello All,
after finishing some experiments successfully i want to share the results. I replaced a 1GOhm resistor by a double diode and found the noise floor getting down considerably. Find attached a document i created to explain how it works..

Sorry, somehow i missed that you mentioned this in the document.



The problem occurs when these start conducting. I measured output of some LDC with a simple opamp unity gain impedance converter, and these can put out 12V peak to peak! At plosives, snare hit and such. You can get away with small capsules.

Take a look at this video, the part about the diodes starts at 8min. Doug decided to bootstrap them.

 

[ccaudle]

Doug decided to bootstrap them.

Uli's document did point out that you need to bootstrap them.

 

[kingkorg]

Uli's document did point out that you need to bootstrap them.

Dang, sorry, missed that.

 

[molke]

Fitting the Hi Z input of an SMD OPA1641 op-amp onto stripboard posed an interesting challenge! ...
I kept the connection between the 1G resistor and the input to the op-amp 'offboard' by not fitting relevant adaptor pin, and mounting the op-amp and resistor as in the attached photo. Seems to work pretty well....

That does seem to leave at least some connection through the adapter PCB with the plastics below to the main PCB/stripboard, but I guess it’s probably good enough. In the video above, the diode package seems to be mounted completely on the PCB.

 

[MicUlli]

I would always recommend the connection between JFET gate and GOhm resistor (or diode) and capsule wire soldered in free air (no contact on pcb). A datasheet from TI (LMC6001) gives useful hints, see attachment

 

[MicUlli]

You mention GΩ resistors are hard to obtain.
Has anyone measured the noise of these 1GΩ? https://www.mouser.com/c/?q=RGP0207CHK1G0
They are "metal glaze" and in stock.

Yes, i use them too. They perform like it should be, no additional noise except thermal noise present

 

[rogs]

That does seem to leave at least some connection through the adapter PCB with the plastics below to the main PCB/stripboard, but I guess it’s probably good enough. In the video above, the diode package seems to be mounted completely on the PCB.

I think the very proximity of the pins to each other on SMD devices are probably more of a concern than the impedance of the plastic?.....
Might be worth trying to lift the pin as you would on a thru hole op-amp ? ... (as shown on fig.29 of the data sheet linked to by MicUlli in post #10 ).
So far (I only have mics constructed within the last 2 years to refer to) I have had no problems with the construction I showed.
It is important to ensure ALL the flux on the adaptor is removed with isopropol alcohol.
(Cleaned after the 1G resistor and capsule lead are soldered of course)

 

[rankot]

So, what are the best diodes for the purpose?

 

[MicUlli]

So, what are the best diodes for the purpose?

As i mentioned in the paper: BAV199 from Infineon or NXP. BAV199 from other manufacturers have a much higher reverse current. Maybe there are alternatives, not tested yet. Small parallel capacitance is also important, BAV199 shows approx. 4 pF.

 

[bremusound]

Bruel&Kjaer/DPA used the BAV45A in earlier 4006 / 4011 models and their variants.

 

[Gus]

Look for the AT patents

IIRC diode use was posted about in an old thread about the AT4060

 

[Admiral-dk]

Certainly NOT 1N4148 as they have a high Leak Current and are rather Noisy.
Most Schottky Diodes as the BAT series are WAY better (but not all of them) .....
I have no experience with BAT199 - Nor this circuit ....

 

[ccaudle]

Small parallel capacitance is also important

Doesn't the bootstrap make the capacitance irrelevant?

 

[MicUlli]

Doesn't the bootstrap make the capacitance irrelevant?

That's true for the signal. But input noise created by the JFET will be amplified a little bit more. In reality it does not play an important role. Find attached a measurement for comparison...

 

[Admiral-dk]

Wauw - that is some Improvement in the low frequencies - is that with the Circuit in the video ?