Author Topic: Extremely high insulation resistance GDT  (Read 1753 times)

0 Members and 1 Guest are viewing this topic.

Offline leblanc

  • Contributor
  • Posts: 48
  • Country: ca
Extremely high insulation resistance GDT
« on: January 18, 2018, 03:15:53 pm »
I am designing a transimpedance amplifier for a 0~100pA current measurement. The resistor in the feedback is going to be a 10GOhm resistor.

When dealing with these high value resistors, care must be taken to make sure that there are no other paths such as leakage through devices or on the surface of the PCB. Special core/prepreg is used, solder mask is cleared in the area, special solder paste is used, surface contamination is cleaned, etc.

I need to protect my op-amp input somehow. TVS/zeners have much too large leakage currents, so I've been thinking of using a gas discharge tube. Unfortunately, most of the ones I've been finding have 1~10GOhm insulation resistance. The leakage through these resistors is going to be much too large, my measurements are going to be way off.

Do any of you know of ultra-high insulation resistance gas discharge tubes (~75VDC spark-over) or other transient protection strategies? It has to be large enough to accommodate a guard ring between the pads.

Also, if you guys have any suggestions for my feedback resistors, that would be pretty helpful. Ideally it'd be a 10% surface mount 10GOhm resistor, with a package large enough to accommodate a guard ring between the pads (i.e. 1206 or larger).

Finally, I'll also be needing a capacitor in parallel with the feedback resistor to maintain stability. I'll be needing a 500fF in a package large enough to accommodate a guard ring between the pads.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 7897
  • Country: de
Re: Extremely high insulation resistance GDT
« Reply #1 on: January 18, 2018, 03:49:37 pm »
For the low leakage parts it might be a good idea to use THT parts and teflon isolated stand offs. Surface mount parts are not really suitable, even with guard rings.
A 500 fF cap would be something like 5 mm of parallel (e.g. teflon) isolated wire, of just some air space. There are essential not 75 V sparc gaps - in air the minimum voltage is around 100 V. Neon lamps could be an option, if the wire spacing it large enough. This about the lowest voltage gas discharge part available. At essentially zero voltage, low leakage diodes like BAV199 are pretty good.
 
The following users thanked this post: leblanc

Offline mark03

  • Frequent Contributor
  • **
  • Posts: 681
  • Country: us
Re: Extremely high insulation resistance GDT
« Reply #2 on: January 18, 2018, 04:13:51 pm »
If the voltage across your input protection device is ~ zero, why does it need such high insulation resistance?
 

Offline Karel

  • Super Contributor
  • ***
  • Posts: 1608
  • Country: 00
Re: Extremely high insulation resistance GDT
« Reply #3 on: January 18, 2018, 04:34:47 pm »
"Gas Discharge Tube 75V 2500A (2.5kA) ±20% 2 Pole Through Hole 10GOhm"

https://www.digikey.com/products/en?keywords=GTCA25-750M-R02-ND
 

Offline CopperCone

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: Extremely high insulation resistance GDT
« Reply #4 on: January 18, 2018, 05:12:32 pm »
If you need higher resistance then you need a good relay to isolate the gdt once the experiment is hooked up. It will always be better to do protection this way, just based on physics
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 5935
  • Country: gb
Re: Extremely high insulation resistance GDT
« Reply #5 on: January 18, 2018, 09:15:13 pm »
What's your input device? If it's a low Ib CMOS opamp then a better strategy is to insert a resistor in series with its input pin to allow the input protection diodes to clamp within their safe current limits.
Chris

"Victor Meldrew, the Crimson Avenger!"
 

Offline Damianos

  • Regular Contributor
  • *
  • Posts: 210
  • Country: gr
Re: Extremely high insulation resistance GDT
« Reply #6 on: January 18, 2018, 10:45:42 pm »
From the datasheet of OPA128LM, that I have used:
Quote
If it is possible for the input voltage to exceed the supply voltage, a series protection resistor should be added to limit input current to 0.5mA. The input devices
can withstand overload currents of 0.3mA indefinitely without damage.
An 1MOhm resistor from the source/feedback node to the Op-Amp input will protect it up to 300V.
For sensitive nodes, use something like this: PTFE Insulated Terminals & Pins, forgetting the SMD components and lifting the Op-Amp input pin from the PCB.
I don't know if polystyrene capacitors with such low values exist, but there is the solution of the TEFLON wires, mentioned earlier.
 
The following users thanked this post: leblanc

Offline leblanc

  • Contributor
  • Posts: 48
  • Country: ca
Re: Extremely high insulation resistance GDT
« Reply #7 on: January 19, 2018, 01:42:59 pm »
I think you've all convinced me to stay away from surface mount. I'll go with through hole parts. I might do the PTFE terminals, I haven't decided yet.

Quote
What's your input device?
I'm using an ADA4530. The bias current is very small, +/-20fA. It's a pretty fancy op-amp with a guard ring driver.

I just found a post from a couple of years ago, there is some useful information here:
https://www.eevblog.com/forum/projects/picoammeter-design/

I've also found a reference design by Renesas/Intersil in the CA3420 datasheet. Damianos, they seem to be doing exactly what you suggested (with the series resistor). I think I'll go with this strategy.


As for the capacitor, what do you guys think of FKP3D001002B00MD00? It's a polypropylene capacitor, 100pF, with insulation resistance >500GOhm. I had originally thought of using a polystyrene capacitor, but I didn't like the idea of something that would break down with IPA, especially when the surfaces here need to be kept clean. Then I thought of PTFE (Teflon) capacitors, but I couldn't find any parts that were suitable.

I think I just won't use a gas discharge tube. I'll might use two ultra-low leakage diodes like ON Semi's FJH1100, which have leakage of 3pA (or even Catalogic PAD1 with a leakage of 1pA) between the input and the rails. Maybe the leakage currents will partially cancel out too.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 4880
  • Country: nl
Re: Extremely high insulation resistance GDT
« Reply #8 on: January 19, 2018, 01:58:22 pm »
The opamp has it's own bootstrapped diodes, don't bother. What's wrong with Mylar capacitors?
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 7897
  • Country: de
Re: Extremely high insulation resistance GDT
« Reply #9 on: January 19, 2018, 02:35:44 pm »
Mylar caps have a considerably higher dielectric absorption so they can be a tricky.
Not having the DA (e.g. in the board)  is also a reason for not going to the PCB with the very sensitive part.

The nice thing about polystyrene caps is that they are really low leakage as they are usually polystyrene all the way. PP caps tend to have a case from a different material and this way extra surface leakage.

The ADA4530 is a fancy expensive part. For first tests a LMC6041 or similar could be good enough (though slow).
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 5935
  • Country: gb
Re: Extremely high insulation resistance GDT
« Reply #10 on: January 19, 2018, 03:46:49 pm »
There's quite a lot of relevant information in this thread:   https://www.eevblog.com/forum/projects/picoammeter-design/
Chris

"Victor Meldrew, the Crimson Avenger!"
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf