Current leak of nA ?

[Festy]

I try to measure an current in nano Ampère but i have a current leak, i try to find the source. Do you have any idea ?
I suppose there are a protect invert diode on the source power and if my current is very low then my current current leak in this diode ? Is it possible ?
If this problem is realy that, I must use dc/dc, regulator ? I take your advice

[kimballa]

A 1nA current leak could come from anywhere. Diodes definitely leak about this much. Check the datasheet.

The 1N4148 datasheet at https://www.vishay.com/docs/81857/1n4148.pdf specifies "reverse current" of up to 25nA at Vr=20V, more at higher voltage and much more at high temperature. Schottky diodes are worse.

At the 1nA level, virtually every component in your circuit (mosfets, ICs) can leak that much. Depending on the board layout, you can see that much current leak between traces, across the soldermask; how well it was cleaned after fluxing will yield different results.

The metrology forum here has folks in it who may be able to give you better guidance, and there are other threads with discussion of builds for 1nV / 1nA scale measurements. But you will probably need to be more specific about the problem you are trying to solve. It is unreasonable to develop a circuit with no leakage anywhere, especially a power supply, where there are so many sources of inefficiency. (The protection diode is one, decoupling capacitors also have leakage. Building your own power supply will also involve passive components like these, which will also have leakage.)

That said, it is possible to create a quiet zone where a specific subcircuit can be monitored with that level of precision...

[Kleinstein]

A leakage current in the nA range is not so uncommon for diodes, but it depends on the diode type. E.g. the fast diodes like 1N4148 are usually more leaky than slower ones. There also low leakage diode with a leakage current more in the low pA range. OP-amps can also have an input bias current in the nA range.

A usefull trick to find a leakage path is local heating of the parts one at a time. In most cases the leakage of semiconductors is high temperature dependent and this often also applies to part that leak because of a defect.  Leakage through dirt / flux on the PCB is less temperaure sensitive - may even go down with temperature as an effect of reduced humidity.

1 nA is not even so super low. Many DMM can use the 10 M input impedance of the mV range to measure currents down to 0.1 nA resolution or so.

[dobsonr741]

OP, pls show a schematics. 1nA is not small at all. DIY Picoampmeter is fairly simple accomplishment: https://www.eevblog.com/forum/projects/picoammeter-design/

[MarkT]

Indeed 1nA is not really small - checkout the ADA4530-1 which has typical input leakage spec that's under 1 femtoamp....  But once you get down to picoamps you need specialized techniques and materials to manage leakage, guard rings, etc.

reverse biased pn junctions generally have very low leakage unless:

elevated temperature
Schottky barrier junction (i.e. not a pn junction),
gold-doped logic diode, exotic RF diode (i.e. not standard doping),
high voltages present.
poor encapsulation, high humidity, etc

High bandgap semiconductors (LEDs are a good example) can be lower leakage than silicon diodes.

[SiliconWizard]

Until we have a full schematic of the device the OP is dealing with, we can't say anything about a few nA leak being "small" or not. No need to try and be smartasses - yes, one can measure pA currents with specialized equipment, but yes, 1 nA is very "small" in practice, unless dealing with very specific designs and applications. But from the - very - succinct descrption of the OP, there is almost no reason to think that a few nA would be abnormal.

As already said above, typical silicon diodes will leak in the order of a couple tens of nA, with only a selected few that can get you lower than this, and not speaking of Schottky, for which it's usually in the µA range and higher.
Tantalum and electrolytic capacitors also leak significantly.
And since we can infer from what the OP says that the "leaky" part is around the power supplies of the device, then, uh. yeah.

[Terry Bites]

I'm trying to read your mind and I'm guessing you have some kind of transimpedance amplifier arrangement.
Show your schematic please.

Input protection diodes: You need very low leakage types, jfets connected as diodes are a good choice here.
Any standard jfet will have leakage at least 100 times smaller than an 1N4148 type. Probably more.
All the high impedance nodes need to be stood of the pcb or protyping board or located inside a pcb guard ring.
The ring can be either connected to ground/ circuit common or the output of an offset control amplifier.

There are many sources of leakage. eg
Dirty pcbs and components, connectors and switches.
Poor cabling or cable types.
Poor PCB design, lack of guard rings on pcb etc.
Poorly chosen opamp or input diodes.
Lack of electrostatic and electomagnetic shielding etc.etc.........

[nctnico]

Ceramic capacitors also leak in the nA range, so that could be a problem as well.

[jonpaul]

Bonjour cher Monsieur : Bienvenue!
Vous n'avez pas reçu de conseils utiles, en raison d'un manque d'informations.
Veuillez nous envoyer des images, des schémas, des photos, etc.
Sans ces informations, il sera impossible de trouver de bonnes solutions ! .
Amicalement
Jon

____________

Hello dear Sir: welcome
You have not received any useful advice due to a lack of information.
Please send us images, diagrams, schematics,  photos,
Without this information, it will be impossible to find good solutions!
Best regards
Jon