Is there any public data, or community insight on best practices to prevent leakage from low-current traces?
For pcb features that involve exposed copper such as component pads and vias, the area can be surrounded with a driven guard. This technique reduces pcb surface conduction especially from deposited oil and contaminates and under varying environmental conditions/humidity, by keeping voltage potentials the same.
But there are also choices about how to route traces. For example a low current trace could be -
- sandwiched between fr4 laminates, versus
- routed on the pcb surface and covered by solder mask
These design choices affect board complexity and cost. So it would be good to know what the relative benefit is/ or if there is any benefit.
Have a look at the “Dielectric Relaxation” section of the ADA4530-1 datasheet. Actually, the whole Applications Information section is superb.
sandwiched between fr4 laminates - increase capacitance and leakage.
covered by solder mask - increase capacitance, leakage and electrochemical effects.
Are there any papers on this subject?
For example, the capacity relative to what increases in the presence of a mask and what does it threaten?
I can also recommend to have a look at the beautiful and fine book 'Low Level Measurement Handbook' 7'th edition. Most of the question are answered there.
I like that handbook, but most preferred for me this book:
A.M.Ilukovich "Technique of electrometry"
Publisher: Energia 1976y.
Link
This is great artefact a golden age of electrometry.
I'm sure it's a great read, but I find the language a little problematic.
If I were a lot younger, I would have studied the russian language.
I don't remember off the top of my head, but I suggest you check the datasheets' of opamps with the lowest bias current, usually they probably have great suggestions.
I like that handbook, but most preferred for me this book:
A.M.Ilukovich "Technique of electrometry"
Publisher: Energia 1976y.
Link
This is great artefact a golden age of electrometry.
I'm sure it's a great read, but I find the language a little problematic. If I were a lot younger, I would have studied the russian language.
Oh, yes! Russian is a little difficult. But some time ago I traslated very useful paper from Chinese into English (not Russian!) using google translate. The information was very valuable to me. So, it sound like there is nothing impossible))
But some time ago I traslated very useful paper from Chinese into English (not Russian!) using google translate. The information was very valuable to me. So, it sound like there is nothing impossible))
Yup!
In fact, translation is not a big problem, it just need a OCR+Dictionary software, as example there is a first 11 pages of this book.
https://ampnuts.com/wp-content/uploads/2022/06/technika_electrometrii2x_translation.pdf
I'm translate it without carefully, but looks are not too bad.
Unfortunately, I do not see any interest EEVBLOG to this book, so I do not consider it necessary to translate it. This is probably due to the fact that this is a Russian book... And now all Russians toys are disliked much more than last year.... 
Moreover, this book is very difficult to translate, because OCR is fail on formulas.
What software have you used?? It has recognized all the photos and formulas!!
In my previous vacations I started translating the book. I couldn't find any free or reasonably priced OCR program that accepted big (one chapter) documents, so I created a small Python script that used OCR to convert an entire page into a string and then it fed it into Google Translate. Problem is, there was no way for it to recognize photos or phormulas, so I had to add them by hand. I got to the second chapter with the help of user forum bsw_m but it was extremely slow having to add all the photos and formulas. Then classes started and that was that...
I'm of course very interested in this book and a couple of others bsw_m showed me. We even asked for help in a translation agency but neither of us could really believe how much the charged for ONE single book.
Dear Atomillo, at first, small note must be clarified: You ask free software from user who live into "The Pirate Bay" country.
I use ABBY FineReader 15 OCR + PROMT Master dictionary + hand work.
I suppose its not entirely moral but at moments like this I would very much like how to use pirate web sites... 120 euro for just one year! Holy moly
The dictionary is not as important, Google Translate often works great and just by reading one can isolate issues. But the scans I see are still a problem
Another question; If an (potentially bootstrapped) op is used to buffer a high-impedance sense input - can the same op output be used for the driven guard? For both pcb guard traces and perhaps external cabling? Or should another op amp source be used?
The argument against the using same op is the extra capacitive loading on the output/emi, and fault tolerance if the guard was shorted by an end-user. I notice some older instrument schematics use a separate buffer op followed by a simple AB amplifier to lower impedance output, but it's not clear how necessary this is.
I can suggest that you have a look at ADA4530-1 datasheet. Figure 99 show an internal structure of the op amp. There is pin 'guard' that can drives an external cable shield and (as I suppose) internal guard rings. To isolate guard unity-buffer from an external capacitance there is an internal resistor of 1 kOhm.
Thanks, I saw Fig99 based on your previous recommendation to check out the ADA4530-datasheet. The extra comment about the purpose of R1 (1k) helps.
I guess the only outstanding question in my mind is - if the separate unity buffer is provided as a convenience for the user, so that the main op can be configured for non-inverting gain. And if gain is not needed then perhaps an ordinary buffer could serve both purposes.
Another question; If an (potentially bootstrapped) op is used to buffer a high-impedance sense input - can the same op output be used for the driven guard? For both pcb guard traces and perhaps external cabling? Or should another op amp source be used?
The argument against the using same op is the extra capacitive loading on the output/emi, and fault tolerance if the guard was shorted by an end-user. I notice some older instrument schematics use a separate buffer op followed by a simple AB amplifier to lower impedance output, but it's not clear how necessary this is.
This is a piece of the bench DMM circuit, which has an input impedance >1TOm. And has an terminal on the front panel of the guard potential device for the use of guarded probes.
Thanks bsw_m, that's extremely helpful.
I'm currently reading carefully chapter 2.
Yes there are some minor mistakes and one case in which the translator made one small mistake and this mistake then conditioned the rest of the phrase for example, but essentially everything is readable. When I finish finals I hope to rewrite a lot as well as do a bit of reformatting to make everything more readable.
In the meantime I'm thinking about the graphics. Yes one could just use Paint to put a white square over the text and then write over it the translation but I'm not sure how well it would fit.
Unfortunately I don't think there is much that can be done in this area, and ultimately in a technical text the main priority should be readability and understanding not presentation.
edited translation of chapter 1. Should be easily readable now.
But can't upload to github repository - see attached file.
chapter 2 will follow...
I'm working on Chp 3. I've skipped the log elements (can someone suggest an example of their use? If you need that dynamic range why not build an I->F?) and I'm formatting the tables separately. I still have to finish it!
Notable I believe is the detailed construction of the air capacitor and the use of guarding.
PS: Title of the document is in spanish but the whole document is english!!
EDIT EDIT: I believe we should get some index first, because for instance the vacuum tube amplifier section or the electrostatic meter section is not very useful today, but the part about dynamic capacitors for example I'm pretty sure contains information that is not easily available anywhere else!
find ch2 attached.
my WYSIWYG is an illusionist - formulas and figures just disappear ... If something is seemingly missing - drop a line.
Will edit ch4 next. This will take much looonger however.
attached is CH4. mostly taxonomic.
will edit CH5 next - but need vacation first.