Timeline for ESD protection for pH measurement electrode input
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 10, 2021 at 9:40 | vote | accept | Medo42 | ||
| Dec 29, 2015 at 16:43 | comment | added | Matt Young | @richieqianle As I said above, this configuration protected to IEC61000-4-2 Level 4 criteria B, take it or leave it. | |
| Dec 29, 2015 at 7:46 | comment | added | richieqianle | Can TVS really "absorb" the esd? I am sure if this is a good design, since TVS increase the clamping voltage at the same time. | |
| Mar 3, 2015 at 14:46 | comment | added | Medo42 | I did not intend to criticise your design. I simply want to understand it before I use it, and unfortunately I don't have an experienced colleague to look over my shoulder and explain things. So if I see something I don't understand, I have to ask directly or remain ignorant. That said, thanks for pointing me to the fact that TVS diodes are meant to absorb, not just conduct away, which is what I had assumed. I'll look into this and open a new question if necessary, since it's really tangential in this one. | |
| Mar 3, 2015 at 14:15 | comment | added | Matt Young | @Medo42 It wasn't a matter of clamping. They just flat out don't absorb any of the energy. That's what the TVS was there for. I had a 2.5kV isolation barrier to maintain, and dumping the entire energy of the ESD hit into the ground plane caused all kinds of problems. I spent a month working on this issue, tried lots of different schemes and devices, and ended up with a product that passed level 4, criteria B. Take it or leave it. | |
| Mar 3, 2015 at 14:05 | comment | added | Medo42 | There must be something I misunderstand here, because as far as I can tell, the SMA diode will only raise the clamping voltage at the electrodes. The ESD current will always have to go through one of the 0603s AND the SMA, so if the 0603s don't clamp the voltage well enough when connected directly to ground, how can they do so with an additional device in the current path? | |
| Mar 3, 2015 at 13:56 | comment | added | Matt Young | @Medo42 For that application, I tried several different TVS diodes, and they had too much leakage current, but absorbed the hit. The 0603ESDAs had virtually none, but didn't absorb enough energy. Putting the two together was the magic bullet that gave me the protection I needed. | |
| Mar 3, 2015 at 11:35 | comment | added | Medo42 | I have just come back to this because I'm looking at protection for some other input now, and now I wonder: Why is the SMA-diode even there? Wouldn't it be enough to directly connect the 0603ESDA diodes to frame ground? Maybe I misunderstand how the 0603ESDA's work... | |
| Oct 15, 2014 at 12:34 | comment | added | Medo42 | Thanks - this might come in handy at some point. However, given the very high impedance of some pH electrodes I mentioned in my question, even 100pA could be significant (1 - 100mV, or up to 1.5pH values). I think by now though that a large resistor in front of the op amp input will be sufficient - 10 Megaohms would limit even a 20kV discharge to 2 mA, which is within the specs for the amp I want to use now - and at 4pA input bias max over temp for the op amp I chose, the error is 40µV max. | |
| Oct 15, 2014 at 1:03 | history | answered | Matt Young | CC BY-SA 3.0 |