0

first let me define the application and circuit. the application of this is to sence currents ranging from 1nA to around 900mA. to accomplish this two resistors are used, a 10K for the nA range, and a 0.01R for the mA and uA ranges. the differential signals from each are then fed into a (and here is what my problem is) analog switch, mux, relay or some other solution. the single differential output is then fed into a x10 op-amp circuit, a programmable gain amplifyer, and finally a pair of ltc2410 24-bit ADC. the solution would need to switch states fast, but would only do so very rarely, as nA range is used rarely in this case.

if the solution didnt require a good amount of accuracy, a good quality analog switch or mux with low on state resistance and good channel-channel matching would be good. however i have heard that these can have problems, such as added voltages (charge injection) (my voltage + added voltage = bad, what i would want is my voltage + ~0 = good) or distortions and leakage currents.

here are two i had in mind (#1 is an unlikely final choice):

(1) https://www.digikey.com/product-detail/en/maxim-integrated/MAX4685EUB/MAX4685EUB-ND/948014 (1nA leakage)

(2) https://www.digikey.com/product-detail/en/analog-devices-inc/ADG888YRUZ/ADG888YRUZ-ND/997607 (200pA leakage)

another solution i have heard used in high accuracy low distortion applications such as this is the use of reed relays. looking on digikey i only found "one" relay i think would work here. board space for me is limited so only smd components can be used. it is a spdt relay so 2 would need to be used.

-https://www.digikey.com/product-detail/en/coto-technology/8L61-05-001/306-1271-ND/1914965

my ultimate question here is which is better, a silicon solution (mux etc) or an electromechanical one (relay)? i have found information about the application of the silicon solutions, but little about using relays. what would be the ultimate advantages and disadvanteges to each in this situation and is there a clear victor? (also am i overlooking any devices?)

super7800
  • 21
  • 4
  • Your range is 9 orders of magnitude!!! Are you serious? I've done as much (almost exactly, in fact) but the peak was about \$10\:\mu\text{A}\$ and not \$900\:\text{mA}\$. It was "hard." One of the earlier solutions was, in fact, to use a COTO relay (electromechanical.) But these are big and expensive. Later renditions used some Burr Brown parts, but I was involved in securing dice rather than packaged (epoxy) parts and I had to do wire-bonding as well as a TE-cooled 3-tier stack. I don't think you'll get a "cheap and quick" answer here. – jonk Apr 27 '20 at 04:20
  • Relays would be slower than your silicon parts. So if you looking for faster response, Silicon is way to go. – Leoman12 Apr 27 '20 at 04:31
  • the only reason for "speed" (the relay has 1ms operate/release time) is not to damage the amplifyers as theoretically going from 100nA to 100mA unpredictably could fry something as the 1ms is on top of the time it takes for the micro to realize it needs to switch. lets say 1.5ms from current ramping up to the circuit responding. i need to consult my component datasheets to see if they could withstand that. https://electronics.stackexchange.com/questions/58021/can-you-feed-in-a-voltage-higher-than-your-supply-to-the-input-of-an-opamp. – super7800 Apr 27 '20 at 04:43
  • also jonk thanks for the input. its meant to be general purpose, but more specifically for development of very low power microcontroller systems (where sleep might be nA but wake might be mA). the accuracy is thus "overkill" but like i said general purpose, and i have projects where that might come in handy. – super7800 Apr 27 '20 at 04:51
  • Can you show us your proposed front end circuit? – Bruce Abbott Apr 27 '20 at 05:13
  • yes of course. it is still a work in progress and is quite messy and hard to understand im sure (hence why i didnt post it). the adc circuit in particular needs alot more attention. i ommitted all the digital stuff. also VON is not connected, nor are their filters setup right on VON or VOP. https://i.postimg.cc/yYrL1KKV/Inked-Schematic-low-burden-current-sensor-2020-04-27-00-35-11-LI.jpg – super7800 Apr 27 '20 at 05:43

0 Answers0