EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: philbona on January 13, 2014, 08:58:53 am
-
Hello all,
I would like to use a (or many) MUX 4052 to switch between multiple K TC as I don't want to purchase multiple max 31855 breakout board.
In my case, the switch time between ports is not really an issue.
Is there any concerns to do so ?
What I suspect would be the accuracy of the value between a K TC directly connected to MAX31855 and the same schema with the MUX in between,
( By the way, I am a newb)
many thanks for your answers,
Phil
-
your best hope at a chance would be to mux both sides of each thermocouple to reduce any signal path thermocouples from creating an offset (treat it like a differential pair (keep them very close) and this for the most part should cancel out most temperature differences,
-
thanks for the answer,
apart from this recommendations, I understand that nothing should contrary the use of 4052 ( 2x4 ports),
I will try soon, just need to buy pieces...
Phil
-
What's the input impedance of the maxim ic? I see only a bias current max, at that max current (100 nA) you could be a couple degrees off from voltage drop with the dual mux approach AFAICS.
-
Back when I worked on sounding rocket payloads, we used thermocouples hooked up with AD595 amplifiers. There were a lot of layout specifics that were required (including metal and solder choice), and as far as I remember, you cannot successfully multiplex them due to noise, unintended thermocouple (two dissimilar metals will form a thermocouple by definition), and impedance issues. However you can wire thermocouples in series and get a sum of the temperatures, which can be quite useful. My $0.02
-
thermocouples in parallel give the average, not series ???
the impedance of the mux was 80 ohms maximum on both channels, and the input independence of the AD595 is very high, so no voltage drop issue there ??? and with thermocouples you want to modify both sides so if a dissimilar metals parasitic thermocouple forms it is added to both sides, thus cancelling
-
thermocouples in parallel give the average, not series ???
the impedance of the mux was 80 ohms maximum on both channels, and the input independence of the AD595 is very high, so no voltage drop issue there ??? and with thermocouples you want to modify both sides so if a dissimilar metals parasitic thermocouple forms it is added to both sides, thus cancelling
We didn't multiplex out thermocouples as we were advised by the manufacturer not to. We literally had 12x AD595 on one circuit board (64 analog-channel telemetry encoder). The AD595 outputs were fed to OP177 op-amps to deliver the +/- 10V differential signal output.
Right on parallel, although series can give average if a precision divider is used (or in the case of cold junctions, you know the temperature of one of the junctions independently).
Looking more carefully the MAX chip has up to +/- 6 degree errors depending on setup, so muxing will probably not be an issue, regardless of imput impedance (which one would assume is quite high).