DDC112: DDC112

Hello Matthiew,

In your case the current to measure is pretty high for this device, so the ideal settings would be the shortest integration time with the highest range.

With the maximal settings we would have an integration time of 333us and a range of 1000pc.

In that case the maximum integrated current would be of:

Imax=1000pc/333us=3uA

Since the current you want to measure is higher than that by a factor of 20, it’s necessary to adapt the circuitry for that.

You can somehow decrease the current at the integrator by using two resistors in that configuration: if both resistors are equal, the current flowing to the integrator would be divided by two…

However in that configuration it would be also recommended to compensate for the error at V-

Indeed here the current equation could be roughly:  

I= Vin/R1 + (Vin - Error)/R2 where r2 is on the integrator side.

The smaller the integrated  current the smaller the error: so you can choose a bigger resistor on the integrator side.

I hope this helps,

Best regards,

Daniel

Edit: updated the equation

Thank you for the input.  I want to clarify a couple things:

1. What voltage are you referring to when you say V-?

2. What is the PACK in this diagram?  Id this referring to a resistor network to reduce the input current to the integrator?

3. Moreover, you mention that lengthening the integration time will induce more error, however it will lower the input current.  How can I calculate the error at certain integration times?  Would I be able to lengthen the integration time along with increase the input impedance to achieve accurate measurements?

Thank you for all you support,

Matt Anderson

Hardware Engineer

AJA Video Systems

Hello Matthew,
1. The voltage i'm refering to is Vin
2. The pack corresponds to the battery pack: is the way you would like to implement it? If no can you indicate how you would like to use the device with the battery pack and the load?
3. You can definitely do that: indeed the lower the current at the integrator side, the better, and then you can also increase the integration time! but you need to keep the Resistor's value high enough in order to not saturate. And this should be pretty accurate, if using accurate resistors... Let me try to calculate the error we would have for different integration times!
Best regards,
Daniel

I can easily use high impedance resistors...........I assume we are talking in the order of 10M or higher? As far as the battery pack, my thought was to tap the input to the integrator from the V+ output of the batt to the load, it doesn't really matter, however we are going to have to leave these batteries in place on our board, and therefore will be tapping into the circuit. What equations are you using for calculating your error? It's been a while since I've had to analyze OP amps and am a bit rusty. That being said our circuit is two 2450 batteries(3.6V nominal) in parallel, each going through a diode and then onto the circuit. I would like to tap into them where they meet the diode and after as well. A brief depiction of our circuit is as follows:

Matthew,
Actually you should use the device on the low side.
The resistor value doesn't need to be that high in this case, however the Ratio R2/R has to be high, so only a small fraction of current will flow through R2, and it will also reduce the error. Let me have a look on how to obtain the error.
Best regards,
Daniel