1
\$\begingroup\$

I'm using the MCP3421 to read voltages from thermocouples on battery power.

I would like to get this working over USB but I think the low common-mode range of the ADC could be problem so I am thinking about isolating the ADC from the USB micro.

It seems like the ISO1540 i2c isolator could isolate data so I would need an isolated way to power the ADC.

What are my options to supply an isolated few milliamps for the ADC and what impact would these have on the common-mode range?

Ideally I would like the solution to be as small and low cost as possible!

\$\endgroup\$
1
\$\begingroup\$

The easiest way is to use an isolated 5V:5V DC-DC converter to take the USB supply and provide some power at the other side of the isolation barrier. Probably something like 1W will suffice.

There will be some noise and some capacitance across the converter (as well as the capacitance across the I2C isolator). You may wish to add additional capacitance across the converter to conduct noise (read the DC-DC converter data as to that, as well as minimum load, maximum capacitance load and other details).

The common mode range, in theory, would then only be limited by the voltage ratings of the isolators and any capacitors you have added.

\$\endgroup\$
1
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

DC / DC DC/DC converter

You take a small power isolated DC/DC converter - unregulated 5V output , let's say 1W or 2W. Then you have to make a low noise supply by using a LDO and input filtering with use of inductors and capacitors. You then feed the analog part with 3.3V filtered voltage, meanwhile you use raw supply for digital isolators.

EDIT: For the sake of simplicity, both analog and dig. isolator Vcc shall be the same. You split these two supply voltages out from LDO, so you won't have troubles with different voltage levels.

Why 1W, because it is widely available. Keep in mind, that LDO will drop the voltage for approx. 1.7V.

\$\endgroup\$
  • \$\begingroup\$ Why would I need as much as 1W? \$\endgroup\$ – Tim Sep 14 '18 at 19:04
  • \$\begingroup\$ I landed almost exactly there for a similar project. Just a word of caution -- these these want a minimum current draw. You might need to add a resistor in parallel to your circuit. \$\endgroup\$ – Scott Seidman Sep 14 '18 at 19:20
0
\$\begingroup\$

http://ww1.microchip.com/downloads/en/DeviceDoc/22003e.pdf

Have you tried just adding some better filtering on the power supply pin? Inductor in series, and a couple caps in parallel:

3.2 Supply Voltage (VDD, VSS) VDD is the power supply pin for the device. This pin requires an appropriate bypass capacitor of about 0.1 µF (ceramic) to ground. An additional 10 µF capacitor (tantalum) in parallel is also recommended to further attenuate high frequency noise present in some application boards. The supply voltage (VDD) must be maintained in the 2.7V to 5.5V range for specified operation. VSS is the ground pin and the current return path of the device. The user must connect the VSS pin to a ground plane through a low impedance connection. If an analog ground path is available in the application PCB (printed circuit board), it is highly recommended that the VSS pin be tied to the analog ground path or isolated within an analog ground plane of the circuit board.

Common mode range refers to the power supply input, yes? >= 0.3V below Gnd to <= 0.3V above the + supply, which is recommended for 5V, while USB is 5V +/- 5% (4.75V to 5.25V).

Are you trying to power it from a higher voltage?

\$\endgroup\$
  • \$\begingroup\$ It will be plugged into USB where ground will be PC ground. If either of the thermocouple wires gets more than a couple of volts away from that ground, you'll surely have issues? Once you've got multiple devices connected to a PC and sensing on large machinery we've got an off the shelf thermocouple logger that has issues with a 5v common-mode range. \$\endgroup\$ – Tim Sep 14 '18 at 18:37
  • 1
    \$\begingroup\$ @Tim Termocouples are very fancy devices. Some of them have tips connected to shell, some are isolated,...there is always a problem with ground currents, so you better stay with isolated daq. \$\endgroup\$ – Marko Buršič Sep 14 '18 at 21:53
  • \$\begingroup\$ Yes, I think tips connected to the shell is the main culprit! \$\endgroup\$ – Tim Sep 15 '18 at 11:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.