Voltage measuement

28 Feb 2011

Hello Everyone,

Can anyone recommend a voltage measurement interface for the Mbed I want to measure DC voltage between 0 and 100V. I am currently using a large potential divider and the Mbed on board A to D inputs. This is obviously power hungry and I am looking for an alternative. I have found some fairly simple chips to do the job but just wondered if anyone could point me in the right direction possibly recommend one they have used.

Thanks in anticipation John

28 Feb 2011

What do you mean by 'power hungry'?

28 Feb 2011

There are alternatives:

  • External SPI-compatible ADC, like the MCP3208
  • Add a single outboard comparator chip, drive one input from the 100 volt signal (divided down with a high-impedance divider, of course); drive the other input from the mbed's DAC output; then implement your own successive approximation algorithm in software
  • Use an external VFC chip like the LM331, then count the input frequency as a proxy for voltage.

But I would think that your current approach is simplest.

28 Feb 2011

Hello Hexley, Thanks for the info, really helpful Plenty to work with, had a quick look at the MCP3208 which looks like it will do the job. Thanks for your help, Kind regards John.

28 Feb 2011

100V across say 1M dissipates 100^2/10^6 = 10mW. This seems managable, otherwise use 10M and get 1mW.

28 Feb 2011

Ad van der Weiden wrote:

100V across say 1M dissipates 100^2/10^6 = 10mW. This seems managable, otherwise use 10M and get 1mW.

Not to split hairs, but it is worth noting that the LPC1768 data sheet specifies a maximum source resistance, Rvsi, of 7.5K. If that is the bottom leg of a 100 volt-to-3.3 volt divider, the top leg would be 220K. So a 10M divider, or even a 1M divider isn't really in the cards.

Of course even with a 227K divider the dissipation is still less than 45 mW.

I agree that such power dissipation seems minuscule. But maybe John has a flea-power source -- like a voltage n-tupler with many stages, for example.

01 Mar 2011

Hello All,

The voltage I want to measure is derived from a Photovoltaic array. The voltage measurement is required for a maximum power point tracking design. A large resistive load (PD) across the panels would be undesirable.

Thanks for all your help

John

01 Mar 2011

Check out this. It's about measuring battery voltage but probably won't be hard to adapt to your case.

01 Mar 2011

When a few hundred kiloOhms is too heavy a load you can try a compensation approach. I.e. you generate a voltage that you can measure equal to the voltage you want to measure. Equal means there is no current flowing in either direction. Alternatively you can use a source follower circuit. The gate is connected to the voltage you want to measure, the source voltage tracks this voltage (minus the threshold voltage). On the source you can use a resistive divider. Both approaches need a high supply voltage (above your 100V).

01 Mar 2011

Thanks Igor, A variation on the Controllable Voltage Divider is the method I am currently using. I am charging a cap and taking a snap shot of the voltage through a PD. Just thought there maybe a better method.

Many thanks John