# 16-bit ADC supporting 0-10 volts for Arduino info needed

Hello, I'm looking for a 16-bit (min) ADC that can support a 0-10 Volt signal from a pressure sensor that plays well with an Arduino Due. Something that has a tutorial that goes along with it to process the signal would be great. So far, I'm unable to find something straightforward, so either I'm not searching correctly or I'm not understanding the results enough to know that's what I'm looking for.

I do need the resolution that a 16 bit ADC will provide. The pressure transmitter I'm using will also send a 4-20 mA as signal output, as well as 0-10v.

This has been a stumbling block for me for a while on this project and any help would be appreciated!

Based on a parametric search at DigiKey.com it looks like a 10V A/D is a rare thing. The only ones I found with an analog voltage input above 9V were 11.5V and up. Have you thought of using a voltage divider so you could use a 5V A/D?

Analog Devices, Texas Instruments, and Maxim Integrated all have SPI interface 16-bit 5V A/D converters in 16 or 20-pin DIP packages.

Yes, I’d use a 5V ADC and just use a voltage divider to turn 0-10V into 0-5V. Keeping in mind that inaccuracies in the resistors of the voltage divider will affect accuracy of the output.

I can't use a 5v ADC...wish I could, then I'd just use the on board one with arduino. I need the resolution I get over 10 volts with 16 bits.

I need the resolution I get over 10 volts with 16 bits.

Then you have to use a 5V one with an amplifier.

Just because you requirements are for one thing does not mean that someone has to supply a component that will specifically do that thing. Part of engineering is making what you have ( or what you can get ) do what what you want.

16 bits over 10V. 0 to 65535. 152.6uV per bit.

16 bits over 10V/2. 0 to 65535. 76.3uV per bit. But that is /2, so if you multiply by 2, that is still 152.6uV per bit of the original signal.

You aren't losing resolution.

So, you're suggesting that a voltage divider over the 0-10v output of my sensor into a 16 Bit ADC would, at worse give me, at worse, 76.3uV per bit. I don't quite understand how then multiplying that number by 2 doesn't water down the result, but I think your saying that it waters it down so little, it probably doesn't matter. Correct?

I’m not explaining myself well.

1 bit represents 152.6uV of the original 0-10V. It doesn’t matter if that is 16 bits over 10V, or if you divide by 2 and feed it to a 16 bit ADC with 0-5V range. 1 bit still represents 152.6uV of the original 0-10V signal.

Honestly, that’s the first explanation of that that makes sense. There are plenty 16 bit 0-5v ADC’s out there…will give that a shot.

I'm looking for a 16-bit (min) ADC that can support a 0-10 Volt signal from a pressure sensor that plays well with an Arduino Due. I do need the resolution that a 16 bit ADC will provide.

The pressure transmitter I'm using will also send a 4-20 mA as signal output, as well as 0-10v.

Just curious why the resolution needs to be so high when the transducer's accuracy would be many times worse?

EXAMPLES

A very accurate pressure transducer would be 0.1% FSS (full scale span). Your requirement is that the resolution be 0.0015% FS (16-bit) In this case, the accuracy is 67 times worse than the resolution

If 12-bit resolution would suffice, then the 0-10V signal could scaled with a voltage divider or op-amp to take full advantage of the 12-bit range of the Due's ADC. In this case, the resolution would be 0.0244% FS (12-bit). The transducer's accuracy would be 4 times worse than the resolution.

Good points, dlloyd.

I can see advantage of sampling 0-10V signal is that external noise night have a smaller affect on very low level signals.

Analog devices has several 10V P-P and 20V P-P 16-bit ADCs http://www.analog.com/parametricsearch/en/10169?mtuid=MTS5KSkLdcfF-PC#10025/p7=16

You can use LTC1609 or ADS8517. Good luck.

CrossRoads: I can see advantage of sampling 0-10V signal is that external noise night have a smaller affect on very low level signals.

Noise from resistors is easy to calculate and check. I think you mean interference rather than noise in the strict sense though. The key to both is using a low impedance, which may mean buffering the signal and then using a divider with fairly low value resistors. Then you have to worry about the noise and distortion of the buffer...