ADC-16, can range exceed +/-2.5 volts when using 12 bits?

Post any questions about our serial port products
Post Reply
dogleg
Newbie
Posts: 0
Joined: Wed Oct 27, 2010 9:01 am

ADC-16, can range exceed +/-2.5 volts when using 12 bits?

Post by dogleg »

I’m monitoring the voltage output from various sensors, where the voltage occasionally exceeds the +/-2.5 Volts range specified in the ADC-16 User’s Guide. Now I found that when configured for 16 bit sampling, that as expected, the ADC-16 would max out at roughly +/-2.49 Volts. However I also found that when I configured the ADC-16 to take samples, using resolutions less than 16 bits, that the ADC-16 would continue to digitize samples well beyond +/-2.5 Volts :idea:

For example I have chosen to use 12 bit resolution, and so when I apply +2.5 Volts to channel #1, I get an integer from the ADC-16 which is equal to 0x0FFF in hexadecimal, however as I gradually increase the voltage applied to channel #1 beyond +2.5 Volts, I get quite linear results up to +4.5 Volts.

Can I use the ADC-16 to reliably monitor voltages that occasionally exceed +/-2.5 Volts but fall within a range of say +/-4.5 Volts without the risk of damaging the ADC-16 :?:

The User’s Guide does indicate that damage would occur if voltages greater than +/-30 Volts were applied :!:

ziko
Advanced User
Advanced User
Posts: 1705
Joined: Fri Dec 01, 2006 10:03 am
Location: St Neots

Re: ADC-16, can range exceed +/-2.5 volts when using 12 bits

Post by ziko »

Hi,

The +/- 2.5V is for the the 16 bits resolution at lower resolutions you can actually get higher voltages.

There is a limit though which is not an exact science, but in your particular case it should be perfectly fine.

Hope this helps.
Ziko

Technical Specialist

ziko
Advanced User
Advanced User
Posts: 1705
Joined: Fri Dec 01, 2006 10:03 am
Location: St Neots

Re: ADC-16, can range exceed +/-2.5 volts when using 12 bits

Post by ziko »

Correction depending on how high the voltage goes it may be ok. When I measured it at voltages above 4 V the accuracy started to reduce significantly. Try it with a multimeter.
Ziko

Technical Specialist

dogleg
Newbie
Posts: 0
Joined: Wed Oct 27, 2010 9:01 am

Re: ADC-16, can range exceed +/-2.5 volts when using 12 bits

Post by dogleg »

Thanks for the reply ziko :) Yes I did some testing with a multimeter, and the samples seemed reasonably linear up to around +/- 4.5V, however I was only applying the voltage to channel #1 whilst monitoring all channels.

I also found that as the applied voltage approached +/- 5.0V, that the linearity started to reduce significantly on channel #1, and that the other channels would then also generate wild voltage swings of several volts, even though those channels were not connected to anything and left floating :?

Do you think that perhaps this oddity is due to the fact that I had left these channels floating at the time :?: Can I trust the samples from the other channels if they were connected to actual voltages, and if the voltage applied to channel #1 were to accidentally exceed +/- 5.0V :?:

On a different note, I had some very pleasing results using the ADC-16 driver interface :D I particularly found using the call 'adc16_get_cycle' to be of good value :o

ziko
Advanced User
Advanced User
Posts: 1705
Joined: Fri Dec 01, 2006 10:03 am
Location: St Neots

Re: ADC-16, can range exceed +/-2.5 volts when using 12 bits

Post by ziko »

Without doing extensive testing its difficult to say, we cannot really guarantee it to be accurate above the quoted range.

Although since you have reduced the resolution it is possible to get higher voltages due to the fact that it is a dual slope ADC. I suppose though it does depend on your supply voltage to your ADC, so it could depend on what voltage your serial port is outputting. Are you using a normal serial port or a USB to serial converter?

Hope this helps
Ziko

Technical Specialist

Post Reply