Differential output

Hi Guntoju,

I don't understand if you actually made a question. Could you elucidate me (and probably other forum users?)

Agree w/Luis - post could have been more carefully composed.

Differential ADC mode demands that you confine the "differential" input voltage excursions to 1/2 the ADC's maximum input voltage - on both the "positive & negative" input voltage swings.    (i.e.± 1.65V for these MCUs when VRef is @/around 3V3.)

Now when you input a differential voltage of 0V0 (that's 0.0V) the MCU will output 1/2 its full scale.   ADC readings exceeding that mid-scale result when the differential voltage drives positive - readings below mid-scale result when the differential voltage drives negative.

Your post "weaves in" a multiplier factor which confuses rather than illuminates.   I'd avoid any multipliers - insure that you do NOT exceed the ADC's max rating and that indeed - the ADC differential readings behave as outlined, herein.   Only - and if/when - that success is achieved - would I add the complexity of, "scaling."

KISS dictates one small step at a time - too much - too soon - over-challenges many!    (especially this reporter)

Hello Luis as per the specifications given in datasheet i configured ADC in differential mode.
I gave (vin = vin+ - vin-)(vin+=0v,vin- = 1.5v) now vin<0 so the result range will be (0x000 - 0x800). When i see the reading in putty or the variable value it is 1100. My doubt is how to check the reading i got is same as i given at input.(If i configure ADC in single ended and i give same 1.5v i got output as 2048 . To get input what i gave i multiplied with 0.7326mv i.e 2048*0.7326=1500mv)So, similarly in differential how to check the output.

Poster cb1 AND Process "KISS" - both slowly, methodically raise pistols to their head...

Hello Anand,

Did you check the meaning of differential Mode? And can you explain what is differential in the example you have?

Regards
Amit

HI Amit Sir,

The voltage sampled in differential mode is the difference between the odd and even channels:

■ Input Positive Voltage: VIN+ = VIN_EVEN (even channel)
■ Input Negative Voltage: VIN- = VIN_ODD (odd channel)
The input differential voltage is defined as: VIND = VIN+ - VIN-,

Thanks and Regards

Hello Anand,

OK and now what the connections are for Vref+, Vref-, VIN+ and VIN-. What is the expected ADC value and what is that you are reading!!!

Regards
Amit

Hi,

Did you noticed this example:

c:\ti\TivaWare_C_Series-2.1.1.71\examples\peripherals\adc\differential.c

for differential ADC use? Please try it, check it and only if successful go further for scaling (should be the same as for single-ended...). This example is still present also in previous Tiva versions, find out yourself where it is.

Petrei 

Let the record note the 3 "!!!" (note: I've "extra" bullets - and another gun...)

HI Amit Sir 

I used internal refernce (From ADCCTRL register Vref+ = 3.3v,Vref- = 0v) and Vin+ = 2.0v and Vin- = 1.0v.

I got reading as 2666 I multiplied with 1.611(mv Per ADCcode value) = 4294.How to know that ADC converted value is +1v.

When i subtracted 3300mV from 4294mV i got 0.99 is this method correct?

If  i  give  Vin+ = 1.0v,Vin- = 2.0v.Reading I got is 1430.(1430*1.611 = 2300) 2300-3300 =- 1000mV.

For some reason unit conversions always give people problems.

First step is to change the conversion from an unsigned value with an offset to a signed value without an offset.

SignVal = ADCVal - CENTRE_PT¹;

This should give you a signed value with 0 meaning zero volts differential and without the imprecisions introduced by floating point.

For clarity I would then multiply by 2.  This should give you the same scaling as a single ended conversion.

Check this and then we can talk about scaling if you have a question.

This is essentially the same advice as Amit's.  Hopefully a change in wording helps.

Out of curiosity, what are you using the differential mode rather than a differential op-amp configuration?

Robert

1 - This is the centre point of the A/D conversion range as explained in the processor manual.

Hello Anand,

Note that 0x800 is the 0 volt differential code. The code you received is 2666 Dec which is 0xA6A Hex.

Now if the code value received is greater than 0x800 then it is a positive voltage. Thus the sign is resolved as positive. The next thing to resolve it the amplitude. Thus the amplitude calculation becomes (0xA6A-0x800) * 1.611mV = 0.99V.

If it were to be negative then (0x800-Value)* 1.611mV would give the amplitude and the fact that Value is < 0x800 would give the sign as -ve.

Regards
Amit

thank u amit sir i got solved my problem
regards anand

Hello Anand,

That is why the data sheet has a figure to explain the code v/s the differential voltage.

Regards
Amit