DT_Analog & 12F1822 internal voltage reference

[retepsnikrep]

Can I use the DT_Analog routine to read the internal 1.024 voltage reference in the 12F1822 chips?

I'm using the system in my BMS to measure the chip supply voltage and it has worked for years using an external voltage ref LM385 on one of the adc pins and a 12F683.

This is just a natural progression to reduce the component count and increase accuracy.

Any issues look likely?

[Acetronics2]

Hi,

USING the internal reference is more than possible ... but READING it looks a bit strange

Why would you use Darrel's routine for such a job ???

Have a look to my last post http://www.picbasic.co.uk/forum/show...020#post117020 ... it' purpose is to read the supply voltage of the chip !!!

Alain

[Darrel Taylor]

Can I use the DT_Analog routine to read the internal 1.024 voltage reference in the 12F1822 chips?

Of course you can.

This is just a natural progression to reduce the component count and increase accuracy.

Well, ... it will reduce the component count. It doesn't use any pins. And it will increase the resolution.
But it won't necessarily increase the "Accuracy".

When they say a "Fixed 1.024V reference" it sounds really good.
Untill you look at the specs. +6/-8%, and -114 PPM/°C

6% of 1.024V = 0.06144V
8% of 1.024V = 0.08192V

So that 1.024V can be any where between 0.94208 and 1.08544V, when it's at 25°C.
Add in some wide temperature fluctuations, and it doesn't look much like a "Reference" anymore.

The 1822 also has an on-chip temperature sensor, so some of those variations can be removed, but the "accuracy" is still off without calibration.
If calibrated, you can get some amazing results using DT_Analog, but every chip must be calibrated individually.

Disregarding the tolerances, here's a routine that will read VDD on a 12F1822 (without DT_Analog).
I'm assuming it's a battery operated device, and only have the FVR turned on while reading it to save current.

DEFINE ADC_BITS 10
ADvalue  VAR WORD
VDD      VAR WORD


;----[Get VDD by reading the FVR]-----------------------------------------------
VDD_Res  CON 10              ; A/D resolution (bits) used for VDD measurement
FVrefMV  CON 1024            ; Fixed Vref Voltage in mV  ( must match FVRCON )
Vref_AD5 CON EXT             ; Calculate FVref A/D reading when VDD = 5.00V
@Vref_AD5 = (_FVrefMV*100000) / (500000000/(1023 << (_VDD_Res-10)))

GetVDD:
    ADCON1 = %10110000       ; A/D Right Justified, FRC clock
    FVRCON = %10000001       ; Enable Fixed Voltage Reference (1.024V)
    WHILE !FVRCON.6 : WEND   ; Wait for fixed ref to be stable
    ADCIN 31, ADvalue        ; Read the FVR channel
    FVRCON = 0               ; Disable Fixed Voltage Reference
    VDD = Vref_AD5           ; calculated Vref A/D at 5.00V
    VDD = VDD * 500          ;   * 500  - two decimal places
    VDD = DIV32 ADvalue      ;   / FVRef reading - converts to VDD volts
RETURN

Just GOSUB GetVDD and the result is returned in the VDD variable with two decimal places (5.00V = 500).
That is actually good enough for most battery powered applications that need to react to the VDD voltage.
It doesn't use the extra current required for a voltage divider, and doesn't use a pin to read it.
A voltage divider would draw current all the time, even when the PIC is asleep. This won't.

Similarly, here's a routine that does use DT_Analog to get 16-bit readings from the reference.

INCLUDE "DT_Analog.pbp"      ; DT's 16-bit Analog Module
DEFINE ADC_BITS 10
VDD      VAR WORD


;----[Get VDD by reading the FVR]-----------------------------------------------
VDD_Res  CON 16              ; A/D resolution (bits) used for VDD measurement
FVrefMV  CON 1024            ; Fixed Vref Voltage in mV  ( must match FVRCON )
Vref_AD5 CON EXT             ; Calculate FVref A/D reading when VDD = 5.00V
@Vref_AD5 = (_FVrefMV*100000) / (500000000/(1023 << (_VDD_Res-10)))

GetVDD:
    ADCON1 = %10110000       ; A/D Right Justified, FRC clock
    FVRCON = %10000001       ; Enable Fixed Voltage Reference (1.024V)
    WHILE !FVRCON.6 : WEND   ; Wait for fixed ref to be stable
    ADchan = 31              ; FVR channel
    ADbits = VDD_Res         ; set the results resolution
    GOSUB GetADC             ; Get A/D value
    FVRCON = 0               ; Disable Fixed Voltage Reference
    VDD = Vref_AD5           ; calculated Vref A/D at 5.00V
    VDD = VDD * 500          ;   * 500  - two decimal places
    VDD = DIV32 ADvalue      ;   / FVRef reading - converts to VDD volts
RETURN

That version gives much more resolution, but only a small amount of accuracy, and only because the steps are smaller.
However, it gives the possibility of VERY accurate readings, if the FVR voltage is calibrated.

If you can determine the actual FVR voltage, the FVrefMV constant can be changed to match, which makes for incredibly accurate readings. (At room temperature).
For now, lets assume you won't need the calibration. Why ... because it's midnight here.

[retepsnikrep]

Thanks for that excellent reply.

OK lets talk about calibrating it for the extra accuracy. ?

[SteveB]

To make this more portable could we do the following?

DEFINE ADC_BITS 10
ADvalue  VAR WORD
VDD      VAR WORD


;----[Get VDD by reading the FVR]-----------------------------------------------
VDD_Nom	 CON 500	     ; Nominal Value of VDD*100. 500 for 5V, 300 for 3V
VDD_Res  CON 10              ; A/D resolution (bits) used for VDD measurement
FVrefMV  CON 1024            ; Fixed Vref Voltage in mV  ( must match FVRCON )
Vref_AD5 CON EXT             ; Calculate FVref A/D reading when VDD = VDD_Nom
@Vref_AD5 = (_FVrefMV*100000) / (VDD_Nom*1000000/(1023 << (_VDD_Res-10)))

GetVDD:
    ADCON1 = %10110000       ; A/D Right Justified, FRC clock
    FVRCON = %10000001       ; Enable Fixed Voltage Reference (1.024V)
    WHILE !FVRCON.6 : WEND   ; Wait for fixed ref to be stable
    ADCIN 31, ADvalue        ; Read the FVR channel
    FVRCON = 0               ; Disable Fixed Voltage Reference
    VDD = Vref_AD5           ; calculated Vref A/D at VDD_Nom
    VDD = VDD * VDD_Nom      ; * Vdd_nom gives two decimal places
    VDD = DIV32 ADvalue      ;   / FVRef reading - converts to VDD volts
RETURN

[Darrel Taylor]

That's great Steve!
You actually saw how it works.

And as a "Single Point" calibration, it should definitely be calibrated as close to the "nominal" voltage as possible.
However, there's an interesting twist to the calibration procedure, which I'll try to cover first...

OK lets talk about calibrating it for the extra accuracy. ?

OK, let's ...

Remember that Calibration can only be as accurate as the Test Equipment used to calibrate it.
Using a handheld voltmeter with it's Low Battery indicator flashing ... probably isn't the best tool.
I rely on a Tektronix Digital scope for my readings, but it has never been calibrated so I'm sure it's not perfect. (There's really no such thing as "Perfect").

Manual calibration of the FVR is fairly simple. All you need is ...

1. The current VDD voltage during the test.
2. The A/D reading of the FVR at that voltage.
3. The Maximun Possible A/D value. (65472 for 16-bit DT_Analog)

For example, on a 12F1822 here I read ...

VDD = 5.11 V <-- Measured VDD voltage
ADC = 13307 <-- Actual A/D reading from the FVR

So you can use a calculator, spreadsheet or your superior mental abilities to do ...

V = VDD / MAXad * ADC
V = 5.11 / 65472 * 13307
V = 1.039

Adjust the constant to match the measured FVR voltage ...

FVrefMV  CON 1039            ; Fixed Vref Voltage in mV  ( must match FVRCON )

And you have a nice single point calibration that will read exactly the same voltage as the test equipment at the test conditions.

So here's the "twist".
Since the calibration formula includes the measured VDD, the math doesn't care if it's 3V or 5V.
The end result is the voltage of the internal reference.

If the test VDD was 3V, and the A/D reading was 22675 ...

V = VDD / MAXad * ADC
V = 3.00/ 65472 * 22675
V = 1.039

So the FVrefMV CON 1039 is the same either way, and it's the only thing that gets changed in the program.
The math in the program uses 5V as a key datapoint, but the actual calibration voltage doesn't have to be that.

But the truth is ... if you calibrated it at 3.00V, the A/D result would NOT be 22675.
The so called "Fixed" voltage reference, varies with VDD.
So for accuracy across the full range, you need at least a two point calibration.

How do you do a "Two Point" calibration of the Fixed Voltage Reference?
I don't know, I haven't done it yet.

Get me started ... I dare ya

.

[retepsnikrep]

Thanks for even more great ideas.

Maths is not my strong suit but i am getting a handle on how it works.
I worked for years using an external reference and a 12F683 calculating it's own VDD for my BMS.

What is confusing me is why we have an EXT variable and then this huge formula in assembly. Why?
Preumably because it can't fit into 16 bit integer maths even with DIV32
In assembler on the 12F1822 do we have 32 bit maths?

Vref_AD5 CON EXT             ; Calculate FVref A/D reading when VDD = VDD_Nom
@Vref_AD5 = (_FVrefMV*100000) / (VDD_Nom*1000000/(1023 << (_VDD_Res-10)))

I appreciate the accuracy will vary over the VDD range i'm more interested in how it changes with temperature for now.

I suppose I could calculate the vref for each chip at various temps from 0-40C at 5C intervals then use a lookup table to pick the right one to use based on the chips own internal temp reading taken just prior to reading the VDD.

All good stuff. I'm going to be doing some bench testing very shortly to see how it works in practise.

I'm very disapointed you won't let me use my $5 multimeter with dying battery to calibrate my whole system

If you had a very accurate calibrated computer controlled psu you could do calibration at any voltage you wanted. But as you say the VDD is already in the maths so why not just calibrate at a nominal 5v and at 3V using a lithium button cell or something as the supply. Then work out the relationship or slope changes between them both.

In practical terms my program might need to do a few things.

1) Have a 3.3v calculated constant vref in eprom. (I'll probably use 3.3v based constant as that's where the cells spend most of the time)
2) Get it's own temp prior to measuring it's own VDD and compensate according to prev temp variation data points.
3) Measure it's own VDD using the 3.3V constant. V = 1.039 or whatever
4) Use the measured VDD to load a new constant value if reqd due to change in VDD. Say cell is at 3.65V in this example
5) Finally use the new constant to read it's own vref again to have a final value VDD

I will be happy with ACTUAL in use accuracy to within +/- 25mv as that's about what I manage at present.

[SteveB]

What is confusing me is why we have an EXT variable and then this huge formula in assembly. Why?
Preumably because it can't fit into 16 bit integer maths even with DIV32
In assembler on the 12F1822 do we have 32 bit maths?

Vref_AD5 CON EXT             ; Calculate FVref A/D reading when VDD = VDD_Nom
@Vref_AD5 = (_FVrefMV*100000) / (VDD_Nom*1000000/(1023 << (_VDD_Res-10)))

That formula is composed entirely of constants, and is assigned to a constant. So, it is evaluated at compile time (by your 64 bit computer) and the integer value is assigned to the the constant "Vref_AD5". For 5v it ends up being 209.

[SteveB]

You actually saw how it works.

I'll confess, it was a head scratcher. After a few reads, an excel spreadsheet to see some numbers, and digging into the data sheet to confirm FVR could be read by the ADC while the ADC Vref was set to 'something' else, the light bulb came on.