Kirk ask for a OpAmp inside a small MSP. The 2013 does not contain a OpAmp.
Just the SD16 with 200kOhm input impedance. To be more specific - he ask for
a high impedance ADC.
I have used the ADC12 as a ADC with a high impedance input. Solution is to
increase sample time to measure high impedance networks. You need to charge
the 40pF input capacitor (this is inside the MSP430). To get the 1/4096
resolution of the 12 bit ADC the sample time has to be at least 9*R*C. (C40pF, R
your input impedance), for 10Megohm this is 3.6ms sample time.
Same is true for the ADC10. The input impedance of the ADC10 is 27pF.
The resolution achieveable with Comparator A is approx. 10bits.
M.
Bill Knight :
> Sorry I caught this thread late. You might have a
look at the
> MSP430F2003/2013.
>
> Regards
> -Bill Knight
> R O SoftWare
>
> On 1/30/2012 2:25 AM, kirkabailey wrote:
>>
>> Lloyd and Al,
>>
>> Thanks for the comments!
>>
>> I've already contemplated a software adc since my sample rate is
pretty
>> slow, so this might be the nudge that justifies the extra code.
>> However, first I need to go dig up the comparator input leakage specs
>> to see what the effective input impedance would be for this approach -
>> probably not worth the effort if I end up still needing an external
>> buffer amp...
>>
>> Has anyone successfully done this at the 14-16 bit level?
>>
>> Cheers,
>>
>> Kirk Bailey
>> --- In m..., Onestone wrote:
>>> You can of course implement a sigma delta converter yourself using
>>> almost any of the smaller parts, provided they have enough memory for
>>> your other apps. The MSP430G2553 looks like a good target, and comes
>>> in the Lunch pad. It depends of course on the sample rates you need
>>> and how efficiently you write your code, but it is quite feasible.
>>>
>>> Al
>>>
>>> On 29/01/2012 11:22 PM, lslonim2 wrote:
>>>> I just went through a similar exercise. I have a fairly complicated
>>>> motor control product. I need a lot of code space and RAM but not a
>>>> lot of pins. I ended up with an 80 pin part. It just seems to be how
>>>> they set up the product line.
>>>>
>>>> Lloyd
>>>>
>>>> --- In m..., "kirkabailey" wrote:
>>>>> Hi All,
>>>>>
>>>> ...
>>>>> Am I missing something? If not, why is this? Most of the
>>>>> simple sensor type apps could really benefit from a high
>>>>> impedance A/D...
>>>> ...
>>>>
Re: Any low pin count 430's with a sigma-delta adc input buffer amp?
Started by ●January 30, 2012
Reply by ●January 30, 20122012-01-30
Oops, jumped in too fast. Saw 'low pin count' and 'sigma-delta
adc' and
spaced over 'input buffer'.
My bad.
-Bill K
On 1/30/2012 9:34 AM, Matthias Weingart wrote:
> Kirk ask for a OpAmp inside a small MSP. The 2013 does not contain a OpAmp.
> Just the SD16 with 200kOhm input impedance. To be more specific - he ask for
> a high impedance ADC.
>
> I have used the ADC12 as a ADC with a high impedance input. Solution is to
> increase sample time to measure high impedance networks. You need to charge
> the 40pF input capacitor (this is inside the MSP430). To get the 1/4096
> resolution of the 12 bit ADC the sample time has to be at least 9*R*C. (C> 40pF, R your input impedance), for 10Megohm this is 3.6ms sample time.
> Same is true for the ADC10. The input impedance of the ADC10 is 27pF.
>
> The resolution achieveable with Comparator A is approx. 10bits.
>
> M.
>
> Bill Knight:
>
>> Sorry I caught this thread late. You might have a look at the
>> MSP430F2003/2013.
>>
>> Regards
>> -Bill Knight
>> R O SoftWare
>>
>> On 1/30/2012 2:25 AM, kirkabailey wrote:
>>> Lloyd and Al,
>>>
>>> Thanks for the comments!
>>>
>>> I've already contemplated a software adc since my sample rate is pretty
>>> slow, so this might be the nudge that justifies the extra code.
>>> However, first I need to go dig up the comparator input leakage specs
>>> to see what the effective input impedance would be for this approach -
>>> probably not worth the effort if I end up still needing an external
>>> buffer amp...
>>>
>>> Has anyone successfully done this at the 14-16 bit level?
>>>
>>> Cheers,
>>>
>>> Kirk Bailey
>>>
>>>
>>> --- In m..., Onestone wrote:
>>>> You can of course implement a sigma delta converter yourself using
>>>> almost any of the smaller parts, provided they have enough memory for
>>>> your other apps. The MSP430G2553 looks like a good target, and comes
>>>> in the Lunch pad. It depends of course on the sample rates you need
>>>> and how efficiently you write your code, but it is quite feasible.
>>>>
>>>> Al
>>>>
>>>> On 29/01/2012 11:22 PM, lslonim2 wrote:
>>>>> I just went through a similar exercise. I have a fairly complicated
>>>>> motor control product. I need a lot of code space and RAM but not a
>>>>> lot of pins. I ended up with an 80 pin part. It just seems to be how
>>>>> they set up the product line.
>>>>>
>>>>> Lloyd
>>>>>
>>>>> --- In m..., "kirkabailey" wrote:
>>>>>> Hi All,
>>>>>>
>>>>> ...
>>>>>> Am I missing something? If not, why is this? Most of the
>>>>>> simple sensor type apps could really benefit from a high
>>>>>> impedance A/D...
>>>>> ...
>>>>
spaced over 'input buffer'.
My bad.
-Bill K
On 1/30/2012 9:34 AM, Matthias Weingart wrote:
> Kirk ask for a OpAmp inside a small MSP. The 2013 does not contain a OpAmp.
> Just the SD16 with 200kOhm input impedance. To be more specific - he ask for
> a high impedance ADC.
>
> I have used the ADC12 as a ADC with a high impedance input. Solution is to
> increase sample time to measure high impedance networks. You need to charge
> the 40pF input capacitor (this is inside the MSP430). To get the 1/4096
> resolution of the 12 bit ADC the sample time has to be at least 9*R*C. (C> 40pF, R your input impedance), for 10Megohm this is 3.6ms sample time.
> Same is true for the ADC10. The input impedance of the ADC10 is 27pF.
>
> The resolution achieveable with Comparator A is approx. 10bits.
>
> M.
>
> Bill Knight:
>
>> Sorry I caught this thread late. You might have a look at the
>> MSP430F2003/2013.
>>
>> Regards
>> -Bill Knight
>> R O SoftWare
>>
>> On 1/30/2012 2:25 AM, kirkabailey wrote:
>>> Lloyd and Al,
>>>
>>> Thanks for the comments!
>>>
>>> I've already contemplated a software adc since my sample rate is pretty
>>> slow, so this might be the nudge that justifies the extra code.
>>> However, first I need to go dig up the comparator input leakage specs
>>> to see what the effective input impedance would be for this approach -
>>> probably not worth the effort if I end up still needing an external
>>> buffer amp...
>>>
>>> Has anyone successfully done this at the 14-16 bit level?
>>>
>>> Cheers,
>>>
>>> Kirk Bailey
>>>
>>>
>>> --- In m..., Onestone wrote:
>>>> You can of course implement a sigma delta converter yourself using
>>>> almost any of the smaller parts, provided they have enough memory for
>>>> your other apps. The MSP430G2553 looks like a good target, and comes
>>>> in the Lunch pad. It depends of course on the sample rates you need
>>>> and how efficiently you write your code, but it is quite feasible.
>>>>
>>>> Al
>>>>
>>>> On 29/01/2012 11:22 PM, lslonim2 wrote:
>>>>> I just went through a similar exercise. I have a fairly complicated
>>>>> motor control product. I need a lot of code space and RAM but not a
>>>>> lot of pins. I ended up with an 80 pin part. It just seems to be how
>>>>> they set up the product line.
>>>>>
>>>>> Lloyd
>>>>>
>>>>> --- In m..., "kirkabailey" wrote:
>>>>>> Hi All,
>>>>>>
>>>>> ...
>>>>>> Am I missing something? If not, why is this? Most of the
>>>>>> simple sensor type apps could really benefit from a high
>>>>>> impedance A/D...
>>>>> ...
>>>>
Reply by ●January 31, 20122012-01-31
Matthias,
Going with the ADC10/12, and adjusting the sample timing, sounds like a good way to go.
Also, thanks for the info on the achievable comparator A/D results. I'm assuming when you said "resolution achievable is approx. 10 bits", you mean the maximum resolution before the accuracy of any additional bits becomes problematic? I have previously reviewed TI Application Report SLAA104, which describes the implementation of a 12 bit Sigma-Delta type converter using comparator_A, but the accuracy of the results weren't really described.
Cheers,
Kirk Bailey
--- In m..., Matthias Weingart wrote:
>
> Kirk ask for a OpAmp inside a small MSP. The 2013 does not contain a OpAmp.
> Just the SD16 with 200kOhm input impedance. To be more specific - he ask for
> a high impedance ADC.
>
> I have used the ADC12 as a ADC with a high impedance input. Solution is to
> increase sample time to measure high impedance networks. You need to charge
> the 40pF input capacitor (this is inside the MSP430). To get the 1/4096
> resolution of the 12 bit ADC the sample time has to be at least 9*R*C. (C> 40pF, R your input impedance), for 10Megohm this is 3.6ms sample time.
> Same is true for the ADC10. The input impedance of the ADC10 is 27pF.
>
> The resolution achieveable with Comparator A is approx. 10bits.
>
> M.
>
Going with the ADC10/12, and adjusting the sample timing, sounds like a good way to go.
Also, thanks for the info on the achievable comparator A/D results. I'm assuming when you said "resolution achievable is approx. 10 bits", you mean the maximum resolution before the accuracy of any additional bits becomes problematic? I have previously reviewed TI Application Report SLAA104, which describes the implementation of a 12 bit Sigma-Delta type converter using comparator_A, but the accuracy of the results weren't really described.
Cheers,
Kirk Bailey
--- In m..., Matthias Weingart wrote:
>
> Kirk ask for a OpAmp inside a small MSP. The 2013 does not contain a OpAmp.
> Just the SD16 with 200kOhm input impedance. To be more specific - he ask for
> a high impedance ADC.
>
> I have used the ADC12 as a ADC with a high impedance input. Solution is to
> increase sample time to measure high impedance networks. You need to charge
> the 40pF input capacitor (this is inside the MSP430). To get the 1/4096
> resolution of the 12 bit ADC the sample time has to be at least 9*R*C. (C> 40pF, R your input impedance), for 10Megohm this is 3.6ms sample time.
> Same is true for the ADC10. The input impedance of the ADC10 is 27pF.
>
> The resolution achieveable with Comparator A is approx. 10bits.
>
> M.
>
