I'm trying to measure battery voltage (12V) using the LPC1768. I was
planning to use some voltage dividers (100K in series and 33K to GND)and connect
to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage
range. I wanted to ensure that the LPC1768 ADC's input impedance was high
enough for these voltage dividers. Upon going through the ADC data in the
LPC1768 datasheet I could not locate any data for the input impedance of ADC but
noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm
max. Is this the input impedance of the ADC? It seems to be very low for an ADC.
Normally I've seen ADCs have impedance of several megaohms.
Can you please reply with your thoughts on this?
LPC1678 ADC input impedance
Started by ●August 25, 2010
Reply by ●August 25, 20102010-08-25
--- In l..., naveen_pn@... wrote:
>
> I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
>
> Can you please reply with your thoughts on this?
>
You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
Figure 29 pretty much says it all
http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
The proper way to do this is to use an op-amp to rescale (and potentially offset) the input voltage while providing a low output impedance. See Chapter 4 of Op Amps For Everyone
http://focus.ti.com/lit/an/slod006b/slod006b.pdf
For example, you may not really want to measure 0V..12V. You might be interested in just 11V..12V You can offset the input by subtracting 11V (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this with one op amp and 4 resistors.
Richard
>
> I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
>
> Can you please reply with your thoughts on this?
>
You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
Figure 29 pretty much says it all
http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
The proper way to do this is to use an op-amp to rescale (and potentially offset) the input voltage while providing a low output impedance. See Chapter 4 of Op Amps For Everyone
http://focus.ti.com/lit/an/slod006b/slod006b.pdf
For example, you may not really want to measure 0V..12V. You might be interested in just 11V..12V You can offset the input by subtracting 11V (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this with one op amp and 4 resistors.
Richard
Reply by ●August 25, 20102010-08-25
--- In l..., "rtstofer" wrote:
>
> --- In l..., naveen_pn@ wrote:
> >
> > I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> >
> > Can you please reply with your thoughts on this?
> > You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
>
> Figure 29 pretty much says it all
> http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
>
> Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
You could probably find the typical input impedance of the A/D by connecting a variable resistance to say, +3.3 Volts and use either a high impedance meter or the A/D conversion output itself to get an idea of how much that A/D input loads the signal down.
Because the A/D input is multiplexed, a meter or scope on that input may show a 2 voltage step wave as it is selected and un-selected by that multiplexer. You may or may not want to add a capacitor from A/D input to ground which should smooth out the mux switching.
boB
>
> The proper way to do this is to use an op-amp to rescale (and potentially offset) the input voltage while providing a low output impedance. See Chapter 4 of Op Amps For Everyone
> http://focus.ti.com/lit/an/slod006b/slod006b.pdf
>
> For example, you may not really want to measure 0V..12V. You might be interested in just 11V..12V You can offset the input by subtracting 11V (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this with one op amp and 4 resistors.
>
> Richard
>
>
> --- In l..., naveen_pn@ wrote:
> >
> > I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> >
> > Can you please reply with your thoughts on this?
> > You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
>
> Figure 29 pretty much says it all
> http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
>
> Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
You could probably find the typical input impedance of the A/D by connecting a variable resistance to say, +3.3 Volts and use either a high impedance meter or the A/D conversion output itself to get an idea of how much that A/D input loads the signal down.
Because the A/D input is multiplexed, a meter or scope on that input may show a 2 voltage step wave as it is selected and un-selected by that multiplexer. You may or may not want to add a capacitor from A/D input to ground which should smooth out the mux switching.
boB
>
> The proper way to do this is to use an op-amp to rescale (and potentially offset) the input voltage while providing a low output impedance. See Chapter 4 of Op Amps For Everyone
> http://focus.ti.com/lit/an/slod006b/slod006b.pdf
>
> For example, you may not really want to measure 0V..12V. You might be interested in just 11V..12V You can offset the input by subtracting 11V (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this with one op amp and 4 resistors.
>
> Richard
>
Reply by ●August 25, 20102010-08-25
ADC have a leakage current that you have to account for. An op-amp solves
that problem, but ADC usually do a sample and hold for a short duration.
This can drag the output of a slow speed op-amp down, so a high speed
op-amp is usually needed.
If you don't need a high rate of ADC conversions, a resistor divider and
ceramic capacitor on the input of the ADC may be all you need. The divider
current must be high enough that ADC current doesn't kill you error budget.
the capacitor has to be big enough so that the dynamic current into the ADC
doesn't change it's voltage, killing your error budget.
DaveS
On Wed, Aug 25, 2010 at 1:32 PM, bobtransformer wrote:
> --- In l... , "rtstofer"
> wrote:
> >
> >
> >
> > --- In l... , naveen_pn@wrote:
> > >
> > > I'm trying to measure battery voltage (12V) using the LPC1768. I was
> planning to use some voltage dividers (100K in series and 33K to GND)and
> connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery
> voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was
> high enough for these voltage dividers. Upon going through the ADC data in
> the LPC1768 datasheet I could not locate any data for the input impedance of
> ADC but noticed a parameter "Rvsi voltage source interface resistance" which
> is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very
> low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> > >
> > > Can you please reply with your thoughts on this?
> > >
> >
> > You will find that the A/D devices inside uC's will require very low
> source impedance. I have never seen the input impedance specified but it is
> not uncommon to have the datasheet specify a low driving impedance - often
> 10k or less.
> >
> > Figure 29 pretty much says it all
> > http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
> >
> > Since you can't find the input impedance of the A/D, you can't calculate
> the voltage divider.
>
> You could probably find the typical input impedance of the A/D by
> connecting a variable resistance to say, +3.3 Volts and use either a high
> impedance meter or the A/D conversion output itself to get an idea of how
> much that A/D input loads the signal down.
>
> Because the A/D input is multiplexed, a meter or scope on that input may
> show a 2 voltage step wave as it is selected and un-selected by that
> multiplexer. You may or may not want to add a capacitor from A/D input to
> ground which should smooth out the mux switching.
>
> boB
> >
> > The proper way to do this is to use an op-amp to rescale (and potentially
> offset) the input voltage while providing a low output impedance. See
> Chapter 4 of Op Amps For Everyone
> > http://focus.ti.com/lit/an/slod006b/slod006b.pdf
> >
> > For example, you may not really want to measure 0V..12V. You might be
> interested in just 11V..12V You can offset the input by subtracting 11V
> (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this
> with one op amp and 4 resistors.
> >
> > Richard
> >
>
that problem, but ADC usually do a sample and hold for a short duration.
This can drag the output of a slow speed op-amp down, so a high speed
op-amp is usually needed.
If you don't need a high rate of ADC conversions, a resistor divider and
ceramic capacitor on the input of the ADC may be all you need. The divider
current must be high enough that ADC current doesn't kill you error budget.
the capacitor has to be big enough so that the dynamic current into the ADC
doesn't change it's voltage, killing your error budget.
DaveS
On Wed, Aug 25, 2010 at 1:32 PM, bobtransformer wrote:
> --- In l... , "rtstofer"
> wrote:
> >
> >
> >
> > --- In l... , naveen_pn@wrote:
> > >
> > > I'm trying to measure battery voltage (12V) using the LPC1768. I was
> planning to use some voltage dividers (100K in series and 33K to GND)and
> connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery
> voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was
> high enough for these voltage dividers. Upon going through the ADC data in
> the LPC1768 datasheet I could not locate any data for the input impedance of
> ADC but noticed a parameter "Rvsi voltage source interface resistance" which
> is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very
> low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> > >
> > > Can you please reply with your thoughts on this?
> > >
> >
> > You will find that the A/D devices inside uC's will require very low
> source impedance. I have never seen the input impedance specified but it is
> not uncommon to have the datasheet specify a low driving impedance - often
> 10k or less.
> >
> > Figure 29 pretty much says it all
> > http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
> >
> > Since you can't find the input impedance of the A/D, you can't calculate
> the voltage divider.
>
> You could probably find the typical input impedance of the A/D by
> connecting a variable resistance to say, +3.3 Volts and use either a high
> impedance meter or the A/D conversion output itself to get an idea of how
> much that A/D input loads the signal down.
>
> Because the A/D input is multiplexed, a meter or scope on that input may
> show a 2 voltage step wave as it is selected and un-selected by that
> multiplexer. You may or may not want to add a capacitor from A/D input to
> ground which should smooth out the mux switching.
>
> boB
> >
> > The proper way to do this is to use an op-amp to rescale (and potentially
> offset) the input voltage while providing a low output impedance. See
> Chapter 4 of Op Amps For Everyone
> > http://focus.ti.com/lit/an/slod006b/slod006b.pdf
> >
> > For example, you may not really want to measure 0V..12V. You might be
> interested in just 11V..12V You can offset the input by subtracting 11V
> (resulting in 0V..1V) and then rescale to 0V..3V (multply by 3). All this
> with one op amp and 4 resistors.
> >
> > Richard
> >
>
Reply by ●August 25, 20102010-08-25
--- In l..., "bobtransformer" wrote:
>
> --- In l..., "rtstofer" wrote:
> >
> >
> >
> > --- In l..., naveen_pn@ wrote:
> > >
> > > I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> > >
> > > Can you please reply with your thoughts on this?
> > >
> >
> > You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
> >
> > Figure 29 pretty much says it all
> > http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
> >
> > Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
>
> You could probably find the typical input impedance of the A/D by connecting a variable resistance to say, +3.3 Volts and use either a high impedance meter or the A/D conversion output itself to get an idea of how much that A/D input loads the signal down.
>
> Because the A/D input is multiplexed, a meter or scope on that input may show a 2 voltage step wave as it is selected and un-selected by that multiplexer. You may or may not want to add a capacitor from A/D input to ground which should smooth out the mux switching.
>
> boB
I haven't seen the A/D input characterized on datasheets. It may turn out that while it is possible to measure the value for a channel, that value may not be the same for all channels and certainly not for all devices.
I would tend to want to design around the uncertainty.
Richard
>
> --- In l..., "rtstofer" wrote:
> >
> >
> >
> > --- In l..., naveen_pn@ wrote:
> > >
> > > I'm trying to measure battery voltage (12V) using the LPC1768. I was planning to use some voltage dividers (100K in series and 33K to GND)and connect to ADC the divided voltage to get a 0-3V ADC input for 0-12V battery voltage range. I wanted to ensure that the LPC1768 ADC's input impedance was high enough for these voltage dividers. Upon going through the ADC data in the LPC1768 datasheet I could not locate any data for the input impedance of ADC but noticed a parameter "Rvsi voltage source interface resistance" which is 7.5kohm max. Is this the input impedance of the ADC? It seems to be very low for an ADC. Normally I've seen ADCs have impedance of several megaohms.
> > >
> > > Can you please reply with your thoughts on this?
> > >
> >
> > You will find that the A/D devices inside uC's will require very low source impedance. I have never seen the input impedance specified but it is not uncommon to have the datasheet specify a low driving impedance - often 10k or less.
> >
> > Figure 29 pretty much says it all
> > http://www.nxp.com/documents/data_sheet/LPC1768_66_65_64.pdf
> >
> > Since you can't find the input impedance of the A/D, you can't calculate the voltage divider.
>
> You could probably find the typical input impedance of the A/D by connecting a variable resistance to say, +3.3 Volts and use either a high impedance meter or the A/D conversion output itself to get an idea of how much that A/D input loads the signal down.
>
> Because the A/D input is multiplexed, a meter or scope on that input may show a 2 voltage step wave as it is selected and un-selected by that multiplexer. You may or may not want to add a capacitor from A/D input to ground which should smooth out the mux switching.
>
> boB
I haven't seen the A/D input characterized on datasheets. It may turn out that while it is possible to measure the value for a channel, that value may not be the same for all channels and certainly not for all devices.
I would tend to want to design around the uncertainty.
Richard
Reply by ●August 25, 20102010-08-25
On 8/25/2010 2:40 PM, n...@yahoo.com wrote:
> I'm trying to measure battery voltage (12V) using the LPC1768. I was
> planning to use some voltage dividers (100K in series and 33K to
> GND)and connect to ADC the divided voltage to get a 0-3V ADC input
> for 0-12V battery voltage range. I wanted to ensure that the LPC1768
> ADC's input impedance was high enough for these voltage dividers.
> Upon going through the ADC data in the LPC1768 datasheet I could not
> locate any data for the input impedance of ADC but noticed a
> parameter "Rvsi voltage source interface resistance" which is 7.5kohm
> max. Is this the input impedance of the ADC? It seems to be very low
> for an ADC. Normally I've seen ADCs have impedance of several
> megaohms.
>
> Can you please reply with your thoughts on this?
It's not the A/D input impedance you are seeing but rather the sample
and hold in front of the A/D.
I've not seen a multiplexed uP A/D input that is anywhere near the MOhm
range. The inputs should always have an RC filter in front for two reasons.
1 - To filter out high frequency signals (especially impulse). Note
this is true even for sources that are supposedly bandwidth limited.
2 - To provide a low impedance source of charge that will allow the
S/H cap to charge or discharge without affecting the voltage level
significantly.
A good op-amp filter/driver could be used as well.
Robert
--
http://www.aeolusdevelopment.com/
From the Divided by a Common Language File (Edited to protect the guilty)
ME - "I'd like to get Price and delivery for connector Part # XXXXX"
Dist./Rep - "$X.XX Lead time 37 days"
ME - "Anything we can do about lead time? 37 days seems a bit high."
Dist./Rep - "that is the lead time given because our stock is live....
we currently have stock."
> I'm trying to measure battery voltage (12V) using the LPC1768. I was
> planning to use some voltage dividers (100K in series and 33K to
> GND)and connect to ADC the divided voltage to get a 0-3V ADC input
> for 0-12V battery voltage range. I wanted to ensure that the LPC1768
> ADC's input impedance was high enough for these voltage dividers.
> Upon going through the ADC data in the LPC1768 datasheet I could not
> locate any data for the input impedance of ADC but noticed a
> parameter "Rvsi voltage source interface resistance" which is 7.5kohm
> max. Is this the input impedance of the ADC? It seems to be very low
> for an ADC. Normally I've seen ADCs have impedance of several
> megaohms.
>
> Can you please reply with your thoughts on this?
It's not the A/D input impedance you are seeing but rather the sample
and hold in front of the A/D.
I've not seen a multiplexed uP A/D input that is anywhere near the MOhm
range. The inputs should always have an RC filter in front for two reasons.
1 - To filter out high frequency signals (especially impulse). Note
this is true even for sources that are supposedly bandwidth limited.
2 - To provide a low impedance source of charge that will allow the
S/H cap to charge or discharge without affecting the voltage level
significantly.
A good op-amp filter/driver could be used as well.
Robert
--
http://www.aeolusdevelopment.com/
From the Divided by a Common Language File (Edited to protect the guilty)
ME - "I'd like to get Price and delivery for connector Part # XXXXX"
Dist./Rep - "$X.XX Lead time 37 days"
ME - "Anything we can do about lead time? 37 days seems a bit high."
Dist./Rep - "that is the lead time given because our stock is live....
we currently have stock."
Reply by ●August 26, 20102010-08-26
Thanks for all of your inputs which are very helpful. It seems like we have to
add an opamp buffer to overcome the low ADC input impedance. We are used to the
Atmel AVR series which have an input impedance of 100Mohms so doesn't
require any additional buffering. Using an opamp buffer means we also need to
add a negative voltage generator to ensure true rail to rail performance for
input signals.
On another note, since the 8 ADC channels provided by the LPC1768 aren't enough for our application, we want to use several external analog mux such as the CD4051. It would seem that the best configuration would be to have the analog input going to the mux first and a single opamp between the mux output & LPC1768's ADC. Or would it be better to have an opamp for every input and put the mux between opamp output and the LPC1768's ADC? Does this make any difference?
On another note, since the 8 ADC channels provided by the LPC1768 aren't enough for our application, we want to use several external analog mux such as the CD4051. It would seem that the best configuration would be to have the analog input going to the mux first and a single opamp between the mux output & LPC1768's ADC. Or would it be better to have an opamp for every input and put the mux between opamp output and the LPC1768's ADC? Does this make any difference?
Reply by ●August 26, 20102010-08-26
--- In l..., Robert Adsett wrote:
>
> On 8/25/2010 2:40 PM, naveen_pn@... wrote:
> > I'm trying to measure battery voltage (12V) using the LPC1768. I was
> > planning to use some voltage dividers (100K in series and 33K to
> > GND)and connect to ADC the divided voltage to get a 0-3V ADC input
> > for 0-12V battery voltage range. I wanted to ensure that the LPC1768
> > ADC's input impedance was high enough for these voltage dividers.
> > Upon going through the ADC data in the LPC1768 datasheet I could not
> > locate any data for the input impedance of ADC but noticed a
> > parameter "Rvsi voltage source interface resistance" which is 7.5kohm
> > max. Is this the input impedance of the ADC? It seems to be very low
> > for an ADC. Normally I've seen ADCs have impedance of several
> > megaohms.
> >
> > Can you please reply with your thoughts on this?
>
> It's not the A/D input impedance you are seeing but rather the sample
> and hold in front of the A/D.
>
> I've not seen a multiplexed uP A/D input that is anywhere near the MOhm
> range. The inputs should always have an RC filter in front for two reasons.
> 1 - To filter out high frequency signals (especially impulse). Note
> this is true even for sources that are supposedly bandwidth limited.
> 2 - To provide a low impedance source of charge that will allow the
> S/H cap to charge or discharge without affecting the voltage level
> significantly.
>
They very well could be using a sample and hold, but it shouldn't be necessary if the mux stays on that channel for the entire conversion cycle.
boB
> A good op-amp filter/driver could be used as well.
>
> Robert
>
> --
> http://www.aeolusdevelopment.com/
>
> From the Divided by a Common Language File (Edited to protect the guilty)
> ME - "I'd like to get Price and delivery for connector Part # XXXXX"
> Dist./Rep - "$X.XX Lead time 37 days"
> ME - "Anything we can do about lead time? 37 days seems a bit high."
> Dist./Rep - "that is the lead time given because our stock is live....
> we currently have stock."
>
>
> On 8/25/2010 2:40 PM, naveen_pn@... wrote:
> > I'm trying to measure battery voltage (12V) using the LPC1768. I was
> > planning to use some voltage dividers (100K in series and 33K to
> > GND)and connect to ADC the divided voltage to get a 0-3V ADC input
> > for 0-12V battery voltage range. I wanted to ensure that the LPC1768
> > ADC's input impedance was high enough for these voltage dividers.
> > Upon going through the ADC data in the LPC1768 datasheet I could not
> > locate any data for the input impedance of ADC but noticed a
> > parameter "Rvsi voltage source interface resistance" which is 7.5kohm
> > max. Is this the input impedance of the ADC? It seems to be very low
> > for an ADC. Normally I've seen ADCs have impedance of several
> > megaohms.
> >
> > Can you please reply with your thoughts on this?
>
> It's not the A/D input impedance you are seeing but rather the sample
> and hold in front of the A/D.
>
> I've not seen a multiplexed uP A/D input that is anywhere near the MOhm
> range. The inputs should always have an RC filter in front for two reasons.
> 1 - To filter out high frequency signals (especially impulse). Note
> this is true even for sources that are supposedly bandwidth limited.
> 2 - To provide a low impedance source of charge that will allow the
> S/H cap to charge or discharge without affecting the voltage level
> significantly.
>
They very well could be using a sample and hold, but it shouldn't be necessary if the mux stays on that channel for the entire conversion cycle.
boB
> A good op-amp filter/driver could be used as well.
>
> Robert
>
> --
> http://www.aeolusdevelopment.com/
>
> From the Divided by a Common Language File (Edited to protect the guilty)
> ME - "I'd like to get Price and delivery for connector Part # XXXXX"
> Dist./Rep - "$X.XX Lead time 37 days"
> ME - "Anything we can do about lead time? 37 days seems a bit high."
> Dist./Rep - "that is the lead time given because our stock is live....
> we currently have stock."
>
Reply by ●August 26, 20102010-08-26
You should read the AVR manual which says:
"The ADC is optimized for analog signals with an output impedance of
approximately 10 kΩ or
less. If such a source is used, the sampling time will be negligible. If a
source with higher impedance
is used, the sampling time will depend on how long time the source needs to
charge the
S/H capacitor, with can vary widely. The user is recommended to only use low
impedant sources
with slowly varying signals, since this minimizes the required charge
transfer to the S/H
capacitor."
A CD4051 will probably introduce switching transients into the system, so a
better analog mux is probably required.
If you put an op amp between the mux and adc, that op amp has to have a high
slew rate and low settling time -- maybe a jfet input amp.
DaveS
On Thu, Aug 26, 2010 at 8:11 AM, wrote:
> Thanks for all of your inputs which are very helpful. It seems like we have
> to add an opamp buffer to overcome the low ADC input impedance. We are used
> to the Atmel AVR series which have an input impedance of 100Mohms so doesn't
> require any additional buffering. Using an opamp buffer means we also need
> to add a negative voltage generator to ensure true rail to rail performance
> for input signals.
>
> On another note, since the 8 ADC channels provided by the LPC1768 aren't
> enough for our application, we want to use several external analog mux such
> as the CD4051. It would seem that the best configuration would be to have
> the analog input going to the mux first and a single opamp between the mux
> output & LPC1768's ADC. Or would it be better to have an opamp for every
> input and put the mux between opamp output and the LPC1768's ADC? Does this
> make any difference?
>
>
"The ADC is optimized for analog signals with an output impedance of
approximately 10 kΩ or
less. If such a source is used, the sampling time will be negligible. If a
source with higher impedance
is used, the sampling time will depend on how long time the source needs to
charge the
S/H capacitor, with can vary widely. The user is recommended to only use low
impedant sources
with slowly varying signals, since this minimizes the required charge
transfer to the S/H
capacitor."
A CD4051 will probably introduce switching transients into the system, so a
better analog mux is probably required.
If you put an op amp between the mux and adc, that op amp has to have a high
slew rate and low settling time -- maybe a jfet input amp.
DaveS
On Thu, Aug 26, 2010 at 8:11 AM, wrote:
> Thanks for all of your inputs which are very helpful. It seems like we have
> to add an opamp buffer to overcome the low ADC input impedance. We are used
> to the Atmel AVR series which have an input impedance of 100Mohms so doesn't
> require any additional buffering. Using an opamp buffer means we also need
> to add a negative voltage generator to ensure true rail to rail performance
> for input signals.
>
> On another note, since the 8 ADC channels provided by the LPC1768 aren't
> enough for our application, we want to use several external analog mux such
> as the CD4051. It would seem that the best configuration would be to have
> the analog input going to the mux first and a single opamp between the mux
> output & LPC1768's ADC. Or would it be better to have an opamp for every
> input and put the mux between opamp output and the LPC1768's ADC? Does this
> make any difference?
>
>
Reply by ●August 26, 20102010-08-26
On 8/26/2010 3:49 PM, bobtransformer wrote:
> --- In l..., Robert Adsett
> wrote:
>> I've not seen a multiplexed uP A/D input that is anywhere near the
>> MOhm range. The inputs should always have an RC filter in front for
>> two reasons. 1 - To filter out high frequency signals (especially
>> impulse). Note this is true even for sources that are supposedly
>> bandwidth limited. 2 - To provide a low impedance source of charge
>> that will allow the S/H cap to charge or discharge without
>> affecting the voltage level significantly.
>> They very well could be using a sample and hold, but it shouldn't be
> necessary if the mux stays on that channel for the entire conversion
> cycle.
A sample and hold is part of any multiplexed uP A/D I have ever seen.
It's unavoidable. Even staying on the same channel doesn't avoid it,
the circuitry will still do a sample and hold. In some systems the
behaviour is completely fixed, in others you have some control over
timing such as the sampling time. In no case that I'm aware of can you
turn it off and you would never want to.
Robert
--
http://www.aeolusdevelopment.com/
From the Divided by a Common Language File (Edited to protect the guilty)
ME - "I'd like to get Price and delivery for connector Part # XXXXX"
Dist./Rep - "$X.XX Lead time 37 days"
ME - "Anything we can do about lead time? 37 days seems a bit high."
Dist./Rep - "that is the lead time given because our stock is live....
we currently have stock."
> --- In l..., Robert Adsett
> wrote:
>> I've not seen a multiplexed uP A/D input that is anywhere near the
>> MOhm range. The inputs should always have an RC filter in front for
>> two reasons. 1 - To filter out high frequency signals (especially
>> impulse). Note this is true even for sources that are supposedly
>> bandwidth limited. 2 - To provide a low impedance source of charge
>> that will allow the S/H cap to charge or discharge without
>> affecting the voltage level significantly.
>> They very well could be using a sample and hold, but it shouldn't be
> necessary if the mux stays on that channel for the entire conversion
> cycle.
A sample and hold is part of any multiplexed uP A/D I have ever seen.
It's unavoidable. Even staying on the same channel doesn't avoid it,
the circuitry will still do a sample and hold. In some systems the
behaviour is completely fixed, in others you have some control over
timing such as the sampling time. In no case that I'm aware of can you
turn it off and you would never want to.
Robert
--
http://www.aeolusdevelopment.com/
From the Divided by a Common Language File (Edited to protect the guilty)
ME - "I'd like to get Price and delivery for connector Part # XXXXX"
Dist./Rep - "$X.XX Lead time 37 days"
ME - "Anything we can do about lead time? 37 days seems a bit high."
Dist./Rep - "that is the lead time given because our stock is live....
we currently have stock."