Curious but the minimal input impedance a the mux input is not given in the
datasheet.
Since it's a sample and hold amplifier, the input impedance on the mux
should charge/discharge the SH capacitor at every next multiplexed channel
voltage value.

Vasile

On Fri, Sep 4, 2015 at 12:06 AM, Bob Blick <bobblick@ftml.net> wrote:

> Hi Mitch,
>
> What you're seeing is normal. You have one A/D converter with a
> multichannel analog switch ahead of it. The issue you are seeing is due
> to the capacitive load the converter has. When you switch to another
> input, the converter's capacitor still has the previous channel's
> voltage on it.
>
> You need to present a low impedance to each input channel. Since you're
> only sampling once per second, just put a "large" capacitor (0.01 to 0.1
> uF) to ground from each input (you'll need 8 capacitors).
>
> That will solve most of the channel-to-channel problems by reducing the
> AC impedance. You will still have input bias current (DC) errors, but
> they will be small compared to what you've been seeing.
>
> Best regards, Bob
>
> On Thu, Sep 3, 2015, at 01:44 PM, Mitch wrote:
> > I am fairly new to multi-channel ADC sampling.
> >
> > I have a ADS7828 and trying to sample all 8 channels - just measure all
> > of
> > them once a second.
> >
> > I found that if I sample only one channel, I get very little noise and
> > the
> > values look reasonable.
> >
> > But if I sample more than just 1 channel, it seems like the value of on=
e
> > sample has some kind of "leakage" to the value of the next. By increasi=
ng
> > the voltage on channel 1, I actually measure a voltage increase on
> > channel
> > 2. The "leak" occurs all the way down the chain.
> >
> > So what should be something like:
> >
> > 1024  0  0  0  0  0  0  0
> > Is giving:
> > 1024 1024 950 870 780 690 600 550
> >
> > Sometime it's even showing a bigger rollover than that:
> > 1024 1024 1024 1022 1020 1017 1015 1012
> >
> >
> > 1 Hz from each channel every second is well below the ability of the
> > chip,
> > but the data sheet doesn't address anything like this. WTF am I doing
> > wrong? Is it an error in my I2C communication? The actual wiring? Bad
> > chip?
> >
> > I'm sending Single-ended input mode, internal ref off, ADC alway on, wi=
th
> > an external reference of 5V:
> > 0b1xxx01xx
> >
> > Any help would be appreciated.
>
> --
> http://www.fastmail.com - A fast, anti-spam email service.
>
> --
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>
--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.