From: sademade (sademade_at_yahoo.com)
Date: 2001-12-20 10:24:02
> Yeah, that one is nice. There are also much higer speed parts at
> amazing low price like ADS1252. I like the idea of using heavy
> oversampling instead of analog filters
> to solve the aliasing problem. Price the same, performance better.
Yes, for this particular application, it might be better, for
my particular application, ADS1211 was better, because had to
have 4 channels, space was limited, the SSOP version of the chip
provided me smallest space with highest performance per that space,
and lowest price.
Here, as it does not make difference if the cost is $15 or $20
more, really, one could surely throw in 4 separate converters,
and make the small risc actually do something for the money spent
on it.
Personally feel 4 channels as a minimum module, as quite many
mood/condition detections require access to 3-4 points, it would
nice to have at least 4 channels, and if absolute
performance is the most important thing, and if one could not afford
a few sample periods of wasted time, one could consider
even 4 of the 1252's, as one could still throw out the additional TL
amp sections, which TL's cost pretty much same as the additional
converters, and take more space, or how ...
> Maybe a DSP would be overkill. But a risc controller in combination
> with the FTDI parts would certainly also be nice. I always thought
> USB would be overkill. But now I'm working on a laboratory
> measuring instrument for plasma physics using USB and since I
> learned about the FTDI parts I think it's very easy :)
Nice to hear that the FTDI parts and drivers are easy to use in
practice as well. Only used Uarts, SPI, I2C, CAN and DSP sync
serial in the past, yet not fearing USB too much, as it
can not be a quantum leap in complexity compared to CAN imho.
> But for now, let's stick to RS232 and try to get the modularEEG
> going.
> Moritz
Appr. your help, and agree that one should absolutely
stick to getting the modularEEG amp section working as the very
first thing, before even any RS232 or USB's. Amp is the most
important module right now.
Q1: Why not just cut the working amp from the earlier "monolithium"
( rs232eeg) as one gets the impression it actually worked with
fairly little trouble ?
Q2: are the blue outlines over the original rs232eeg board
for copper-box-shielding the amp-sections from 60 Hz, similar
to many commercial PC A/D cards that otherwise would suffer
from too much noise ... ?
Q3: Is there any realistic way to get low-noise enough design without
the copper/nickel plated shieldboxes ?
Q4: Why is the right-leg grounding best, instead, of, say, left-
ear grounding, or middle-finger grounding ?
Q5: Could you please let me know what is your favorite book or
resource for learning analog shielding/grounding/guarding techniques
and their optimal PCB layout techniques, so I could attempt some
alternativewhat-if trials, even if they are worthless, but just would
like to learn to come up with decent low-noise, nano-microvolt range
amps, as there are those UV-photodiode applications also, where one
could benefit from learning the low noise techniques.
Thank You A Lot, Moritz,
Sade M.
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:35 BST