Voltage amp solution needed

Hi Group

Ive recently purchased a 10-bit resolution data acquisition card with a
range of -1V to 1V sampling at upto 20MHz.
Upon running a few tests ive realised that the raw voltage signal that i am
measuring varies between around -20mV and 20mV. My problem is that the
resolution in that range is not all that great and i fear it is too late to
exchange for a 16-bit card (which would do the job at a much higher cost).
Is there a basic solution where i could construct or purchase an amplifier
off the shelf that would condition the raw signal to use the entire span of
the card(-1V to 1V) and is also capable of performing at frequencies upto
20MHz?

Thanks

Peter

Pete wrote:
>
> Hi Group
>
> Ive recently purchased a 10-bit resolution data acquisition card with a
> range of -1V to 1V sampling at upto 20MHz.
> Upon running a few tests ive realised that the raw voltage signal that i am
> measuring varies between around -20mV and 20mV. My problem is that the
> resolution in that range is not all that great and i fear it is too late to
> exchange for a 16-bit card (which would do the job at a much higher cost).
> Is there a basic solution where i could construct or purchase an amplifier
> off the shelf that would condition the raw signal to use the entire span of
> the card(-1V to 1V) and is also capable of performing at frequencies upto
> 20MHz?
>
> Thanks
>
> Peter

You can build a fairly simple 5X ampifier that is pretty flat up to 20
MHz with something like an LM7171 ($2.80 from digikey) if you can come
up with a positive and negative supply. The offset specs are not
perfect, but you can measure that with your A/D converter board and
subtract it from the other measured results.

--
John Popelish

Pete wrote:
>
> Thanks John
>
> I have had no prior experience with OP-amps so this will be a pretty steep
> learning curve of which i am keen on pursuing if it will do the trick. Any
> chance that you would have a circuit schematic to use in conjunction with
> the LM7171 that would suit this application?

The video line driver on page 16 of the data sheet:
http://cache.national.com/ds/LM/LM7171.pdf

gives the general idea, except that the two resistors in the voltage
divider to the - input have to have a ratio of 4 to 1, instead of
being equal, to give an overall gain of 5 (say 510 to ground and 2k
feedback, with, perhaps a 100 ohm trim pot between them (with the
wiper to the - input) as a gain trim. And I am assuming that you
would put this right next ot your A/D input, so no series terminating
resistor would be needed on the output.

Read the application notes section for how to bypass the supply lines
for such a high frequency opamp.

--
John Popelish