[hpsdr] Neat front end from Phil

[Ahti Aintila]

On 29/07/06, Phil Harman <pvharman at arach.net.au> wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> Phil's front end does look very nice and it will be interesting to see the
> performance figures.
>
> Just a cavet when selecting amplifiers for this type of circuit. There are
> actually two noise sources that need to be considered  - the amplifier
> current noise and amplifier voltage noise.
>
> Not taking the current noise  and source impedance into account can lead to
> the wrong conclusions.
>
> Based in voltage noise one would select the SSM2019  over the OPA227 .
> However, if you do the math and include both noise sources and the effect of
> the source impedance then this what happens
>
>      Source Z (ohms) SSM2019 NF (dB) OPA227 NF (dB)
>      10 4.2 8.7
>      50 1.7 5.4
>      100 1.1 4
>      150 0.8 3.3
>      300 0.5 2.2
>      500 0.5 1.6
>      1000 0.6 1
>      2000 0.9 0.6
>      5000 1.7 0.3
>      10000 2.7 0.3
>
>
> As you can see the even though the OPA227 has 3 times the voltage noise of
> the SSM2019 it also has 1/5 the noise current so wins with high Z sources.
>
> At low Z voltage noise dominates so we need an opamp with low voltage noise.
> At high Z current noise dominates so we select one with low current noise.
>
> I suggest that we need to look for an opamp designed to be fed from a low Z,
> like a microphone preamp, for this application.  Also using the device in an
> non-inverting mode can have a significant impact on the system noise figure.
>
> Given the above the LT1128 may perform better than the OPA1632.
>
>
> 73's Phil...VK6APH

Phil_H,

This is really a place of compromise for the impedances and used
gains. The noise specifications of the different opamps are all given
in different ways. At first SSM2019 looks very good, but it is good
only at high gains (40 dB). Those other types you mentioned are
better, if the gain is low. For wider dynamic range we possibly need a
rather low gain to avoid ADC overloading at strong signals.

On the other hand, in this sampling configuration the opamp
(integrator or plain sampling capacitor) "sees" the low impedance of
the input transformer 50% of the time. Another 50% is determined by
the feedback resistor (or any other load across the capacitor).

Without any additional math and exact specifications, I still feel
that the performance with either LT1127 or OPA1632 is not very much
different. If I am wrong, please inform. Now I'm drafting the test
layout for OPA1632 and no LT1127 available locally. If needed, I have
to ask for samples from Linear Technology.

73, Ahti OH2RZ

_______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at hpsdr.org
> Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> HPSDR web page: http://hpsdr.org
> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
>

 1154162755.0