bandlimited pseudo noise

[Magnus Danielson]

From: Martin Czech <czech at Micronas.Com>
Subject: Re: bandlimited pseudo noise
Date: Mon, 14 Aug 2000 09:18:01 +0200 (MET DST)

> Magnus, all very well, but specialized hardware means time
> I do not have. 

Martin, the lowest range of device is just an upgraded version of what I have
in the CMOS catalog, and it contains

4011   1
4013   1
4015   2
4016   1
4519   2
A clocksource
4 resistors
A capacitor
2 pushbuttons

Add two or four 4015 and you are getting the LMS length you want. Add a fixed
antialias filter and you are pretty much there. Output gain is missing.
Input and output buffers is good.
Then you need to some recording method.

It is really not very much hardware, that's my point. Trying to do advanced
filters and/or noise-sources that you bring with you is really alot of
hardware. Only if you want to make measurement speed higher you are required
to do more advanced tricks. You can make several improvements on the curcuit,
better CMOS switches, longer PRN, antialias filter, etc.

Given the complexity you want it should be a one or two evening project on a
normal veroboard.

Now, if you still consider such a project out of your league, then naturally
you should look on what you can do by using the stuff you have.

> So I need to use the hardware which is allready there, and 
> this will suffer from alias when taking the reference signal
> by electrically tapping the noise generator.

Maybe one really should check how much of a problem antialiasing is on the
DAT. If one records a sequence of frequencies > 20kHz and see what pops out
on the other side, what strengths and what frequencies.

> I see no problems in doing corelation at home
> on a cheap computer.
> 
> But another type of hardware could be usefull:
> 
> An instant speaker optimizer via inverse filtering.
> So far the units you can by are in the 10000$ dollar range...

Oh... those... yes, they are expensive...

Cheers,
Magnus