[sdiy] Cheap audio spectrum analyzer

Wed Oct 28 20:32:56 CET 2015

Ditto likewise. I have a lab full of moderately ancient, well maintained
and truly great instruments that in many cases cost me almost nothing. They
let me get the job done, so I can ask for little more.

I will confess to an excessive liking for bench multimeters, however,
particularly the better Keithleys and HPs...

Sarah

On Wed, Oct 28, 2015 at 12:12 PM, <rsdio at audiobanshee.com> wrote:

> Excellent article, Neil.
>
> For the same reasons you've described, I've not built my own SMD oven out
> of a used Toast-R-Oven plus Arduino. I'd rather pay the local assembly shop
> $30 and have them use the right tools for the job (not to mention, they
> have pick-and-place robots in addition to the SMD ovens).
>
> I'd rather design things that don't exist yet, or at least don't exist any
> more (vintage synths). Designing things that already exist is a great
> exercise for learning, but it's a bad choice if you need to get other
> things done.
>
> Brian
>
>
> On Oct 28, 2015, at 4:18 AM, Neil Johnson <neil.johnson71 at gmail.com>
> wrote:
> > Hi Rick,
> >
> >>
> http://www.cnx-software.com/2013/07/24/359-red-pitaya-board-combines-an-oscilloscope-a-spectrum-analyser-a-waveform-generator-and-more/
> >>
> >> ~360 US$
> >
> > I wrote up my thoughts on this board last year on muffwiggler:
> >
> >
> https://www.muffwiggler.com/forum/viewtopic.php?p=1576218&highlight=#1576218
> >
> > -------
> >
> > I think I'm about to be flamed for the following. This project is an
> > over-priced half-baked time-waster. Allow me to justify that.
> >
> > Firstly, the price. I just checked the RS website, and including VAT
> > one of those would cost me £378.00 + postage. For that I get a fragile
> > board that might be good as a wideband noise source, but not much
> > else. To make it a useful instrument, one that could be fairly
> > compared to "old T&M gear on eBay" it would need (a) a case, (b) a
> > power supply, (c) input and/or output interfaces, (d) a display. And
> > if you're serious about your hobby then it will also need calibrating.
> > Once you add up all those, plus the time taken to design/assemble/test
> > I think you'll find you have significantly less money to spend on
> > synthesizers, and significantly less time to play them.
> >
> > Secondly, the half-baked. It's like holding up a steering wheel and
> > proclaiming "I has a car!". No, before this board could be even
> > remotely comparable to even a cheap old 20MHz dual trace scope from
> > eBay or somewhere like Stewarts it is going to need at least two input
> > pre-amps, and an external trigger input. Those pre-amps need to have
> > switchable gains, AC/DC coupling, very wide bandwidth (DC-20MHz or
> > so), over-voltage/ESD protection, and so on. And that's the easy bit.
> > Reliable triggering is what really makes or breaks a scope - Tektronix
> > got this nailed, HP spent years trying to get it right. Now, surely
> > the point of using a fancy doodad CPU is you can control it all in
> > software. So, off to RS or Farnell and buy some expensive Teledyne
> > signal switching relays to do the gain switching. And then put it all
> > together and make it work.
> >
> > Suppose you only want a signal generator? That's a bit simpler, since
> > you only need output drivers, able to cope with short circuits and
> > externally-applied volts (e.g., plugging the cable into a VCO output
> > or bench PSU...oopsie). Depending on what you intend to use the signal
> > for you might also need to investigate additional filtering to ensure
> > a clean output.
> >
> > Finally, the time-waster. Ok, so you got the board. Then you need to
> > put it in a case (you wouldn't want that bare board laying on a bench
> > with pliers, components, clipped leads, solder splashes, etc). Next
> > you'll need to add proper interfaces, as mentioned above. Then, as you
> > say, load and maybe modify applications to do what you want. The
> > power! Great. So instead of debugging the real problem (VCO not O'ing,
> > whatever) you're debugging your tools.
> >
> > Don't get me wrong - this has the potential to make a great component
> > within a piece of test equipment. But it is only a part, like a
> > steering wheel. On its own it is a long way even from cheap T&M kit
> > from eBay. Given enough time, effort, expertise and expense you could
> > make something comparable to what you can get on eBay. The project's
> > kickstarter page has a great sales pitch BTW!!!  *sheesh*
> >
> > Or put another way, what can I get for £378?? Recently sold on eBay (uk);
> > * Hameg dual-trace 20MHz scope, £50-£60
> > * Thandar TG-501 function generator, £60
> > * HP 3478A 5.5-digit bench multimeter, £90
> > And have plenty of cash left over, or maybe choose higher-spec
> > instruments (Tektronix scope, get the 3478A calibrated, etc)
> >
> > Beyond that and you're into higher end, such as:
> > * HP 3580A spectrum analyzer, £320
> > * HP 8903B audio analyzer, £214
> > * HP 8594E RF spectrum analyzer, £550
> > * HP 34401 6.5-digit bench multimeter, around £300
> > * HP 3325A synthesizer function generator, about £200
> > None of the above can be achieved by the redpitaya without significant
> > time and expense - in short, the above options are way cheaper if you
> > really need them.
> >
> > Personally I would rather spend my limited hobby time designing,
> > building, repairing, and playing synthesizers than futzing around with
> > building standard test equipment. YMMV of course
> >
> > -------
> >
> > Neil
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>

-- 
[s]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20151028/a613f1b5/attachment.htm>