My little IN-13 Spectrum Analyser

222 views
Skip to first unread message

Paul Parry

unread,
Feb 27, 2018, 4:32:00 PM2/27/18
to neonixie-l
Hi All,

Rather than another clock ( not that I don't like making clocks! ) I thought I would have a go at a Nixie Spectrum Analyser.
Here's what I came up with. 18 channels and all analogue electronics. No DSP or FFT. 
Probably more work than needed on the supports for the IN-13's but that's me :)


Cheers,
Paul


gregebert

unread,
Feb 27, 2018, 4:51:31 PM2/27/18
to neonixie-l
Neat!

I'm surprised that the response on the IN-13's is pretty fast; I just assumed they were sluggish.
How much current did it take to get maximum-scale ? I can see from the video you get very close to the end.

When I did some bench testing on the ones I bought a few years ago, static testing above 3/4 scale often got weird behavior, such as lighting from the opposite end.
Did you ever encounter that ?

threeneurons

unread,
Feb 27, 2018, 8:06:44 PM2/27/18
to neonixie-l
That's a beaut ! I have a current obsession with color organs. Going to make a couple with thyratrons. That plus neon, and my eyes definitely lit up !

IN-13's use 4.5mA max. x18 channels its 81mA, so its not really that bad.

What order filters ? Or trick it a little by upping the Q ?

Nick

unread,
Feb 27, 2018, 8:26:20 PM2/27/18
to neonixie-l
Lovely as always, Paul.

Are you using a multiplexed switched capacitor filter like the ltc1059, or lots of distinct channels?

I only ask as I looked at something similar a couple of years ago - the traditional way was to use a whole bunch of individual discrete analogue bandpass filters. The approach I was looking at used a single switched capacitor filter that was analogue multiplexed across the bands I was interested in both in terms of the bandpass frequency and the output sample and hold. Used about 1/10th of the components.

Cheers

Nick

John Smout

unread,
Feb 28, 2018, 12:37:08 AM2/28/18
to neoni...@googlegroups.com

> On 28 Feb 2018, at 01:26, Nick <gerbilp...@gmail.com> wrote:
>
> Lovely as always, Paul.
>
> Are you using a multiplexed switched capacitor filter like the ltc1059, or lots of distinct channels?

It’s lots of distinct channels - I laid out the PCB :)

John S

Nick

unread,
Feb 28, 2018, 1:08:32 AM2/28/18
to neonixie-l
Top man as always, John!

Nice job, though - it looks quite beautiful. I'm wondering what folks' experience is with the lifespan of IN-13s?

Cheers

Nick

Paul Parry

unread,
Feb 28, 2018, 3:21:07 AM2/28/18
to neonixie-l
Hi Nick,

I was quite surprised by the lifespan on the IN-13 tube.. someone did a forensic laboratory analysis https://www.saltechips.com/lab/the-shine-experiment/ and found if the tube was on continuously, it would take 14.7 years to reduce in brightness by 30% from new. 
The Issue with the IN-13 is they poison very quickly, if they display a static level for a week or so it is enough to poison them. However on the spectrum analyser they will never stay at a fixed point, so I'm not thinking that the lifespan is going to be much of an issue.

Cheers,
Paul

Paul Parry

unread,
Feb 28, 2018, 3:27:45 AM2/28/18
to neonixie-l
Thank you, they take about 4mA for FSD, however they rarely need to go right to the top - a little reservoir capacitor helps.
The IN-13 has an extra lead, and has I guess like a pilot light at the bottom which helps the glow start at the right end. I've had a couple where the glow broke up, or did not reach the top of the tube, however you can burn them in just powered up for a couple of hours to fix that.

Paul Parry

unread,
Feb 28, 2018, 3:29:15 AM2/28/18
to neonixie-l
No, current pull is not bad really, 125v and they are not all going to be lit all the time.
The filters are Fliege filters, seem to do the job ok.

Jon D.

unread,
Feb 28, 2018, 8:05:32 AM2/28/18
to neonixie-l
I haven't done much of anything with filters, but isn't a Fliege filter a notch filter.  How did you turn it into a bandpass filter?  Apologies if this is a dumb question.

Paul Parry

unread,
Mar 1, 2018, 3:39:10 AM3/1/18
to neonixie-l
It is indeed a notch filter, all that happens is that I get a DC level out of the filter, which then feeds into a comparator chain, which effectively inverts it. 

Jon D.

unread,
Mar 1, 2018, 2:18:23 PM3/1/18
to neonixie-l
I like your description.  Inverting a DC level makes sense.  Then you have to have the inverted notch voltage drive your IN-13 to the max at the center frequency of the notch filter.  I suppose I was thinking too much about the frequency bandwidth of the bandpass.  18 somewhat narrow notches for the 18 audio channels up to 15KHz is probably a decent approximation for the audio range of frequencies.  The video looks great !!

Thanks.

Robert L

unread,
Mar 3, 2018, 8:37:14 PM3/3/18
to neonixie-l

Hi Paul,
 
I've lost count of how many times I've viewed your video...

A very nice job indeed!

Best,
Bob L. 

 
Reply all
Reply to author
Forward
0 new messages