Vishay PBG Plasma VU-meter...photos

[Igor]

Hi to all!

Some time ago, I found this on net:
http://www.jogis-roehrenbude.de/Leserbriefe/Glimmroehre-Stereo-Anzeige/Bargraph.htm

The appearance of these cutie plasma's is really cool,
but level detectors were too simple at my bad taste, I wanted
more precision response and peak/average option.

Some uP's were burned, and baby was breadboarded...

Then, after some freakin', level detector and dc-dc converter were replaced...

Here's some photos of first proto' and schemes.

http://i251.photobucket.com/albums/gg291/diy33609/pbg_disp_sch3.gif
http://i251.photobucket.com/albums/gg291/diy33609/pbg_disp_sch2.gif
http://i251.photobucket.com/albums/gg291/diy33609/pbg_disp_sch1.gif


Gigathanx to German guy for providing source code!!!

 

[nrgrecording]

Whaoo!!! :thumb: :thumb: :thumb: :thumb:
I have the same Vishay PBG12201 bargraphs at home but I didn't liked the input circuit, too. A half precision input circuit for a high precision is not that good, eh?

I also tried the nixie tube bargainmeter... ahem bargraphmeter. :wink:
http://www.nrgrecording.de/temporary/RTW3.jpg
http://www.nrgrecording.de/temporary/RTW1.jpg

I thougt about doing a meterbridge with these plasmas. But its hard to see 24 "over-levels" at the same time when the bargraphs have the same colour under and above 0dB. Red plexiglas for the levels blow 0dB could probably help. The levels above 0dB (without the plexiglass or with clear plexiglass) will look brighter then.
By the way... where is the PIC and the transistors on your pcb? :?: Is there another sub-pcb?

Frank.

 

[EZ81]

@Igor: Looks nice and compact. :thumb:


[quote author="nrgrecording"]But its hard to see 24 "over-levels" at the same time when the bargraphs have the same colour under and above 0dB.[/quote]
Hi Frank.
You can bright up sections of the display by slowing down the motion of the light dot for certain sections of the display (the bar is actually a dot that moves over the display 70 times a second). IIRC SSL did this for all dots after ? dBu (with a bucket full of logic ICs etc), but one can also add marks in ten dB steps. etc. like this:

It does not look great on the photo, but is actually quite visible in real life. My plasma meter never left the breadboard stage, i got stuck at the signal rectifier, too....
EDIT: fixed some typos

 

[Igor]

Hi Frank!

I seen this in-13 circuit on net, the response doesn't seems nice too...
it uses 2252 as rms detector, right?

Some time ago, I experimented with IN-13's, and finnished with a lot of 074's and tone of matched transistors in log detector,
then Dorrough time constants, see:
http://www.groupdiy.com/index.php?topic=22456

btw, in-13's are quite sensitive to their B+ voltage, 1% deviation can give
1db mistake....

I had no 2252's handy, but now Farnell have them in stock.
Good idea to use cheap in-13's and 2252 then avg/peak
time constants, like this:

http://i251.photobucket.com/albums/gg291/diy33609/in13_log.gif

Whole circuit becomes quite simple...

Problem is for perfect stereo tracking, 2252 is not best precision,
and in-13's are a way sloooppy...

circuit should have calibration for in-13's and 2252's, then,
cal procedure becomes bit complete... but whole vu-meter cost
on in-13 is cheap.

By the way... where is the PIC and the transistors on your pcb? Is there another sub-pcb?

Backside photo:
http://i251.photobucket.com/albums/gg291/diy33609/pbg_2.jpg

@ez-81: did you used microcontroller in your project?

 

[ClaudeTucker]

Igor, nice work!

You may have tried it but you can put a 22M Rfb around the 2252 to improve the low-level accuracy below ~-50dBu.

See: http://www.thatcorp.com/datashts/2252data.pdf#page=6

There's a pretty convincing graph with and without the Rfb.

Claudelle

 

[Igor]

Claudelle, I know about 22 meg, but even with it 2252's needs
adjustment. AD636 etc more precision, but more costy.

 

[cannikin]

Nice work Igor, once you get the kinks out - fab the boards, I buy some. I would love to have a bunch of those puppies dancing in my studio.