20ledx2 dualcolor bargraph...use and enjoy!

Hi to all, here is usefull thing...
This is 2 rows of 20 leds, each row can display
output level as green bargraph and gain reduction
as red point at the same (more or less time.
Possible to use as peak/rms indicator too.
It uses minimum parts and scale can be configured
(lin/log3db/vu/combined) by replacing 3914 to 3915/16.
See schemo and comp.placement.
Have to replace battery on camera to take photos.

http://img.photobucket.com/albums/v436/IJR/20ledx2bargraphsch.jpg

http://img.photobucket.com/albums/v436/IJR/20ledx2bargraphpcb.jpg

Hi Igor,

What supply rails are you running into the 4052's? It looks like +12V only on the signal switcher. Won't this cause a problem with negative excursions of your signal inputs without Vee?

OK, I answered my own question :grin:
The signal inputs are already rectified before entering the signal switch.

Photos:

http://img.photobucket.com/albums/v436/IJR/PIC_0222.jpg

http://img.photobucket.com/albums/v436/IJR/PIC_0223.jpg

http://img.photobucket.com/albums/v436/IJR/PIC_0224.jpg

I've got an error when I clic the links :sad:

All five links work fine from here.

it look very useful Igor.
I don have a lot of experience with this kind of circuits, can a circuit like this make some noise on the signal? the multiplexer is changin the load isn`t? can a buffer solve the problem?

Rafael

The signals are already "prepared" and buffered,
I used this on my own SSLMixBuzz ripoff,
and _IF_ all grounds are OK, it does not adding ANY noise.
BTW, got s/n on ssl ripoff about -88 db with this indicator
on test variant built on one 2181c in VCA, and
I going to use 4x That 2181 "a" grade in parallel for each VCA in future.
As far as I know Gyraf's GSSL ripoff measured from some people -71 db without any additional noise-adding stuff ))

:shock: :shock: :shock: :shock:
woooaaaaa very cool work!!!


I was thinking about connecting it direct to the output to measure Vu, but soon I remembered the rectifier.

Rafael

One consideration with VCAs at least the older ones, there was a tradeoff between current density and distortion. In other words you could make it very quiet, or very low distortion just not both at the same time. Adding chips in parallel will help, similar to using larger geometry devices at higher current density.

I am a fan of using the Pk/VU meter with compression so you can see in real time how you are changing the dynamics (crest factor) as you compress a signal. For this to work well you want the Pk/VU scale to be calibrated in equal dB step size. Then the crest factor is visible as the gap between PK and VU reading. I find it easier to read as a peak dot, floating above a VU solid bar. This can even work with one color since PK will always be larger than VU if you dial in similar release times. The series of led driver chips in this design do have a 3dB/step model available.

The second color could be used as a gain reduction display, but you would have to play with the control voltage to get that to track on dB meter scale, since VCA control voltages are usually linear. Either log the audio for display, or anti-log the CV. I did something similar back in 80s and had the gain reduction meter move from top (0dB) down (-xdB per LED), but the display was still busy. Easier to read IMO if GR is separate meter, but could be next to audio level meter.

JR

You could also multiplex a linear display chip and a log display chip to the same set of LEDs to get your peak VU display along with gain reduction.

Yes and you could probably do the whole thing even cheaper with one small micro processor.

I'm an old analog dog, but some things just scream, do me digital. The output is very digital after all.

JR

One consideration with VCAs at least the older ones, there was a tradeoff between current density and distortion. In other words you could make it very quiet, or very low distortion just not both at the same time. Adding chips in parallel will help, similar to using larger geometry devices at higher current density.

Click to expand...

the alan smart c2 achieves a s/n of -100db (I measured this myself) it is based on a completely symmetric signal path with a vca for each phase of the input signal.
the thd was also very low (I measured 0,005%) even with the c-grade vcas

CV signal is lin/db, and cheap solution is to use 4301 in sidechain, it has inside good e'nuff vca for sidechain, 3 opamps and kinda rms detector
(not very precision imho) to convert output level
into lin/db function...money' and space saving solution.

_achieves_
liked this "achieves"!
put here input signal 0db (ref @+4) make-up out +14 you will get -102s/n at least.
Put 4x2181 in parallel, it will be around -106...-110. Nuff new.
BTW, I hate symmetric stuff.
Take ssm2142 for xampl feed-in 1k square signal, then sum outs thru
10k 0.1% than see whaappens.
Music is never sin-like signal.
I allways use SE in my design opamps, unbalanced(pin 3 to gnd thru same Z as on pin2) or trafocoupled outs on my gear. Sounds better.
Dats it.

I would say everything around -100 is good enough...

but as far as i understand you get a better s/n by not unbalancing the input signal (symmetric design) or just use an unbalanced input...
do you know why?

but as far as i understand you get a better s/n by not unbalancing the input signal (symmetric design) or just use an unbalanced input...
do you know why?

Click to expand...

No, sorry, Matthias, thought I expressed myself right....
OK, electronically balanced input is good thing, it helps us to re-solve some gnd issues.
At the point we trying to split signal like electronically balanced
output, it have to be done with high precision, and anyway,
when you summing back + and - phases, on square signal,
it does not makes better transient responce.
This is what I meant.
And, just my own preference, I can spend
30-80 ma of power supply more per discrete stage but make the output
in class A and single-ended. I.e. instead of like 990's output
I use CFP with current source out on Albatross (my own design amp.stage),
input folded cascode fet+bipolar sounds with this just EXCELLENT.
I mentioned them in tread about my own console, IIRC posted
some photos with 1k load 0.002%thd and heavy loaded with about 0.08% thd
2nd about 30 db higher than third.
Never got same results and sound on tubes, BTW...
Sorry, got out of topic ))))

Igor,

when can we see more details on your Albatross project, please? 2nd 30dB higher than 3rd sounds heavenly!

Martin

Ahem))
Buy preamp with Albatross stages, scrap down compound, etc ))
I will like to keep this in secret some time, later will post info before someone will scrap compound from modules.
Anyway, all ideas you can found in Internet.
All I can share with people at this time, I do.

Fair enough, thanks for the answer!

Are you saying your Albatross pre-amp is available commercially? Is there a website? How much?

Best,

Martin

Hi Martin, for now, I have enough money from other stuff, and going into commercial manufacturing of an audio units is too much headache for me at this time, but... thinking about this as well.