bigugly

a one knob sqeezer
« on: December 06, 2010, 04:28:10 PM »
Hey all,

I've been needing to tame 4 channels of vox for a small PA rig that I use with my brothers band and this seemed interesting.- http://silonex.com/audiohm/compressor.html - see fig. 4  I already have a dozen of their NSL-32SR3 optocouplers on hand for use in my guitar amps. This little comp should fit the bill.

I added the current mirror from fig. 6 and a full wave rectifier that feeds the Attack/Release RC network. RV1 will be panel mounted and labeled "Squeeze". RV2 and RV3 will be trim pots which will be adjusted during testing, although they could be brought to the front panel also. I couldn't find any info on subs for the 2SC3068 but it doesn't appear to be anything special so I'll be using BC550Cs since I have a large stock of them. Any ideas or suggestions would be much appreciated.

thanks,
James
"...the land of the free*, and the home of the brave"
*some restrictions may apply, see Congress for details


abbey road d enfer

Re: a one knob sqeezer
« Reply #1 on: December 06, 2010, 06:16:36 PM »
The problem I see with this is that the amount of compression is governed by the gain of the circuit; basically, increasing compression relies on increasing gain. I think it is not desirable for vocal applications because of the risk of going into feedback when wanting more compression and having to reach the fader to compensate.
To make things clearer, do you intend to use that in insert in a mixer?
In order to avoid this problem, I would use the signal across the LDR, through a buffer.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

bigugly

Re: a one knob sqeezer
« Reply #2 on: December 06, 2010, 08:32:58 PM »
The problem I see with this is that the amount of compression is governed by the gain of the circuit; basically, increasing compression relies on increasing gain. I think it is not desirable for vocal applications because of the risk of going into feedback when wanting more compression and having to reach the fader to compensate.
To make things clearer, do you intend to use that in insert in a mixer?
In order to avoid this problem, I would use the signal across the LDR, through a buffer.

I thought having the make-up gain increase as the compression increases would be a good thing. I mean, on my dBx166 if I lower the threshold there by increasing the amount of compression I would inevitably have to increase my make-up gain to get the level back to the same apparent loudness. But, I can also see how having a "two knob squeezer" might give me added flexibility while still maintaining simplicity. Using your idea and taking the output from the top of the LDR is probably the prudent thing to do.
"...the land of the free*, and the home of the brave"
*some restrictions may apply, see Congress for details

abbey road d enfer

Re: a one knob sqeezer New
« Reply #3 on: December 07, 2010, 03:11:47 AM »
The problem I see with this is that the amount of compression is governed by the gain of the circuit; basically, increasing compression relies on increasing gain. I think it is not desirable for vocal applications because of the risk of going into feedback when wanting more compression and having to reach the fader to compensate.
To make things clearer, do you intend to use that in insert in a mixer?
In order to avoid this problem, I would use the signal across the LDR, through a buffer.
I thought having the make-up gain increase as the compression increases would be a good thing. I mean, on my dBx166 if I lower the threshold there by increasing the amount of compression I would inevitably have to increase my make-up gain to get the level back to the same apparent loudness. [/quote] The dbx163x did that, with one single fader. It was utterly unusable for live vocal applications.
Quote
But, I can also see how having a "two knob squeezer" might give me added flexibility while still maintaining simplicity. Using your idea and taking the output from the top of the LDR is probably the prudent thing to do.
I see that you've put a gain pot in the signal path. This can give you about 40dB gain. I don't see the point. There's no way you can increase the gain by 40dB and not have feedback. And you would need a reverse-log pot in order to achieve relatively smooth gain control.
« Last Edit: December 08, 2010, 02:49:28 AM by abbey road d enfer »
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

jdiamantis

Re: a one knob sqeezer
« Reply #4 on: December 07, 2010, 04:16:38 PM »
Hey all,

<snip>

I added the current mirror from fig. 6 and a full wave rectifier that feeds the Attack/Release RC network. RV1 will be panel mounted and labeled "Squeeze". RV2 and RV3 will be trim pots which will be adjusted during testing, although they could be brought to the front panel also...<snip> ...thanks,
James

James,

2 quick things.  1. The schematic shows a precision half-wave rectifier, not full wave. 2. Release time will be the parallel combination of R11 & the sum of RV2 and R4. To make release time just the combination of RV2 and R4, you need to move D1 outside of the feedback loop with the anode connected to the junction of RV2 & RV3 and the cathode connected to the junction of pin 7 of the OA, D2, C6 and R11.

Doing this will cause the threshold of limiting to be about 2 X Vbe or about 1.4 Volts. This may or may not cause a problem with overload, depending on input signal, and the position of RV1 and RV3. Best thing to do is breadboard one up and check it out.

jD
If it don't fit force it;
If it breaks, get another one.

jDiamantis

PRR

Re: a one knob sqeezer
« Reply #5 on: December 07, 2010, 09:14:52 PM »
> tame 4 channels of vox for a small PA rig that I use

There's two feedback loops.

Gain is ALREADY correct: 6dB to 2dB below acoustic howl. (There's no other way to work vocals with a band.) Your mixer already does this. Your box should be unity-gain except when levels are excessive.

There is also feedback through the performers. When your box reduces gain, they will automatically scream louder. A 1mS attack 1Sec release limiter will "duck", giving hoarse throats and "can't hear!" complaints.

In live performance, clipping is not illegal. It IS useful to clip some peaks to give a "loud" impression with a non-ample sound system or in concert with distorted instruments (guitars). OTOH it may be polite to not have long periods of gross clipping.

There's no very low frequency, no product spec sheet, no need to compromise for low-low bass THD numbers.

You want a moderately slow attatck (touch of clipping on attack) and fairly rapid release to restore loudness ASAP.

You don't have a $35/hour engineer to watch a meter; and what would it tell you? And meters break. IMHO you DO want a light to know when the box is mucking with gain, so you know when screaming louder is self-defeating.

K.I.S.S. always. LEDs can peak-detect. LDRs have time-constants.

Build the below.

Set up the band, set vocal gain just under howl, where you usually do.

Insert limiter. Start with pot full down, confirm no change in system operation.

Scream. Turn up pot until eye-candy LED flickers and amp/speaker distortion is reduced. Not eliminated: it won't do that, and a dead-clean result from a non-ample system will probably be "lame". Let the spit spatter, just hold the vowels down so they are vowel-sounding and not square-waves.


 

Related Topics

  Subject / Started by Replies Last post
1 Replies
1347 Views
Last post April 21, 2006, 01:50:18 AM
by Black Dog
4 Replies
1666 Views
Last post May 04, 2006, 12:30:41 PM
by buschfsu
14 Replies
7241 Views
Last post January 13, 2014, 09:58:01 AM
by Crash
3 Replies
1498 Views
Last post July 23, 2010, 01:57:14 PM
by mitsos