la-light build

[widdly]

I built an la-light compressor following johan's schematic in the gallery ( http://www.aronnelson.com/gallery/v/johan_0/LA_Light_last_one_001.gif.html )

I'm running on a mixer insert and it compress's properly and sounds nice at mild settings.  Thanks Johan for the elegant design.

I have one problem though....

When I set the threshold so there is a lot of compression happening, I get some distortion from loud signals.  I sounds like it is in the low frequencies.  If I change the threshold (reduce the compression) the distortion disappears.  I'm using two led's and two ldr's all sandwiched together. I'm guessing the distortion is caused by my LDR/LED setup.  Any ideas if this is normal or a way to reduce this effect? 

I bought a 50uA meter to use and it wobbles away when it's compressing. If I want a VU meter I'll have to print one up and glue it on the inside.  Is there a way to calibrate the meter so the scale is meaningful?  Using a standard VU, the +ve dB settings would be useless since the circuit will never increase the gain, only reduce it.

[darron]

sounds like there possibly just isn't enough headroom for you. it might be fine with less compression, but as you start to crank that up you hear it come through. bass is what you often start to hear break up first as you reach headroom limitations. maybe try increasing the voltages to the opamps to a few volts before their full potential. don't do this for the LED's opamp though as it doesn't need it and you can potentially burn the leds out if they expect the opamps 9volt limitation. the LEDs and LDR are only there to set gain level for the opamps, so the noise shouldn't originate from them directly.

typical guitar voltages seem to be around a volt or so maybe at first moments of pluck? dropping down to hundreds of millivolts. i don't have a scope so i haven't had a great look into it. if you are reaching headroom breakup levels then a range that goes a bit past 9 volts might be for you.

what voltage range are you giving the pedal?

anyone else got a different theory? please correct me if i'm being misguiding, or jump in and agree.

[widdly]

That's what I thought at first.  I don't think it's a headroom problem because the distortion occurs when it is compressing the most.  If I turn the threshold up, so there is less compression, the signal is louder and the distortion disappears.   I'm running at 15volts so that gives +/- 7.5 volts. 

I had a look at some other similar circuits and found a recommendation for placing a 100uF cap in parallel with the LED's to get rid of bass distortion ( http://sound.westhost.com/project92.htm ).  The author suggests it is caused by the LDR itself.  Sounds like the problem I'm having.

[Johan]

did you set up the LEDs so they both shine on the LDR?. depending on the LDR, if only one LED shine on it, it might get to recover enough on low frequencys so that you get something that would look like halfvawe rectifying. the trick to this circuit is that you get the bennefits of fullwave circuit, but simplyfied greatly  by using two LEDs ( each LED will only shine half wave..)  obviously, the timeconstants depend on the LDRs
with the trimpot for the meter, if you set it so the meter reads 0db, it should give you a pretty good tracking of the gainreduction

johan

[widdly]

I'm using two LED's and two LDR's all pointing at each other.  I redid the LDR/LED part last night to try a closer lineup between the LDR's and LED's and that seems to have sorted the distortion. 

The meter is working fine and sits at full scale when not compressing and goes down to about 1/4 scale when really compressing hard.  I was thinking of running an opamp buffer on the output so I could switch the meter between showing compression amount and output level.  I'll see how that goes.

Thanks again for the design Johan.  I'm really happy with this one now.  I'll post some pics and samples once I get it boxed up.