LA-4 Help Thread!

[Rybow]

In my first channel I had the NSL 6910 + LED configuration. I tried the 20K resistor in parallel with no luck, but increasing the trimmer to 1K gave me a bit more control. Not enough to get the ratio adjusted properly, but it did make a difference.

I did finally get a chance to try out a different LED and the MC33079. I used a small 50 degree 800 mcd led, and I noticed right away that I had more control over my threshold. I could actually dial in moderate amounts of compression where as before, it would slam everything. That was when I had a 10mm 35 degree 2700mcd LED in there. I am not sure if it was the LED's or the MC33079, but the sound is much more transparent which I love. It used to do something odd in the low mids which just added some mud. Distortion is less as well. On the downside, I cooked something with all the components I have been changing out. With the input sensitivity switch in high mode, everything sounds OK, flip it into low, and my head almost got taken off by the white noise. I can turn the noise up with the output control, and it reacts to audio in that it seems to distort the audio. For now, I am pulling this channel out. Will probably order a new PCB and remake it as I have no idea where to begin searching out the issue. It was so quiet before I started messing around in there! Oh well, live and learn.

My second channel should be going in (finally) soon. I am using the NSL 32 in this one, and am pairing it with the cinemag CMOQ 2S OT. I did a way better job on this board, so I am looking forward to hearing how it sounds!

 

[Luny Tune]

I'm very sure that it's the LED that determines whether or not you need to change the trimmer or the 82K resistor (R13). My suspicion was correct in that the 2700 MCD LED would hammer the signal down. (I gather) That's because the light is so strong as soon as the LED starts to light up that it brings the resistance all the way down to minimum instantly in the LDR. What we need is a softer light that moves the point where the LDR works to a different range that, due to the LDR's logarithmic curve, also gives us different attack and release times.

If you can't find LEDs at 100-200 MCD I'll be happy to send you some of the big 150 MCD ones I have. Just say the word.

I still have the bigger ratio trimmers in but I'm not sure if it's actually necessary. I have only just ordered more 6910's and will experiment with even softer LEDs and the specified trimmer when they get here.

 

[pedroplanet]

today I could calibrate the correct ratio with nsl32, it distorts more than it should for sure,
it maybe "attack/release" times on those I got.
By the way I'm using the MC33079 on sidechain..

 

[Rybow]

I'm very sure that it's the LED that determines whether or not you need to change the trimmer or the 82K resistor (R13). My suspicion was correct in that the 2700 MCD LED would hammer the signal down. (I gather) That's because the light is so strong as soon as the LED starts to light up that it brings the resistance all the way down to minimum instantly in the LDR. What we need is a softer light that moves the point where the LDR works to a different range that, due to the LDR's logarithmic curve, also gives us different attack and release times.

If you can't find LEDs at 100-200 MCD I'll be happy to send you some of the big 150 MCD ones I have. Just say the word.

I still have the bigger ratio trimmers in but I'm not sure if it's actually necessary. I have only just ordered more 6910's and will experiment with even softer LEDs and the specified trimmer when they get here.

Yeah, I didn't even look at the rating when I bought them. Anything under 800 MCD at the 50 degree angle has been very hard to find. Mouser has some, but they are all green. That would be amazing if you could send me a couple to try out! Although, I am on the other side of the world.

Allied was not great for me. I ordered through them when I started this project just to get my hands on the NSL 6910's. Cost me $15.00 CAD for shipping. My order was small, so it wasn't a space thing. It also shipped ground and took over 2 weeks to get to me. Digikey by comparison is $8.00 shipping next day delivery.

 

[Luny Tune]

Well, it looks like Allied is not particularly more expensive than others. I can certainly live with $15 in shipping costs. Stupid that I have to get that kind of small indication of it from someone in here. How can they not know when you know??

Well, thanks. Looks like I don't have to fear a $-150 surprise on my account if I order from there some time...

 

[Rybow]

I know. I had to email them directly to find out what company they used to ship stuff. Its not one of the standards (fed ex, ups). Some other company. I find that $15 for shipping a bit steep when it takes 2 or more weeks to get to you. Especially when the competition has better rates and/or next day shipping. $150 is pretty steep considering that shipping from Canada to Europe for a small parcel (1kg 30x30x5cm) is around $60 or so for delivery within 4 business days, half that for regular air. I have to place an order with Allied soon for some more 6910's, if you need some stuff, I'll let you know when I am about to order, and I can just tag it on my order and post it to you.

 

[Rybow]

Channel one out, channel two in!
I finally got my second channel in. Its not dual mono yet due to some issues with channel one, but channel 2 is in and working! Thought I'd do a little report. In channel 2 I have the NSL 32 loaded with the MC33079 in the side chain. I used all vishay dale 1/2 watt resistors, Wima and panasonic caps, hairball meter and grayhill rotary switches in this build. I did a very quick adjustment procedure on it, and then tested it out with a DI guitar.

I have ratios! Unlike my last channel, all the ratios are there, and how! Major sound differences between them all. Just like I'd expect from a compressor.

Distortion is very low, even at 20:1 while taking off around 10db's of GR, I could not detect a whole lot of distortion at this point. I know a lot of other builders have had issues with distortion using the NSL 32, but on my guitar signal, I could not hear any overt signs of distortion unless I really smashed the signal. Even then it wasn't that bad. Maybe I just got lucky? On my other channel, I would get a crackling distortion as soon as the threshold was reached. Not here though!

Noise- noise floor seems pretty good. If I crank the output on my LA 4, I do get some noise, but it does seem to be very low. I'll try and measure it more accurately soon. I still have some wiring to tidy up.

Common mode rejection- I am still a little confused about this one. I am using my DAW to send the tones out for adjustments, but I am wondering about the best way to make this adjustment.

The overall sound is very clean and transparent, very nice. I can tell the NSL 32 is quite a bit snappier then the 6910. I really like it so far, but we shall see after I try it out on some more sources. I still want to try out the VTL 5C4 as well.

 

[ilfungo]

I just finished my third LA4
In this  I used urei 11148 transformer
beautiful sound and 10 dB of extra gain ...
by hairballaudio are reproductions of this transformer, which could perhaps be a good option to OEP ....
I still think the desire to build one with  input transformer, is there a way to baypass entire input section to insert input transformer?
Thanks

 

[Rybow]

Did you have to do anything special to hook up the UREI transformer? I am about to get a cinemag to try out. Unless it gets lost in the mail that is.

 

[ilfungo]

I have not had time to directly compare  UREI and OEP transformer
but the feeling was very great sound
10db extra gain
and seemed less noise ...
these days I will test and comparisons and will let you know
I would like to post my LA4 Wav files but do not know how ???

 

[pedroplanet]

I used all vishay dale 1/2 watt resistors, Wima and panasonic caps, hairball meter and grayhill rotary switches in this build. I did a very quick adjustment procedure on it, and then tested it out with a DI guitar.

I have ratios! Unlike my last channel, all the ratios are there, and how! Major sound differences between them all. Just like I'd expect from a compressor.

Distortion is very low, even at 20:1 while taking off around 10db's of GR, I could not detect a whole lot of distortion at this point. I know a lot of other builders have had issues with distortion using the NSL 32, but on my guitar signal, I could not hear any overt signs of distortion unless I really smashed the signal. Even then it wasn't that bad. Maybe I just got lucky? On my other channel, I would get a crackling distortion as soon as the threshold was reached. Not here though!

I made my unit using 1/2W resistors only on 15ohm and 47ohm...(R43 and R45 positions)

May the distortion issue be the rest 1/4W resistors??

 

[Rybow]

I made my unit using 1/2W resistors only on 15ohm and 47ohm...(R43 and R45 positions)

May the distortion issue be the rest 1/4W resistors??
[/quote]

Might be the issue. There are a couple of other posts in this thread relating to replacing 1/4 watt resistors with 1/2 watt resistors to reduce distortion. I know on my first channel I did the same thing you did using only 1/2 watt resistors for the 15 and 47 ohm values. I ended up replacing them all except the 4.99K (couldn't find them at the time) with 1/2 watts in the end and the distortion did get better. Although it still crackled when it hit the threshold. I found that taking a long time to really dial in the adjustments helps out a lot as well.

 

[pedroplanet]

I've just got the vactec VTL5C4/2. distortion problems gone! 

 

[Rybow]

Nice man! I am definitely going to give these a try. Is there a big difference in attack and release with the Vactecs?

 

[ilfungo]

Opto Matching??? ???

applying various resistance to the vactrol led
I measured several VTL5C4 / 2
I found only 2 Vactrol who behaved in a similar way (differences of a few ohms)
1 - 660ohm
2 - 665ohm

is this the matching way?
Thanks

 

[pedroplanet]

Nice man! I am definitely going to give these a try. Is there a big difference in attack and release with the Vactecs?

I can't realy tell because my nsl32 were almost useless...

It seems to have a slower release time

 

[ilfungo]

After building three LA4 and be very happy with the sound (Luny THANKS for your great project)
I want to build a rack with 8 x LA4 to use live
How many amps should be the toroidal?
Can this work?
http://www.antekinc.com/details.php?p=62
Thanks

 

[regularjohn]

So has anyone had a chance to use their LA4's as a drum buss comp?
I'm asking because I'm trying to decide what opto's to use.  I generally like a slow attack and fast release for drum buss applications, and I know Luny said the NSL32 has a fast release, but is it disturbingly fast in a way that would seem unnatural?
Also, I just read through all 24 pages of this thread  , and unless I lost consciousness at some point I don't think anyone has made much mention of the attack characteristics of the different optos.  Any thoughts on that?

 

[benlindell]

I've tried mine as drum bus compression a couple times now, release time is not very fast IMO, It kind of seems to chill the drums out and adds some tone and distortion when pushed hard, for me it it didn't seem to add punch like an 1176 or a distressor can, just color. Be digging it on vocals, controls them nice. I haven't played around with different optos but it did change out the side chain opamp like Luny said and it cleaned up the sound a touch.

I'm enjoying using it, thanks again for making this project happen luny.

 

[Luny Tune]

It's all labour of love, guys. You're most welcome. ;D

Glad to hear you agree that the other op-amp in the sidechain has a good effect. I'm constantly questioning myself, after having been wrong about a few things in this process, but that particular thing stuck with me.

As a drum bus comp I think it's all about what you're looking for exactly. The 1176 is more controlling and keeps the signal in an iron grip while maintaing/adding punch and you can certainly shape the sound in a number of creative ways. The LA-4 just does what it does but what I like about it is that it's a bit more...bouncy. It's strikes me as maintaining a certain musical elasticity. I love that just as much as I love what the 1176 does.

Remember that I'm using the LED/6910 for opto.

There are different nsl-32 versions, by the way. Maybe some will behave even better/more like the original than the basic one. The snag is, I believe, that when the LED in the opto is very bright we're moving the working range to the steep part of the LDR's range. That kind of moves against that we usually want the biggest dynamic range out of an opto, but we probably just don't need such a big dynamic range for it to work fine musically. Think about it; The VTL5C9 has a 112 db dynamic range....Who has ever squeezed a signal by that much?? That's WELL above what could be reproduced on a CD! (16 bit=96dB dynamic range) The thing is that with a comp/limiter with no options for setting or adjusting attack and release times you do need those times to be set somewhere and that's where the right combination of LED and LDR comes in. It would be very interesting to see if there's a encapsulated opto out there, but most of the time, I can't find specs about the LEDs in them.

Now, there are still lots of stuff I don't understand in this circuit but I think I'm right about this. The amount of gain reduction we need in a comp/limiter is easily done with just a third of the VTL5C9's (just as an example) range, so if a given LED can only drive the lower third of the LDR's range, mission accomplished. We don't just need dynamic range. We need the right behaviour.

All that said, DIY rocks and we are free to put any opto we like in there! It's just my personal obsession with the original that's pulling me in that one direction all the time....