Just built an LA-4. Anyone got an original one I can compare with??

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Luny Tune

Well-known member
Joined
Sep 11, 2005
Messages
414
Location
Elsinore, Denmark
My buddy and I just made PCB layout, milled it on his PCB prototype face mill thingy (WHAT a piece of equipment!!! :p) and built it.

It works just fine. (I'm making a complete project out of it so everybody can get access to it of course.)

However, I don't have the original to compare it to. Well it's built exactly like the schematic says (changed the opamps to TL074 though...) but I think it's slightly slower working than I remember it. I worked on one some years ago and loved it but I remember it a bit different sounding than this diy'ed one.

It could be due to my using a VTL5C4 instead of the LED/LDR combo they used originally and I've got the LEDs and am expecting the LDRs any day now and I'll experiment a bit. The romantic nerd in me would like to build his own optos and I have a pretty good idea what was in the old original optos...

I'm writing here in the hope that if someone out here has got an original LA-4 (We built ours from a '79 schematic) they would maybe be so kind and provide me with a test file??

I have a short file with an electric bass that I use for my crude tests and I would so appreciate if I could send it to someone who would run it through their original LA-4 and send the resulting file back to me.

The file has som pauses and some noise on it and turned out to be quite usable since the slow release time is actually audible when compressing hard and the bass wasn't compressed when it was tracked so it's also quite well suited dynamically.

Anyone who will help?

Find the raw file here!

Set the unit something like this.
Keep the input safely under clip level. A couple of dB margin...
Threshold - Middle
Ratio - 8:1
 
We have two, a blackface and a silverface, don´t know which year they´re built. I can run your file through them monday when I´m back at work. The threshold controls are a bit fiddly to work with, how much(-ish) do you want me to compress?

/Dave
 
Great!! Thanks a million!

Hmm...right now I don't have a meter on it...

This is what it looks like at the moment:
la4test.jpg


...but if you could maybe make one in 8:1 with 6-8 dB gr and one in 20:1 with the threshold all the way down to give me a sort of look into an extreme setting...

That would be awesome!

Thanks!! ;D
 
Luny Tune can I ask wich LDR you have order ? thanks I been experimenting with a home made T4b myself

many thanks .............Gary O.
 
But there's no T4b in the LA-4. The LDR in the T4b is a Silonex NSL5910 (though it needs to be tested and selected for it's response times to be within specs for the LA-2).

The 5910 has been replaced by the 5912 now as far as I can tell. The difference seems to be the removal of the glass lense.

The 5910 is most likely chosen for it's response to the wavelength of the light emitted by the EL-panel. It would be silly to use an LDR that responds poorly, or not at all, to the light that's meant to make it all happen. That's why I believe the only logical choice for LDR in the LA-4 opto is one that responds to the light emitted by the LED which I know is red, and that's the NSL6910.

I'll post it when I can confirm which LDR it is.
 
Thanks for info sorry thought t4b was in LA4 ...yeah done lots of light source & and LDR experiments still trying to find best LDR tried couple NSL5910 but no joy works but ..... little neon bulbs work well for AC light source by the way.
 
Remember the lightsource for 5910 has to be the right color. Maybe you tried it with the wrong color? Should be some sort of green-ish color i believe. There are others out here who knows. I know that the right EL panel can be bought at Farnell. Maybe it's this one:
http://dk.farnell.com/pacel/50x60mm-ylw-gr/el-sheet-50x60mm-yel-grn/dp/141926

EDIT: Just did a search on the forum and I can see you've already discussed this with guys much more experienced than me so I'm sure you know a lot more than I can tell you by now... ;D
 
No I dont no much....I have tried just about everything that lights up in every colour & every LDR I can get my hands on but still not found the right one if you know what i mean...I always like to hear other folks findings on the subject so many thanks
 
Ok, I got the sound down. The "buzz" isn't really there and that's on one hand giving me cleaner audio....but the "buzz" of the blackface one in particular is really awesome. The blackface is the one to aim for in terms of the "buzz".

I'm taking the consequence of that and will redesign the PCB to hold an option for putting in either TL074 or the old 4136 opamps and maybe even switch between 2 x 18V and 2 x 15V supply. I don't see that much else different in the designs of the two.

The Vactec VTL5C4 is waaay too slow for this comp. I'm "building" the optos myself and I've ordered som Silonex NSL32 to test. They have the same specs as VTL5C2 but they're only half the price. It sounds to me like the NSL32 will fit the bill nicely with it's half the attack time and one third the release time. I'm not absolutely sure if it's enough on the attack time but we'll see if it performs differently with home made "standard red LED/NSL6910" optos... I WILL nail it.

Thanks a million for the test files, Craptical! Thanks to you, we'll all soon have a well tested new project to dig into! ;D

One thing puzzles me though, Craptical...! How come the very squashed 20:1 file has even more distortion on it?? Can you explain? I don't see how the 4136 opamp should distort more with less signal?? Or is it also depending on the speed of the optos??? If the latter is the case then I understand, but if not... ???

I'll post the whole thing later. :)
 
Just played the file with the original blackface to the bassplayer performing on the test file. He absolutely loved it and wants one with 4136 chips in it.

Think twice before upgrading your units some day, craptical!! ;)
 
Farnell....or rather UPS didn't deliver today so the NSL32 vactrols won't be here until tomorrow.

By the way, if some og you haven't seen how a PCB face mill works (is face mill the right word at all for it? Or is it "cutter"?) I can show you a photo of the copper side after the contours have been milled/cut and most of the components are soldered in.

loddeside.jpg


It's a fantastic way to prototype and also test the gerber files before the PCB is multiplied into hundreds.

Really expensive though...
 
Yeah, I noticed that (distortion) too. Looking at the Waltzing Bear schematic (being no expert in op-amps) I can only think of three possibilities:
1. The output of IC1C gets severely loaded down as the LDR resistance decreases (despite 82K resistor???).
2. Internal distortion in LDR (wierd... maybe smoothing of SC signal gets bad/uneven at extreme setting?).
3. If IC2C has low input impedace, it loads the output of IC1D at lowest threshold setting (R14 fully open).
Clipping LED did not light up at any time. There was no difference in input signal level between settings/takes.

???

/Dave
 
Luny Tune said:
By the way, if some og you haven't seen how a PCB face mill works (is face mill the right word at all for it? Or is it "cutter"?) I can show you a photo of the copper side after the contours have been milled/cut and most of the components are soldered in.

It's a fantastic way to prototype and also test the gerber files before the PCB is multiplied into hundreds.

Really expensive though...

Simply it is a flat bed cnc router. I had two ISEL (German) machines. They are also marketed as Tecno ISEL in US. Small desktop DaVinci does not cost much.  Galaad software provides PCB routing and costs only a couple of hundred euros.
 
Yeah, and I thought it was the much talked about 4136 issue that I could hear. Well, it's not!

And the VTL5C4/2 isn't too slow either as I claimed a couple of posts ago. Sorry for any confusion. It's a question of my not being able to understand how it all works yet.

How do I know? Well I tested the NSL32 and though it's about three times faster than the other opto it still sounded exactly the same... I found the error I had made while stuffing the board and suddenly it was all good!!

Here are the test files:

Here's the original blackface LA-4.

Here's our LA-4 with Silonex NSL32 optos.

I am SO thrilled! I absolutely LOVE IT!! :p

And thanks to you, craptical, I know that it sounds as it should! Thanks SO much for your time! Can I send you a "Thank you-board" when I get some made?

@Sahib
Ok, thanks for the info. I think I'll keep an eye out for one of those maybe.
 
A board would be absolutely brilliant! The LA4 is my go-to comp for bass, mostly in 4:1 ratio, never too many of those... Good luck with your project!

:)

/Dave

 
It's a deal. When I got some money to put into it, I'll have a batch of boards made, and then you will get one! ;D

Just tried it on various things. Holy hell! I love it!

When I have used it in the past it has always been under time constraints in other studios so I haven't ever had the opportunity to play aound with it. It just sounded great on bass and vocals and was sooo easy to dial in with great results. I've wanted it bad for so long.

I really don't get all the dislike I sense towards it on various forums. (Here too...) Like it's not worth while the effort to build it yourself, it's not good quality enough or something...

I think it absolutely rocks and in my opinion it's a perfect companion for my other weapons off mass compession! :p
 
Uh yeah, a fellow fan!! I'm totally with you...

I'll certainly post it out here when the boards are ready. :)

PS: Now that you got the originals, do you know of any adjustment procedure to follow? I have a couple of ideas but it would be cool to know how they did it originally. Like it says in the 1176 manual. It's just not there in the LA-4 manuals that I've seen...
 
Back
Top