D-LA2A Support Thread

[dandeurloo]

I added a switchable 10db pad right after the output transformer. That way you can really push the output gain for some nice distortion without going over +4dB output.

What kind of Pad did you add?  I'm wondering which kind of pad will not effect the impedance and balancing.

Kingston and Hank, do you find 10db pad to be to much?  Would it be more useful if it was maybe 8db or something like that?  Just wondering what you have found over time in regards to this.

 

[johnnyscotch]

If you don't want a switched pad, would it just be a matter putting the applicable resistor(s) at the output XLR?

Have you discovered what combination of resistors gives which dB reduction? I suppose it's just ohms law, I'm absolute shite at figuring it out though

 

[dandeurloo]

I am talking about 2 different things that all happen at the outputs. 

1. output loading / termination  - I think this what you are talking about.  I have been doing that and like a 1.8k across pins 2&3. 

2. padding down the output - this is so you can drive the comp harder and get more color out of it and also hopefully get more resolution out of the makeup gain pot. 

I'm trying to figure out what kind of pad would not effect the loading when switched in and out. 

 

[Hank Dussen]

Since the output of the transformer is 600 ohm, I used a 600 ohm T attenuator.
You can use this pdf to find the correct resistors: http://www.groupdiy.com/index.php?topic=45632.msg571875#msg571875
So for a 10dB pad you'd need 2 x 620R, a 287R and a 1K33 resistor
I did 'reverse' my attenuator compared to the attenuator drawing in the link. The 'load-side' should be the transformer, and hot and cold on the XLR side. I'm not sure if this matters but it seems logic and worked for me...

I found 10dB of attenuation to fit perfect. No matter how hard I push the gain I never seem to get more then +4dB at the output.
But I guess this largely depends on the tube that's distorting.

Of course you can also use 600 ohm attenuator potentiometer. http://hairballaudio.com/shop/product_info.php?cPath=27&products_id=65
Or a rotary switch with different attenuator settings...

 

[Kingston]

I've a 600-ohm stepped attenuator for debugging. I found out a 12-15dB output pad gave me the strong effect-like distortion I was looking for. Others like different so don't take that value as a recommendation. Then I just made a 14dB 600-ohm H-pad using some calculator I found on the net. I switch it in and out with a DPDT relay.

Switching in and out that series 68k input resistor increses the gain/distortion pallette even further.

I wouldn't spend any brain cells or time with output loading selectors with 600-ohm output transformers. I've found they have little to no effect in any project I've ever done. Maybe some transformers are more picky, or maybe I've been lucky. No weird high frequency shifts in any of my projects from any output loading scenarios. But then I've mostly 600 ohm I/O everywhere.

 

[dandeurloo]

Great info guys!  Thanks for all the feedback.

I wouldn't spend any brain cells or time with output loading selectors with 600-ohm output transformers. I've found they have little to no effect in any project I've ever done. Maybe some transformers are more picky, or maybe I've been lucky. No weird high frequency shifts in any of my projects from any output loading scenarios. But then I've mostly 600 ohm I/O everywhere.

Kingston this is interesting.  I have found the loading makes really cool sonic flavors. I am using cinemag CM-9589L's and I am testing just going back into the DAW.  So that may have something to do with it. 

I also just swapped in a 6072/12AY7 and seems to have a little less gain.  May need to try the r109 & r113 mods and then wire in the pot again for feedback and hear whats happening. 

 

[Kingston]

Kingston this is interesting.  I have found the loading makes really cool sonic flavors. I am using cinemag CM-9589L's and I am testing just going back into the DAW.  So that may have something to do with it. 

CM-9589L is great. I've used in a few places too. The minor sonic differences I've found are:

1) about 1dB gain drop with 600ohm loading recommended by the datasheet with very flat response to +30khz.
2) 1-2dB high frequency roll off from maybe +10khz onwards with very light load (AD input for example)

experiences with edcor and sowter have been similar with negligible (in my opinion) differences and at some point I stopped caring. Mic input transformers are another thing, there I've used whatever they recommend in datasheets (if available).

From what I gather from experienced vintage transformer twiddlers here on the forum, some of them (input and output) can be very tweaky with different loads so I don't doubt if people are hearing significant changes with loading.

 

[dandeurloo]

Switching in and out that series 68k input resistor increses the gain/distortion pallette even further.

Do you mean R11?  I am playing with that now.  I can't decide if I want to put that on a pot or a switch with a fixed point.  Are you just switching it out completely?

 

[Kingston]

Switching in and out that series 68k input resistor increses the gain/distortion pallette even further.

Do you mean R11?  I am playing with that now.  I can't decide if I want to put that on a pot or a switch with a fixed point.  Are you just switching it out completely?

I mean R6/R106 depending on what schematic you are reading. It's an internal input pad. Bypass that, and you'll have the LA2A working as a "mic preamp" of sorts. More gain available. I wouldn't bother playing too much with the line amp stage at all (that R11 feedback resistor for example). It is what it is and it needs drastic changes to get anywhere closer to nicer sounding and softer clipping. The 12AY7 swap with the minor resistor changes often suggested is nearly pointless. It's still going to clip hard and nasty. No euphonic tube sound in LA2A's (less than 0.1%THD unless it's broken), just utilitarian clean line level gain. The points of interest (soundwise) are the transformers and the optical stage response.

 

[desol]

Switching in and out that series 68k input resistor increses the gain/distortion pallette even further.

Do you mean R11?  I am playing with that now.  I can't decide if I want to put that on a pot or a switch with a fixed point.  Are you just switching it out completely?

It is what it is and it needs drastic changes to get anywhere closer to nicer sounding and softer clipping. The 12AY7 swap with the minor resistor changes often suggested is nearly pointless.

I don't think CJ originally suggested the swap to make it sound better....even though he did say that. Pretty sure it was just about allowing more movement on the input....which it does.

 

[ethervalve]

hi all, I asked this earlier but it got buried really quickly:
Just wondering if anyone else had to use a different value for C4 on the left and right channel to get the frequency response equal?
A 100pf got me flat response on channel one, but i needed to use 250pf on second channel. Just wondering if this is normal... (Edcors on in/outputs)

 

[Kingston]

Pretty sure it was just about allowing more movement on the input....which it does.

The what now? Swapping the tube or changing its operating point has absolutely no effect on the compression. It has a minor effect on the first stages of the line amp. The amp is there only for the make up gain, and has no interaction with the compressor circuit whatsoever.

LA2A signal chain:

input transformer -> light/optical element based gain reduction -> very linear medium gain line amp for make up gain -> output transformer

Just wondering if anyone else had to use a different value for C4 on the left and right channel to get the frequency response equal?
A 100pf got me flat response on channel one, but i needed to use 250pf on second channel. Just wondering if this is normal... (Edcors on in/outputs)

This is normal. Some people use trimmer capacitors to perfectly match the channels, but it's not that critical really.

 

[desol]

Yes, you're absolutely right. In fact, i remember that it struck me as being odd the first time i used it? Compared with other units...

The tube swap doesn't do that much. It does allow a little more movement on the pot for adjusting makeup gain(which is otherwise at ~ 2) and maybe it sounds a smidge different. Although, i wouldn't know...cause i didn't bother comparing tubes. I think i also put the 25k pot in there with the 75k resistor. So that allows even a bit more...

Plus...12ay7 remains cheaper than 12ax7....which is a little bonus. 

 

[Hank Dussen]

I added a switchable 10db pad right after the output transformer. That way you can really push the output gain for some nice distortion without going over +4dB output.

FWIW, I checked, and it seems to be closer to a 12dB pad...

 

[dandeurloo]

Hank thanks! 

Yes, you're absolutely right. In fact, i remember that it struck me as being odd the first time i used it? Compared with other units...

The tube swap doesn't do that much. It does allow a little more movement on the pot for adjusting makeup gain(which is otherwise at ~ 2) and maybe it sounds a smidge different. Although, i wouldn't know...cause i didn't bother comparing tubes. I think i also put the 25k pot in there with the 75k resistor. So that allows even a bit more...

Plus...12ay7 remains cheaper than 12ax7....which is a little bonus. 

I've compared both the 12AY7 and the 12AX7 and for me I think I am going to use the 12AY7.  It does allow a little more control on the gain pot and I like the sound.  It is a little warmer, less crunchy.  Not great description but I hope that helps.  I am leaving R11 and the plate R's alone.  Those mods didn't do anything for me.  Plus then I can just switch in 12AX7's if I want to.

I also did more work on the 25k pot.  I found that by strapping a 27k resistor across pins 1 & 3 I got more resolution across the whole pot.  The center stays the same pretty much but the edges are much better.  It lowered the value of the pot to about 12k so you have to make up for that on the other resistors.  I also put a 220r on the ground side of the pot.  So it doesn't turn all the way off.  No need for that we have BYPASS! 

 

[Kingston]

For anyone trying to justify the tube swap with better gain pot control: please don't, there are better ways. Leave the line amp the ancients designed alone and don't ruin it's specs or the utilitarian "signature sound". It's not exactly the greatest tube line amp design in the history of mankind, but you wanted LA2A, right?

Let's concentrate on the gain pot itself. The main issue here is that the line amp has way too much gain for modern usage, like +20dB too much. You can knock this down with other tube choice with the cost of more THD due to unoptimal tube operating points and other issues, but really all you need to change/control is the gain pot.

The 100k value for the pot is uncritical by the way. anything between, say 50-250k works fine. This way you are also free to scale those resistors with much more tolerance.

My solution? 24-step rotary switch with 1-3dB steps from about -50dB to 0dB (relative to input). I think I even posted the exact values somewhere on the forum years back, maybe even on this thread.

 

[dandeurloo]

Kingston, isn't that kind of what I did to my gain pot?  I'm not 100% but it did sound better and was much more full range then a standard 100k pot. 

The other thing I am confused by.  Doesn't the overall value of the pot have to remain 100K.


Thanks for your help.  I am glad guys like yourself and others are willing to help!

NOTE:  I am interested in building a stepped switch like the one on this.
http://www.requisiteaudio.com/products/studio_electronics/l2m_limiter/index.html
I have a number of friends who have one and LOVE it.

 

[Kingston]

Doesn't the overall value of the pot have to remain 100K.

Not at all. This part of the circuit is "reasonably high impedance". It means the stage before the pot can only drive weak high impedance loads. It's not written into stars it must be 100k. Maybe the original designer picked that because there were gazillion 100k log pots around and thought it was a safe choice. But we can just as well pick 50k or 250k. Gain and frequency response will stay unaffected. Very basic electronics. And as you may at some point find out stepping out the painting by the numbers dighole, audio electronics is not ultra exact discipline. Instead we recognize the limits of certain points of the design and shoot in that general direction. The values are uncritical. Tubes are especially forgiving.

I would go out on a limb and say you can safely deviate 30% in any direction of all the parts of an LA2A unit and it would still function quite perfectly. Long a go, you couldn't even ask for better tolerance for many of the parts in the first place.

 

[johnnyscotch]

Can you gents help a brotha out with wiring the power trafo for this puppy?
I just need 110, I'm not putting a switch in for dual voltage. Two pairs of yellow and white wires? I can only find help in the threads for 220 or both, not just 110.

I assume the green/yellow wire goes to star ground and the IEC ground goes to star as well?

Also, my IEC has the built in fuse, does this complicate wiring the power switch in?

 

[desol]

For anyone trying to justify the tube swap with better gain pot control: please don't, there are better ways. Leave the line amp the ancients designed alone and don't ruin it's specs or the utilitarian "signature sound".

I wasn't the original one to suggest it?? It was him... >>

Sheesh.  ;D


Edit: Besides, wasn't it originally designed for a different application? One that the ancients designed it for?