All things LA2a related

[SSLtech]

Dave,

Bore da, Boyo...

The LA2a electro-luminescent panel in the T4 DEFINITELY responds to WELL above 2kHz... It simply HAS to, because if I put a (constant-amplitude) swept tone in, I get essentially-consistent LDR reduction ALL the way to the top of the audio band...

If the EFL panel was truly going 'dark' above 2kHz, there'd probably be some significant reduction in Gain Reduction above that frequency.

But, you touch on an interesting point.

IF the panel is (linearly) less efficient at a certain frequency, then the RATIO at that frequency would be reduced. If the ONSET of illumination was higher, then the THRESHOLD at that frequency would be different.

It's hard to measure inside an LA2a circuit, because it's inside a feedback loop. If -for a simple example- the light turn-on "threshold" was slightly higher, the resultant signal increase would -in turn- increase the drive AC signal to the EFL panel, which would eventually result in gain reduction... the effect doesn't actually DISAPPEAR, but it is 'concealed' by the feed-back nature of the circuit.

For example, if take the cover off a T4 on even a dimly-lit workbench, the GR caused by ambient light leaking onto the optos reduces the signal to the EFL panel so much that the EFL panel almost never lights up. (the drive signal to the EFL panel is sourced post-attenuation, so 20dB of GR caused by ambient light spill, means a 20dB reduction in voltage to the EFL panel... Unless you wind the "peak reduction" control all the way towards the end stops, the T4 will NEVER light up. -If you ever want to see the EFL panel glow in a T4, you essentially have to have the room DARK, DARK, DARK!

Now, taking the T4 OUT of a feedback operation to measure them (as I've done) by slapping a Crown amplifier across the EFL panel connections, and slapping two Fluke 87's across both sets of opto terminals (one DVM measuring each opto) and then feeding it with a constant amplitude swept tone, you can see that the EFL panel does indeed respond comfortably to the upper end of the audio spectrum.

In my measurement setup, I was using a Neutrik A2, which has a calibrated output control, so using a nominal resistance amount (say 1000Ω for example) it was an easy matter to stop the sweep at any given frequency of interest and see how much change in drive level (in dB) would be needed to 'correct' any deviation in opto resistance (which would of course correspond to any nonlinearity in EITHER the opto or the EFL end of things.

Yes there was some fluctuation, but nothing TRULY significant. All the values for opto resistance deviation for each of the T4s which we tested, (including two original Universal Audio T4s, one New-Old-Stock JBL T4, two 'BLOO' T4s and one T4 manufactured by CJ himself) were entered into an excel spreadsheet, for an amplitude sweep/resistance values table. A second test was started for frequency over dB change needed for same opto resistances, but was abandoned early on precisely because it was largely dismissed as not being significant...

So...

It would be silly of me to suggest that you're not seeing what you're seeing, but I do wonder WHY you might be seeing it.

My first question would be -are you certain that you're using a directly comparable material? -Please forgive my ignirance, but are there perhaps more than one type of fluorescent design? How are the electrodes and what effect might surface area/linear size have? -How about series impedance? -Also, have you considered that the series capacitance in a T4 might be there to equalise or compensate for a drooping light output with rising frequency?

While I've not investigated most of these last points, it's not that I have no interese, but the need was never really "pressing" if you know what I mean. -I measured flattish results in-circuit (don't forget that the LA2a also has response-shaping abilities in the peak reduction drive circuit...)

So there's some unfinished investigation results from Keith-land... the only other comment that I'd really like to add is that CJ did send me a beautiful T4 which he himself created, and he's definitely done PLENTY of measuring, though I don't know all of the exact tests he's run... He's published some on his site, and I've run other tests some of which I abandoned when they largely started to corroborate his earlier findings. (no point asking all the same questions twice, after all! :wink

You raise an interesting point however: -Maybe there is an inherent HF decline in most EFL materials which is 'corrected' for in the implementation... or maybe there's more than one type of EFL panel. -Maybe there's a DIFFERENT answer waiting to be proven... Only one way to find out...

Test, measure, publish!

Keith

PS: -Whereabouts in the Welsh hills are you? -And does your surname resemble a christian name following a 'P'?

 

[dave P]

Hi SSL Tech,
Where to start?
I never made the EL panel which is about 6"x3" but it was supplied to us by a former customer. I guess it was made of different luminescent material to the T4 units you have access to. I was driving an EL84 direct from a sig gen and the panel was a bluish green type, maybe the capacitance was wrong or the luminescent material was crap but it never made 10kHz. At first I thought there was a physical limit to the speed the electrons tunnel their way through to the veins of copper needles in the crystals and that was causing the problem. It is well known in the industry that these panels were made for 50-60Hz and that higher frequencies make them wear out faster due to gradual migration of the holes and copper ions I believe. The only way to restore the luminescent powder is to scrape it off, anneal it and put it back on!

I think my driving voltage was constant because I then stuck a TX on the EL84 and drove a bridge rect. straight on to the green diode which at 550nm was a perfect match for the LDR, and it went up to 100kHz which was why I sent the original post. However I will recheck my original drive to the panel to see if Iv'e cocked it up somewhere!

The light dependent resistors stuck on the panels are veeery sensitive as you say and they waste the drive signal so the panel never lights if you try them out of the can. I noticed how much top boost there is in the LA2 drive circuit but I thought it was for the radio broadcast pre emphasis application, maybe it compensates for the drooping response of the panels as well.

I live outside Builth Wells in Powys but I'm not Welsh just an English "incomer" as they so charmingly put it. I work half the week in Stevenage which is where I am now mate.
thanks for taking the time for such a long response.
all the best
Dave

 

[dave P]

Hi Keith,
I re-checked the circuit and the output falls as I said from 1-2kHz and as the panel has a capacitance of 68nF I guess this is the main reason, it's asking a lot of an EL84 to drive a load like that. As the panel is so big its capacitance must be much higher than the T4 panels, they must have a much smaller capacitance and a higher impedance to allow higher frequencies I guess.

I did the test in as near a darkened room as possible and stuck an LDR on the panel and measured the resistance as follows:-

100Hz 6.9k
200 Hz 3.8k
500Hz 3.3k
1kHz 7.9k
2kHz 680k
6kHz 950k
10kHz 1030k

I stuck a constant 7v into the EL84 so boosting above this would only put the tube into cut-off.
By contrast a green LED on the same LDR can get the resistance down to as low as 60 ohms so it would be possible to achieve more than 40dB reduction using one of these.

best
Dave P

 

[shabtek]

ive only used panels out of night-lights---aprox. 2"x3". blue and green, they respond to a constant voltage swept through pretty much the entire audible spectrum... light level does not seem to fluctuate too much, but light frequency/color does get more 'visually saturated' at a certain 'corner frequency'

 

[SSLtech]

Fascinating...

I have a very singular recollcetion of Builth, Wells. -That Jazz festival that they have there... I played there once with "Thomas Lang" Maybe 1988..?

{edit} -Ah yes.. the Brecon Jazz Festival! -Marvellous!!! {/edit}

Middle of a thunderstorm, outdoors, I was mixing front of house, and the "tent" roof above the band s-l-o-w-l-y filled with water during the set, and sagged lower and lower until the band were -no exaggeration at all- playing on their knees... -I'm pretty certain that I've even seen a reference to it on the Thomas lang web page.

I remember lead-grey skies, incredibly low headroom doorways here and there, and an even lower headroom stage canvas by the end of the set!!! (I think I banged my forehead on a doorway lintel on the main street!)

Anyhow...

EL84 = 12BH7 I presume? (sorry, not too familiar with equivalences, my failing, not yours)

So... You may have to scope the feed to the panel... there may be some droop to it. -I used an octal screw-terminal base, which made all the measurements fabulously easy using the methodology which I described earlier.

Part of the problem in that test might be the sheer SIZE of the panel.. a real panel is under a square inch, you were loading the drive circuit with about 18 square inches, that puts more load on at HF, I'd imagine.

Stevenage, bit of a commute isnit?

Keith

 

[AnalogPackrat]

EL84 = 6BQ5 which is a slightly bigger brother to the 6AQ5...

6BQ5 = Voxy goodness
6AQ5 = LA-2a sidechain happiness

I think Keef is onto something with the area thing. Surely the capacitance of the EL panel is related to its area. Anyone measure the C of a T4b El panel? I just ran a back o' the spreadsheet calc and the reactance of 68nF drops below 1k at only ~2500Hz. 3.8nF (1/18th) still has a reactance of over 2k at 20kHz...if my late evening math is not flawed.

A P

 

[CJ]

Keefer, ol boy, what is your John peel library like?
got any tapes, etc?


Oh, tell peter purpose tyht I want his dond, does it taste like bamgers and mash witjh HP sauce addaed?


Did you ever meet him?

I have a Saucer Fuyll of Sectets tape that is pretty weirdass, dirt under the figner nail saof david and rodger.


is this a roast?
no,

 

[dave P]

Thanks Guys,

thats it then, its all down to the size of the panel. CJ have you tested the Capacitance on one of your panels? Is it about 3nF?

I have nearly finished the translation into English of the V76 technical description, (I did the others V's last year) can someone host it for me if I email them the pdf? I don't have a website as yet.

another day another mystery solved,
thanks to all

dave P

 

[Bluzzi]

I'm just starting to verify my second LA2A before I plug it in and I think I screwed up.

I wired up an A24 as input (I find the A10 too much) but I think I should have wired it backwards to get a 1:5 that I want.

If I'm not mistaken as an output it is wired as 5:1. As an input I should have wired it 1:5 so I should have it backwards compared to the A10 right?

jim

 

[khstudio]

I JUST finished my third LA2A using DRIPs PCB (my first 2 were Point to Point)
EDCOR XSM Series Transformers I/O

NO Meter Switch - hardwired for GR
(put the Limiter Response on the front in place of it)

I'll post Pics soon. It's sounds great. :green:
It seems those EDCORs are going to be fine & saved me a TON of money.

 

[khstudio]

[quote author="Bluzzi"]I'm just starting to verify my second LA2A before I plug it in and I think I screwed up.

I wired up an A24 as input (I find the A10 too much) but I think I should have wired it backwards to get a 1:5 that I want.

If I'm not mistaken as an output it is wired as 5:1. As an input I should have wired it 1:5 so I should have it backwards compared to the A10 right?

jim[/quote]

I'll tell you man... 1:1 (actually 10k:10k) for the input is the way to go.
All 3 of mine use them & I LOVE em'. :wink:
I couldn't imagine 1:10 or even 1:5 in front of it... especially if your planning on interfacing it with Modern 10k GEAR.

 

[Bluzzi]

Kevin - Both of mine are P2P too! I started before.

1:1 really! Interesting. All I have though for now is some extra A24 so I feel its better than the A10 at least.

Anyone can confirm if I goofed on the wiring?

jim

 

[crisotop]

did you have a center tap ?
otherwise youll have to put some resistors to ground and to each leg of the heaters.

I don't have a center tap on my 6.3V secondaries, does that mean I'll use two 100Ohm 2W resistors as suggested in the drips opto manual and solder them from each heater leg to star ground?

Many Thanks
Christoph

 

[khstudio]

[quote author="crisotop"]

did you have a center tap ?
otherwise youll have to put some resistors to ground and to each leg of the heaters.

I don't have a center tap on my 6.3V secondaries, does that mean I'll use two 100Ohm 2W resistors as suggested in the drips opto manual and solder them from each heater leg to star ground?

Many Thanks
Christoph[/quote]

Yes!
I don't think they have to be 2watt's... I've used 1/2 or 1 watts before.
Also, any ground will work... around the Power Supply GND would be best though

 

[cannikin]

Hey Kh,

How did you hard wire your GR switch?

 

[khstudio]

[quote author="cannikin"]Hey Kh,

How did you hard wire your GR switch?[/quote]

Check Here:

LA2A Complete - with DRIP PCB
http://www.groupdiy.com/index.php?topic=26151

You may be able to see it in the pics... if not I explain it about halfway down.

 

[crisotop]

Yes!
I don't think they have to be 2watt's... I've used 1/2 or 1 watts before.
Also, any ground will work... around the Power Supply GND would be best though

Many thanks Kevin! I'll give that a try - hopefully I won't end up with hum problems

cheers
Christoph

 

[crisotop]

Another short question - does anybody know a supplier in europe that sells 22GA (0.75mm) shielded 3 conductor wire? I can only source 2 cundoctor wires?

thanks in advance
Christoph

 

[feza]

Hey,

After reading A LOT of info about this phat compressor I want to build it myself. So I am now doing basic soldering projects and eventually doing this 'bad boy'. Posted on another site some questions about more expensive added parts didn't really helped me a lot. But posting these questions again here might reveal more. :wink:

And by the way Tat Purusha and Drip I am the one who mailled you about the parts. Finally got activated. Really guys it gives me a good feeling being here and thinking we might grow old on this forum.

I'm off reading more then 160 threads

 

[feza]

Yo,

I did some questioning on another forum about adding more expensive parts then the parts given with Drip's guide. Here's what I want :

Jensen caps, Ceramic Gold plated sockets, more expensive tube, Ohmite resistors and probably a stepped attentuator. But I don't know which FRED will work with the LA2A?

Basically I want to build a no-compromise LA2A. Are there any more "enhancements" to be done? Keep in mind that I am not yet a very skilled person. So some pointers would be great! Furthermore building a tube compressor is dangerous. What are the things I must look out for? The caps keep electricity and ESD comes to mind. Good working tools are mandatory so I am collecting the right things and reading a lot about soldering.

Last but not least you can power-up the unit without having it all assembled. How 'barebone' should it be when checking if it works? The T4 and no tubes out right? And how long should it take to get no harm from electricity from the unit?

Well I am off reading more but these are the questions I got in mind before the build. I'll wait about 3 or 4 months attempting to build the unit myself and not planning to questioning much more what can be teached by myself, mates. :wink:

Greetings,
Dennis