DIY T4B, matching EL Panels and Cells

[CJ]

you guys need any info/

 

[shabtek]

CJ!!!            yes, are you ok. how can we get to the smokehouse..my enter and question mark keys just went dickey

 

[DaveP]

Radardoug,

Its really hard to get sufficient voltage out of the transformer to drive the LED's so series was not an option, they both turn on instantly in practice in any case.  In fact you need RURP8100 schottky diodes, or similar, to minimise the voltage drop on the bridge.


Mitsos,

The LED's came from RS, part no. 654-2538, they are Osram 5mm link below:-

http://uk.rs-online.com/web/search/searchBrowseAction.html?method=searchProducts&searchTerm=654-2538&x=20&y=14

For the housing I used a solid Aluminium rod about an inch long and diameter wider than the Allied photocell, In fact it was a length of doorhandle!  Hey, this is DIY.  I drilled a 5mm hole through the centre and a wider hole at the other end for the photocell, you need snug fits for this.  I should emphasise how critical it is to get absolute light tightness for this as the photocells are very sensitive. Same applies to T4B, in fact.

You need to draw out a little sketch showing the angle of light output from the LED so that it matches the dia of the photocell, this gives you the optimum distance between the opposing devices.

Once you have made these, they will outlive you and never need changing or touching again, they are easily mounted inside the chassis on pipeclips or small cap clips.  I call this an LA-2D, its so simple to use and I use it all the time; really smooth sound on vocals.
best
DaveP

 

[ruckus328]

Dave, curious - what's the max possible voltage to the LED if using a 30:1 transformer like you've shown?  I'm not very familiar with the LA2A.  Looking to figure out how this approach could be tried with an LA3A, which looks like it could eliminate the step up transformer used.

 

[SSLtech]

Nice and inventive!

It bears pointing out however that there are a few changes which will affect the resulting behaviour and compression characteristic.

The LED/opto caps could be the same off-the-peg solution as  I used in my LA-4 clone... should work perfectly, and MUCH less work. It deals tightly around the opto, since they're slightly elastic.

HOWEVER... I hoonestly don't think that it'll behave the same as an LA-2a because of the way the LEDs are used... I've tried Parallel LEDs in the past, and found that the turn on points weren't consistent. Also, the turn on threshold behavour and the increase in light output above threshold won't be the same as an EL panel. Also, there's a particular response curve that the EL panel introduces into the sidechain, since it's a highly capacitive load. -Sweep and scope the sidechain of an LA-2a to see what I mean.

The transformer approach (a la LA3, but in reverse, so to speak) is novel, there'll be a bit of DC involved, which will doubtless place some imposition on the transformer, but I can't speak from experience, since I haven't built one.

I found that the LEDs HAD to be in series, or turn-on threshold was simply too unreliable. -I also found that I had to modify the sidechain response in order to track the rather complex LA-2a effective-ratio-versus-frequency curves. Rochey and I have had an EL-free T4 module essentially ready-to-go as a plug-in module for about 2 years, but it's taken this long to get it into a T4-sized module, which requires using SMT components. The remaining issue which is holding this back as a saleable item is opto matching and a standard test procedure.

Once that's all done, we have a product which plugs into a T4 socket, and works indistinguishably from a UA T4 module. It is El panel-free, they've been running for about 28 months and showing ZERO signs of aging. They've been left on driving 30dB compression on a tone for about 3 hours steady, trying to 'burn' or 'wear-down' the any part of the light source circuit, and so far they STILL match brand-new T4's perfectly.

In order to finish the job, we have to get the opto matching nailed down reliably and consistently... whether you're making EL-panel T4's, LED T4's or 'Turbo' T4's, that's one thing which ANY builder will always face, as Joe (Kenetek) correctly points out.

So the question arises... Who prefers the approach of making a modified circuit such as this, as opposed to redesigning the T4 module in and of itself, to eliminate the EL panel WITHIN the unit, but still allow the user to plug in a 'genuine article' T4 module, which would of course not be possible with a circuit modification approach such as this?

I'm curious as to whether we should continue to develop ours?

For reference, I've had this switcher box installed in a studio with two new-condition reissue LA-2a's side by side:

It's plugged into the back of  one of the LA-2a's thus:

Which allows people to switch between an old-stock UREI 'vintage' T4B which is slightly aged, a used T4-A, a new Universal/Teletronix T4B, and the Expat Turbo.

Results are that most people can identify the UREI T4B and the T4A, because they have strong characteristic differences... but NOBODY has been able to tell which is the 'real' UA T4 versus the prototype 'Turbo' unit. -The switch positions are numbered, but the sockets are not, and they are sequenced in a non-intuitive order, so that people can't tell which is which by looking at the order of the devices under test.

Thoughts, anyone?

Keith

 

[SSLtech]

Dave, curious - what's the max possible voltage to the LED if using a 30:1 transformer like you've shown?  I'm not very familiar with the LA2A.  Looking to figure out how this approach could be tried with an LA3A, which looks like it could eliminate the step up transformer used.

...And you basically build yourself an LA-4.

 

[joe-electro]

Good work here. I've been fooling with all kinds of LED's for years and never got them to sound the same as an EL panel.  I'd be very curious to put one of your units on my T4B tester and see what it looks like. PM me if interested.

Thanks,

Joe

Nice and inventive!

It bears pointing out however that there are a few changes which will affect the resulting behaviour and compression characteristic.

The LED/opto caps could be the same off-the-peg solution as  I used in my LA-4 clone... should work perfectly, and MUCH less work. It deals tightly around the opto, since they're slightly elastic.

HOWEVER... I hoonestly don't think that it'll behave the same as an LA-2a because of the way the LEDs are used... I've tried Parallel LEDs in the past, and found that the turn on points weren't consistent. Also, the turn on threshold behavour and the increase in light output above threshold won't be the same as an EL panel. Also, there's a particular response curve that the EL panel introduces into the sidechain, since it's a highly capacitive load. -Sweep and scope the sidechain of an LA-2a to see what I mean.

The transformer approach (a la LA3, but in reverse, so to speak) is novel, there'll be a bit of DC involved, which will doubtless place some imposition on the transformer, but I can't speak from experience, since I haven't built one.

I found that the LEDs HAD to be in series, or turn-on threshold was simply too unreliable. -I also found that I had to modify the sidechain response in order to track the rather complex LA-2a effective-ratio-versus-frequency curves. Rochey and I have had an EL-free T4 module essentially ready-to-go as a plug-in module for about 2 years, but it's taken this long to get it into a T4-sized module, which requires using SMT components. The remaining issue which is holding this back as a saleable item is opto matching and a standard test procedure.

Once that's all done, we have a product which plugs into a T4 socket, and works indistinguishably from a UA T4 module. It is El panel-free, they've been running for about 28 months and showing ZERO signs of aging. They've been left on driving 30dB compression on a tone for about 3 hours steady, trying to 'burn' or 'wear-down' the any part of the light source circuit, and so far they STILL match brand-new T4's perfectly.

In order to finish the job, we have to get the opto matching nailed down reliably and consistently... whether you're making EL-panel T4's, LED T4's or 'Turbo' T4's, that's one thing which ANY builder will always face, as Joe (Kenetek) correctly points out.

So the question arises... Who prefers the approach of making a modified circuit such as this, as opposed to redesigning the T4 module in and of itself, to eliminate the EL panel WITHIN the unit, but still allow the user to plug in a 'genuine article' T4 module, which would of course not be possible with a circuit modification approach such as this?

I'm curious as to whether we should continue to develop ours?

For reference, I've had this switcher box installed in a studio with two new-condition reissue LA-2a's side by side:

It's plugged into the back of  one of the LA-2a's thus:

Which allows people to switch between an old-stock UREI 'vintage' T4B which is slightly aged, a used T4-A, a new Universal/Teletronix T4B, and the Expat Turbo.

Results are that most people can identify the UREI T4B and the T4A, because they have strong characteristic differences... but NOBODY has been able to tell which is the 'real' UA T4 versus the prototype 'Turbo' unit. -The switch positions are numbered, but the sockets are not, and they are sequenced in a non-intuitive order, so that people can't tell which is which by looking at the order of the devices under test.

Thoughts, anyone?

Keith

 

[SSLtech]

Good work here. I've been fooling with all kinds of LED's for years and never got them to sound the same as an EL panel.  I'd be very curious to put one of your units on my T4B tester and see what it looks like. PM me if interested.

We might be able to make that happen... Let me ask Rochey and see what he says.

Also, I might have to 'pot' the light source part first, if that's okay with you. Using OEM optos, I've had indistinguishable results. Using modern off-the shelf optos has introduced the component-selection-and-matching lottery, as I'm certain you can appreciate.

By the way, that one in the picture is the old HUGE 'development' can. -It now fits into a 'standard' can, basically identical to the Drip Electronics one. We can even switch the El panel curve emulation characteristic from standard to more-linear and overall slightly more aggressive. -Subtle effect, but useful.

I'll talk to Rochey and see what he thinks.

Keith

 

[joe-electro]

That's cool.  Just let me know.

Thanks,

Joe

Good work here. I've been fooling with all kinds of LED's for years and never got them to sound the same as an EL panel.  I'd be very curious to put one of your units on my T4B tester and see what it looks like. PM me if interested.

We might be able to make that happen... Let me ask Rochey and see what he says.

Also, I might have to 'pot' the light source part first, if that's okay with you. Using OEM optos, I've had indistinguishable results. Using modern off-the shelf optos has introduced the component-selection-and-matching lottery, as I'm certain you can appreciate.

By the way, that one in the picture is the old HUGE 'development' can. -It now fits into a 'standard' can, basically identical to the Drip Electronics one. We can even switch the El panel curve emulation characteristic from standard to more-linear and overall slightly more aggressive. -Subtle effect, but useful.

I'll talk to Rochey and see what he thinks.

Keith

 

[SSLtech]

Oh, Dave P, looking at your schematic, it looks as though your HT neon is strapped directly across between the HT line and ground...

It might be worth being VERY clear that such a neon would have to be the type with a series resistor, I'd have thought, since they tend to 'clamp' voltages, and a NON-series-resistor version is in fact used for this precise purpose in the original version, stabilizing the GR meter deflection. I think if you were to use the original type of neon, the results might be a little 'spectacular'...

Also, I recall that in my earlier playing-around, I noticed another characteristic difference in this LED behaviour: it's only half-wave detection. A much better result was achieved by wiring a second diode in "reverse-parallel", and pointed also at the photocell. It made the physical mounting somewhat more complicated, but in listening tests, the results were closer to the EL panel, PARTICULARLY with largely asymmetric waveforms such as human voice.

Keith

 

[SSLtech]

That's cool.  Just let me know.

Will do.

Actually, I'm almost going out of my mind right now, clearing out the last remnants of my old workshop, and in the middle of a relocation which is driving me compeltely round the f#@%&*ing pipe!

And yesterday, to cap it all- I found a big bag of about 20 LA-2a trimmer capacitors, and spent about half an hour asking myself why I would have them...





...Ahem...





I think I was supposed to mail them to you, about three years ago.




...Sorry 'bout that!

Still want 'em? -I'll toss them in the mail if you do.

Keith

 

[lassoharp]

Results are that most people can identify the UREI T4B and the T4A, because they have strong characteristic differences... but NOBODY has been able to tell which is the 'real' UA T4 versus the prototype 'Turbo' unit. -The switch positions are numbered, but the sockets are not, and they are sequenced in a non-intuitive order, so that people can't tell which is which by looking at the order of the devices under test.

My question would be:  Have the listeners shown a sonic preference for either type over the other?

 

[SSLtech]

That's one thing which is all over the map.

Different music, different instrument, even same instrument different TUNES from the same band produces different preferences -even from a single person.

Preferences are things which I don't try to second-guess. -Replication to as deep a level of accuracy as I can achieve... now that's something I can put my weight behind.

However, NUMEROUS modules from across several years have been cycled through the switcher, both for listening AND measuring comparisons. Similar types of different ages, and vastly different types also (T4-A, T4-C etc). I eventually chose two of those which were most obviously different to leave for the long-term switching/listening tests.

Keith

 

[DaveP]

Hi Ruckus328,

The LED operates at ~4V dc and the bridge has ~3V ac across it, you can get ~100V on the primary, you need the big step down to get enough current for 2 LED's.  Putting them in series would need a more expensive transformer than the simple $10 RS one I used and a tube with more poke.

SSLtec,

Not everyone trusts themselves to mess with original circuits so your product will satisfy the majority I guess, my advice is carry on.

I was building my own from scratch so it was no problem to redesign anyway.  As I mentioned earlier, I think absolute matching between pairs is not so critical because a lot depends on the VU meter used.  I used a genuine VU meter from Sifam costing over $100 and I had to modify the associated bridge to calibrate the output to what it said on the meter.

I really don't think the rise time of the LED/EL panel matters too much, we are only talking a few microseconds here, what matters more is the recovery time of the photocells and that has nothing to do with LED's or EL panels.

best
DaveP

 

[DaveP]

SSLtech,

Yeah your right, my neons have a built in series resistor.  There was I thinking how lucky you were to have my carefully drawn schematic too!

best
DaveP

 

[joe-electro]

PM'd re: trimmer caps.

Thanks Keith!

Joe

That's cool.  Just let me know.

Will do.

Actually, I'm almost going out of my mind right now, clearing out the last remnants of my old workshop, and in the middle of a relocation which is driving me compeltely round the f#@%&*ing pipe!

And yesterday, to cap it all- I found a big bag of about 20 LA-2a trimmer capacitors, and spent about half an hour asking myself why I would have them...





...Ahem...





I think I was supposed to mail them to you, about three years ago.




...Sorry 'bout that!

Still want 'em? -I'll toss them in the mail if you do.

Keith

 

[DaveP]

Ruckus328,

Correction #2, Although I tried putting a bridge on the TX I found the LED's worked better without, so strike my earlier comment.

You get 4V dc on the diodes and about 100~120V ac audio on the primary from the 12AT7, thats clean before clipping too.

best
DaveP

 

[ruckus328]

...And you basically build yourself an LA-4.

Well, don't know I agree on that one.  Alot more differences between the 2 units besides the G/R element, as you know.  And the point here is Dave indicated there were little to no sonic differences between a T4B and his method.  Don't want to spill the beans quite yet but I have a new opto design in the works that I'd like to try out Dave's method as well as your pluggable version too Keith.

Sidenote - where the hell you been?  ;D

 

[peter purpose]

Dave.... v interesting that little drawing. I was toying with the use of leds, but would have stuck em right on top and bolloxed it up. I feel an experiment or two in the works.

 

[joe-electro]

I got the best results by putting the LED's behind a piece of white, translucent plastic to diffuse the light and leaving some space between the LED's and the plastic to get the light to disperse more evenly across the entire surface of the plastic.  I've also found that the bright blue LED's have the best dynamic range.

Joe