T4B Photocells

The UREI LA-3A schematic shows an opto-attenuator with 4 photocells:


but just Googling around for off-the-shelf replacements, many including UA's look like they only have 2 photocells.

In the schematic, 2 are paralleled and another is shunting the luminescent panel supply. Is paralleling really the same as just one? Why parallel? For improving tolerances? Is the one shunting the panel to linearize the response?

Does anyone know how these things work and if only 2 photocells are really necessary?

The one in the middle looks like a trimmer across the EL element. Opto comps Only need 1 for signal squishing, 1 for metering...la3a copy might be different. I thought t4b were interchangeable

The LDRs have the lambda symbol.

PRR said:

The LDRs have the lambda symbol.

Click to expand...

Yeah? Is there some meaning behind that?

I have to wonder if some of the stuff in that dashed box is actually just some horse sh-radish  put in there to throw people off. Just like the box labeled "UA4706 T+C" found in the 1176 schematic.

All of the parts to make these cells are readily available and cheap. The photocells are on Allied and Digikey (not Mouser), the electroluminescent panels are all over the web cheap, RF metal shields to block whatever are not hard to find. This could all be mounted directly on little PCB sort of like a discrete op amp and without the enormous socket.

And the B11184 auto-transformer looks a whole lot like the primary of a 70V line transformer which is ~$10 on parts-express.

Clones of the T4B or homemade versions had been extensively done here and on the web.
Not easy to get the same attack and release times of the original, some people got it better than others.

Search for DIY T4 in google

> Yeah? Is there some meaning behind that?

I don't understand your antagonism(?).

The original LDRs were extensively tested and sorted for their attack/decay rates. A "T4B" built of random oil-burner photoeyes will not do it right. But do it if you want to. I have.

Note that PbS cells are fading from the market, already illegal some places. Make your lifetime buy now.

The glow-panels can easily be re-rigged with LEDs.

PRR said:

> Yeah? Is there some meaning behind that?

I don't understand your antagonism(?).

Click to expand...

I don't either.

lambda could be used to mean  (light)wavelength\photoresistor, if you add an 'a' lambda becomes the forbidden dance

..Aah - a hidden warning about the later-to-be-forbidden substance..

indeed. no mere coincidence either that you are seeing red; both the color and the spanish pronunciation

Here is how you can simply do your DIY version, easy stuff:

https://www.youtube.com/watch?v=HB86Mw4mZLw

Check how a T4B is made, you have assembly pics:

https://sites.google.com/site/diypartsstore/catalog/new-t4b-opto-el-panels


more stuff:

http://www.igsaudio.com/t4bx

http://www.kenetek.com/T4AT4BUnits/KenetekT4BModules/tabid/300/Default.aspx

I thought that Cadmium is a banned substance....

Bri

  Don't mention that word or UE will shout down the forum!

  The ones with the lambda are the photocells, the others are just trim pots, that's the arrow as we have seen the trim pots drawn for the last century. The lambda was later replace for a few small arrows pointing the light sensitive device.

JS

joaquins said:

The ones with the lambda are the photocells, the others are just trim pots, that's the arrow as we have seen the trim pots drawn for the last century. The lambda was later replace for a few small arrows pointing the light sensitive device.

JS

Click to expand...

That is interesting and it makes sense. Thanks.

I have read the various responses, links and materials about these T4B cells. I am posting a summary of my findings here so that I can reference it later. Maybe it will help others like me who are starting with no knowledge of how the T4B works. Please feel free to correct me if I misunderstood something.

The T4B consists of two Clairex CL-505L Cadmium Sulfide photoconductive cells, one for gain reduction and one for metering. Both are illuminated by the same electroluminescent panel which is driven directly by the AC input signal filtered and stepped-up to no more than ~90VAC peak. The EL panel has a 4n7 cap in series and a "trimmer" resistor in parallel. The gain reduction cell has a trimmer resistor in parallel although some earlier units seem to also have a CL-705HL 332 in parallel.

Here are some key specs for the origial Clairex CL-505L from the Clairex Photoconductive Cell Application Design Handbook:



The light and dark resistance of this cell looks like it varies between <1K to >1M in the dark.

So the Type 5 CdS are most sensitive to yellowish light and the "Rise" time is quite slow taking several seconds to fully recover. There is a huge variation in the performance of these parts so two units must be carefully selected. There is also a "memory effect". If the device spends most of it's time in the dark, the response will be significantly different from that of a unit that spends most of it's time in the light.

[Although this effect is explained in some detail in the above mentioned handbook, I don't completely understand it. But my impression is that it could be important to compression behavior. If there is a moment of silence in the input, when the input signal is restored, the cell will "remember" the previous resistance level and thus generally serve to provide a consistent signal level.]

The modern replacement part of choice for the CL-505L seems to be the Silonex NSL5910 CdS photo cell (although UA use NSL-02-042 0310). Unfortunately the new cells are measurably faster and the EL panels turn on at a significantly lower voltage and are generally brighter. Although the Kenetek part seems to be quite good they do acknowledge in their video that it is simply not possible to replicate the behavior of the original TB4 using new parts. I don't know how significant the differences are in practice.

So yet again, "better" is not always desirable.

squarewave said:

I am posting a summary of my findings here so that I can reference it later. Maybe it will help others like me who are starting with no knowledge of how the T4B works.

Click to expand...

I just wanted to thank you for the summary here, being exactly the person you described

I've been reading into the subject so as to feel comfortable enough to measure, match and ultimately replace a failed 5910 in one channel of my Requisite L2M Mk2 - I feel like I actually learned something today  8)

In hindsight it's probably a much safer option to buy one from Kenetek. That guy obviously has a process that is going to be difficult to match and the price is more than reasonable. I recommend a "slow" cell since the older LDRs were apparently very slow. That's sort of the characteristic of a T4B. I used one of the Kenetek cells in my LA3A and it can take multiple seconds to fully recover if I hit it hard enough.

Anyone knows why 2 LDRs in parallel inside the T4B on the GR circuit?

The very right resistance element is a trimmer, not LDR

I had a Telektronix unit, and when I replaced the EL panel, I seem to recall three CDS cells, one of which was bigger..

Could that be? or am I imagining things? (it was 15 years ago, or more)

Yes I confirm, just opened a T4B from an LA3A, was a smaller LDR in parralel of the big one at one side !