Tape Echo Unit Bulb Replacement

[sarakisof]

Recently found a nice old Lem 120&180W  mixer PA amp rack unit with Lem baby tape echo on top. I'm in the middle of repair and restoration.
Amongst others, the indicator miniature lamp which is related to the mixer pre is dead. Desoldered from circuit and confirmed measuring open by multimeter.





It hasn't to do with mains power indicator light (this is illuminated inside on off rocker sw).

Apart from the huge power trafo (mains amp power) which is located on the bottom rack, there's a second little one (sec: 2 x 34VAC) on the top echo mixer which provides voltage to the mixer/tape echo. It has two small secondary windings.

First pair provides 34VAC to the mixer.
Second pair provides 34VAC to that miniature bulb.

Voltage on mixer rails is +28VDC.

I'd like to find something close/similar to this. Incandescent miniature bulb 30V should be fine(?), but what about Wattage? What specs I'm looking for?

Needless to say there's no scheme for this unit anywhere (it's different from the common LEM Baby which is available online).

Any thoughts?

[PRR]

I was gonna ask: do you need it? And can you use LED? But then I worked out that this is a side-fire lamp housing for use on a dark stage. You might need it (though you can find fine LED clip-on book-lamps) and a plain LED won't spray sideways like an incandescent.

Here in the US, I would look for a telephone lamp, 48V and very small Watts. On 34V they would live another 50 years.

https://www.grainger.com/category/lighting/light-bulbs-lamps/miniature-light-bulbs-lamps/telephone-slide-base-miniature-light-bulbs-lamps?attrs=Voltage%7C48V+AC&filters=attrs
https://www.jkllamps.com/48MB
https://www.ship-store.com/product-page/790856-telephone-switch-board-lamps-48v-40-ma-5-5x30-mm

Of course with the death of both wired telephones and incandescent technology, it may be a scavenger hunt finding old bulbs.

That 3rd link is in İstanbul, your part of the world; but they do seem to be about "to 50,000 HP diesel engines" more than dollar lamps.

[antonis]

Second pair provides 34VAC to that miniature bulb.

If it's directly and only provided to that bulb, estimaste (or measure by connecting appropriate resistive load for voltage sagging) secondary current and go for a bulb of 10 times lower current rating..

[sarakisof]

estimaste (or measure by connecting appropriate resistive load for voltage sagging) secondary current and go for a bulb of 10 times lower current rating..

How could I estimate? I should know trafo's core dimensions I guess? It's a small one though, think smaller than an output one in tube amps.

Of course with the death of both wired telephones and incandescent technology, it may be a scavenger hunt finding old bulbs.

Fortunately I have access to a large quantity and variety of'old school' incandescent miniatures.
So far, I've found these:

Wire Ended 36V 20mA


Telephone bulb 36V 20mA


E10 / BA9S bayonet 36V 20-50mA


Which one would you prefer?
I guess a 36V and 20-50mA would be a safe zone choice right?

[antonis]

How could I estimate?

Measure "open" (unloaded) voltage..
Say, 36VAC..

Place a 1k resistor across it..
If measured voltage remains unaffected, lower resistor's value till voltage starts to sag..
Calculate current via Ohm's  Law..

P.S.
From your photo I vote for 300mA (or so) transformer..
So, any of the above bulbs should be fine..

[sarakisof]

Perfect.

P.S.
From your photo I vote for 300mA (or so) transformer..
So, any of the above bulbs should be fine..

Yeap me too. From my little experience on old equipment too I don't think there were such miniature incandescent bulbs rated below 20mA (specially for such volts) back then. We have to think out of "led world" here.

Cheers everyone 🥂

[Rob Strand]

How could I estimate? I should know trafo's core dimensions I guess? It's a small one though, think smaller than an output one in tube amps.

I suspect the transformer is powering more than the lamp.   When the transformer is powering the rest of the circuit the transformer is already loaded and is some regulation sag already present.   You can get old solid-state circuits where the circuit actually takes less power than the lamp so the sag is caused largely by the lamp!   Another factor is whether the lamp winding only powers the lamp or is also powering other circuits.

To get an accurate model of the situation requires too many measurements.  A simple estimate would be to turn the unit off then measure the DC winding resistance feeding the bulb, assuming the bulb is powered from AC.   From that you can ball-park the sag with different lamp loads.

If you measure all the winding resistances and voltages and the dimensions of the core I can estimate the transformer VA rating.  However there is still a missing piece of the puzzle: how much the rest of the circuit is drawing.   A simpler estimate of the transformer VA can be made from the mains voltage and the primary winding DC resistance, and the outer dimension of the core (the wider dimension of the core, not the mounting).   What that could do is put an upper limit on the bulb size.

I think the biggest piece of information missing is the power rating of the bulb (in watts).  That will determine the brightness.  It's not so simple working out the power rating of a dead bulb.

Running a bulb under voltage will reduce light output but will increase life.  Running a bulb over voltage will rapidly decrease its life!

[sarakisof]

I suspect the transformer is powering more than the lamp.

Rob, as mentioned on my first posting this trafo has 4 secondary windings.
First pair provides 34VAC to the tape echo preamp (top of the unit).
Second pair provides 34VAC to that miniature bulb and 6 mixer channel preamps.

I don't think a 20mA bulb would hurt so much, nor that this dead to be replaced bulb was originally rated below 20mA's.
We talk about mid 70s here, have never met any indicator incandescent lamp rated below 50mA on any of my repaired old gear so far.
Reading your inputs got me anxious though, thinking leave it open as is.
Don't need it anyways, as PRR correctly said, it's just a fire-light for use on dark stages.
Only reason I'd like to fix, is my passion & persistence in "fix to original state" habit.
And I feel most of you guys can feel me,
that's why I love this place.
Most of "consumer electronics" / arduino / house electrician friends of mine were like "you don't need it, just duct tape those two winding wires" or "get a chinese led & resistor, fix" ..

Another factor is whether the lamp winding only powers the lamp or is also powering other circuits

1A. It also powers the mixer faders/pre's/eqs.
34VAC, rectified by BY164 silicone bridge, gets filtered by two caps (1000&2200 - 35&63V) then providing the channels/faders with +28VDC.
1B. And the lamp.

2. The other two trafo's windings which power the top tape echo preamp/circuit follow a similar path (+34VAC , BY164 bridge, etc).






The circuitry for tape echo & mixer is similar (not the same though) with the common LEM Baby tape echo. Big differences come in power amp (with mine having two dedicated racks with huge trafo's for two PA amp ss heads 120 & 180w), if that helps a bit.
LEM Baby scheme can be easily found on net. Still waiting for a diagram response from GEM company in Italy. Don't have big expectations though.

[Rob Strand]

I don't think a 20mA bulb would hurt so much, nor that this dead to be replaced bulb was originally rated below 20mA's.

20mA at 34V is 680mW.   0.9W/1W lamps are pretty common for general indicators.  Control panels and things like that can use considerably higher power bulbs.

To me the transformer looks a bigger than a little 2VA to 4VA type.   If you remove the mains and measure the DC resistance across the plug that will give you the primary DC resistance.    For a 240V main 2VA to 4VA primaries are going to be over say 800 ohms to 1k or so.  4VA with a 1W bulb is still on the feasible side.  If you see 300 ohms to 600 ohms then it's highly likely the transformer has significant reserve to drive a 1W lamp, maybe even higher power lamp.  One caveat if your mains aren't 230V/240V those resistance figures need to be changed.

[sarakisof]

230V Mains after going to fuse & rocker on off switch on top unit, travel via Hirschmann/Molex kind of heavy duty connectors all the way down to the bottom Power Supply rack where the mains power huge trafo is located. This 400VA trafo gives then 230V secondaries and share them to every rack shelf (120W , 180 W and Tape Echo / Mixer rack units). Each rack unit has its own trafo.

The Echo/Mixer unit (which is our subject here) has the trafo pictured above. Its primaries are the 230V secondaries of the bottom power supply 400VA trafo.

Removing this little trafo's (let's call it Mixer trafo) primaries (left unconnected) they measure 247 ohms.
Mixer trafo's secondaries which go to the lamp measure 19.2 ohms.

[Rob Strand]

Removing this little trafo's (let's call it Mixer trafo) primaries (left unconnected) they measure 247 ohms.
Mixer trafo's secondaries which go to the lamp measure 19.2 ohms.

I'd estimate that small transformer to be about 10VA (upto say 12VA).  If we assume the two ~34VAC transformer windings are of equal power that means you have 5VA each winding.

From the resistances I'm guessing the secondary is wound over the primary.

With a 20mA lamp load, the voltage transformer voltage will drop by about 0.5V when the lamp is added, so a final voltage of about 33.5V.

As far as a nominal transformer voltage goes I'd say it's about 33VAC (32VAC at a squeeze) but since the transformer isn't under full load, even with the lamp present, the actual secondary voltages will be higher than the nominal.  (Nominal transformer voltages are given at full AC load.)

It's possible to do a little better if we factored in loading due to the rest of the circuit.  However, those numbers are good enough to work with.   You could use even a higher power bulb without any issues.

[sarakisof]

Hats off Rob 👍

Cheeers 🥂

[sarakisof]

After success story with the lamp I'm back with a last mystery I'd hopefully wish to solve.

Amongst others, previous owner had replaced the original tuchel/DIN input & output connectors

with plastic mono switching jacks for the six inputs (switching function left out of play, wired as common mono guitar open jacks, probably those were all he had in hand at that moment & just used them) and XLR s for the two outputs (wired as mono, left one pin lug unused).







He was right as circuit calls for mono connections. From the mixer PCB cards there are just instrument/mono/shielded interconnect cables like those used in guitar pups.

Schematic for LEM Baby unit (which, for that part of the circuit, is almost the same with mine) calls for just mono connections everywhere. No electronically balanced circuitry anywhere.


But, found a YT vid from an Italian guy restoring the lem baby and took some close prtnscrn shots from the insides of the back of his unit, showing the original input/output wiring originally used on that LEM Baby model which had jacks.
As you can see if you zoom in they are  open mono switched jacks with the Tip lug connected to GND lug  (what I'm missing here, maybe some jacks were different back in the days?)...



Anyways, what was the purpose of a shorted/switched connector originally used there? How do you think DIN connectors could be wired originally in my unit? I think it was a pseudo 3 prong connector by connecting "-" lug to gnd lug..

Now that I already have those plastic insulated switching jacks that previous guy put in there, I should connect their gnd lug (= PCB cable's shield) to chassis GND (mains "star" gnd) to every jack , right?

[Rob Strand]

Anyways, what was the purpose of a shorted/switched connector originally used there? How do you think DIN connectors could be wired originally in my unit? I think it was a pseudo 3 prong connector by connecting "-" lug to gnd lug..

It's very common to short unused inputs at the socket to prevent noise pickup.
(With some layouts / designs it's possible shorting the input increases the hum but even then it's still shorted.)

I see the sockets wired like this, which looks 100% OK/normal.

[sarakisof]

Ok seems the chinese mono switched jacks I had in hand and compared to, get the tip & sw lugs reversed  that's why I was confused.



Makes sense now.
So, if I wire those plastic insulated jacks previous guy had installed there, like below, I'm back to original wiring I guess right?

[Rob Strand]

Ok seems the chinese mono switched jacks I had in hand and compared to, get the tip & sw lugs reversed  that's why I was confused.

Yeah, crap like that does your head in  .   I *never* trust anything to do with pin outs and mechanical parts.   As it happens I have a couple of old switched mono sockets on my bench and they look like the ones in the unit (ie. the pic I posted).

Makes sense now.
So, if I wire those plastic insulated jacks previous guy had installed there, like below, I'm back to original wiring I guess right?

To first order yes.   Just use the switch on the tip of the plastic sockets.

But beyond that is a big can of worms:

Where things get different is the original sockets don't isolate the (socket) ground from the chassis whereas the plastic units do.

With single ended audio there's also sorts of hum and noise issues related to grounding sockets.   There's also a number of variations to the wiring:  ground at both ends, ground at the socket end, ground at the PCB/circuit end. 

In short I find the plastic sockets much less problematic for ground loops.  However this requires the ground to be connected the PCB.   However if there is no ground at the PCB/circuit end of the coax you will need to ground somewhere, perhaps at the single ground point show on the left of your pic.

Notice that as shown there is a single ground point where the ground bus wire connects and that the sockets also connect to ground.

At minimum each input socket needs to have a ground path back to the circuit.

Beyond that is trying to come up with best grounding which avoids ground loop issues *inside* the unit and *outside* the unit when external devices are connected to the unit.

I have no idea how the grounds are connecting at the other ends of the coax in your pic.

If you don't what to think or experiment with this stuff too much perhaps try to copy what the original did.   From this perspective I think you will need to run a ground bus wire across all the sockets then back to the single ground point.   Just do that and leave it there - unless you have problems.

[PRR]

get the tip & sw lugs reversed

What, the plastic "Marshall" jacks?

If you take the metal strips out, and swap them left to right, tip and switch are reversed.

This is a subtle point (I never noted it before). Probably some one guy once got it different, his customers said "做乜鬼???!!", and learned to swap and swap again, or stare at how the metal strip actually switches, until happiness happened.

[duck_arse]

If you take the metal strips out, and swap them left to right, tip and switch are reversed.

I;ve done this. the locking pip on the lug destroys the plastic it needs to hold the lug in place, on the way out. then when you plug in a plug, the finger slides and looses grip.

but on jaycar jacks, the plastic melts as you solder - then you throw it away.

[sarakisof]

Where things get different is the original sockets don't isolate the (socket) ground from the chassis whereas the plastic units do.

Sure. Forgot to mention I will connect each plastic socket's sleeve lug to both  coax cable's shield and ground bus wire for every socket.

If you look carefully there was actually a gnd bus wire originally there with brass big rings running across all the sockets (for the original DINs sleeve) and some small solid wires on each ring probably for soldering them to DIN's sleeve lug. Some of them saved from previous owner's brutal job..




So if I go like that for every socket  I'll be back to original wiring, right?

[Rob Strand]

but on jaycar jacks, the plastic melts as you solder - then you throw it away.

When I was a kid I'm sure I heated up connectors well over the intended time and they still survived.   The plastic was really hard almost like Bakelite and all the riveting was tight.  Nowadays I'm almost hesitant to solder plastic connectors.