LA2A meter lamp to LED conversion

[jwhmca]

Burnt out lamps again... I'm really thinking I should just replace with LED. I did some research on it and have a couple of questions.

Since it's AC I know I can just use an LED and resistor, but wouldn't that allow DC back into the transformer? Is that a problem?

Better way would be to strap a diode across it... but having a pulse on the line cause any issue with the heaters? Audible noise?

 

[JohnRoberts]

If you power more than 1 LED you can alternate polarity of LEDs and steering diodes to balance your chi and not disturb the force.  Also LEDS will probably draw less current reducing system noise.

JR

 

[jwhmca]

If you power more than 1 LED you can alternate polarity of LEDs and steering diodes to balance your chi and not disturb the force.  Also LEDS will probably draw less current reducing system noise.

JR

Like this?

 

[JohnRoberts]

works for me....  8)

JR

 

[fpliuzzi]

An alternate idea, to the perfectly good one that you have drawn in your diagram, is to just use two LEDs placed in reverse-parallel and leave the two diodes out of the circuit.

The pair of anti-parallel LEDs will light on alternating half-cycles of the AC sinewave from your transformer's secondary. The current will be shared between the two LEDs so you'll have to reduce the value of the current limiting resistor by half if you want each LED to get its full 20mA.

Also, sanding the LEDs with 400 grit paper will allow the LED light to be a bit more diffuse.

Regards,
Frank

 

[JohnRoberts]

There are bi-color LEDs packaged that way but for White LEDs you might want to check the specific data sheet. I recently did a little white LED project and two of the data sheets I collected (LUXEON) specifically say they are not designed to be operated with reverse voltage. OTOH an Everlight white LED shows a roughly 5V zener when driven in reverse.

My guess is that two Everlights would work anti-parallel since the 5V zener is well above the nominal 3.6V forward voltage at 20 mA.

I generally try not to ignore manufacturer's data sheets when they say not to do something I usually don't.

Diodes are cheap. At least check the data sheet for the LEDs you plan to use.

JR

 

[fpliuzzi]

JR, am I missing something? The second LED that is wired in reverse-parallel to the original LED also happens to be a diode that conveniently emits light. The pair of anti-parallel LEDs will light on alternating half-cycles of the AC sinewave. Each LED should protect the other from reverse voltage.

I have seen this technique used successfully to replace lamps in recording consoles, etc. I've used it myself on various pieces of vintage audio and test gear.

Respectfully,
Frank

 

[JohnRoberts]

JR, am I missing something?

my words...?

The second LED that is wired in reverse-parallel to the original LED also happens to be a diode that conveniently emits light. The pair of anti-parallel LEDs will light on alternating half-cycles of the AC sinewave. Each LED should protect the other from reverse voltage.

Umm not exactly... LEDs generally exhibit 2+ volts or so in the forward direction. So two LEDs connected anti-parallel, one will see +2V forward, while the other is seeing that same -2V reverse. This is generally inconsequential, as i mention they sell bi-color LEDs in 2 lead packages hard wired this way...

I have seen this technique used successfully to replace lamps in recording consoles, etc. I've used it myself on various pieces of vintage audio and test gear.

Respectfully,
Frank

As i posted I suspect this would work for the Everlight white LEDs whose data sheet spec 3.6V (max) forward voltage at 20 mA and 5.2 V (min) reverse zener voltage. Since back to back the -3.6V across the reverse biased LED is less than the -5.2V zener voltage.

But as i also posted I have two data sheets that say their white LEDs (high output) are not designed to be operated with reverse voltage bias.

BUT I ALREADY SAID THIS... does repeating it make a difference.?

I try not to give advice that may not work, since it irritates the OP and embarrasses me. 

I might as well repeat my closing comments.

Check the data sheet for the actual white LEDs you plan to use. See what they say about reverse bias...  If the LED can handle more voltage in reverse bias than forward, the back to back topology will safely clamp the reverse direction.

Diodes are cheap... So the first approach will work for any white LEDs as the reverse voltage is clamped at roughly .5V

If building a large number of these the approach using no diodes will be cheaper, as long as it works.

I do not know why one manufacturer advised against reverse bias. I suspect they have a reason, but I try not to ignore the manufacturers advice.

Feel free to ignore me... my advice is cheap.

JR

 

[jwhmca]

In calculating the Current etc... would I use 6.3as the volts or because it is AC some RMS number?

 

[dfuruta]

in the la2a, the lamp is also serving the purpose of regulating the voltage for the meter section;  if you replace it with leds, you might want to also put a zener or something in parallel to do the same.

 

[JohnRoberts]

I'm not sure how the lamps would regulate the voltage, perhaps their current draw is anticipated in the nominal supply voltage.

If the LEDs draw less, perhaps an extra power resistor to soak up some current?

Regarding your resistor value I don't know. Depends on forward voltage of the specific LEDs and brightness of them, so you may need to experiment some to determine a final value. 

As a starting point 6.3 AC, - (3V+ ? look up on data sheet for specific LED and current + diode drop (0.5V) if using my suggested approach (for back without diodes don't add the diode drop).  is the voltage across the R  So this V/R= current. You want this continuous current to be 2x the current for each LED since they are only conducting alternately half the time each. So 40 mA total will deliver 20mA ave to each LED.

So something around 100 ohms as a starting point.  Or not...

JR

 

[Kingston]

in the la2a, the lamp is also serving the purpose of regulating the voltage for the meter section;  if you replace it with leds, you might want to also put a zener or something in parallel to do the same.

There is no "regulation" of any kind by the bulb on AC heaters. Or DC for that matter. Incandescent bulbs eat raw current and that's it. The amount of current drawn by any common VU meter bulb (somewhere around 100mA) is negligible compared to all the tube heaters. Which themselves are plain incandescent bulbs, by the way. You may see very minor voltage change on the AC heaters with the bulb removed. It has no effect on the lifetime of the tubes whatsoever.

That said and getting back on the original topic, I also modded my LA2A's to LED's quite recently. Got sick of changing the bulbs, too. No noise removal or protection tricks for me, just a bunch of LED's and resistors. Specs did not change one bit so feel free to install whatever configuration you want without worry. It's just the heater line we're talking about.

[edit]

I just realised, dfuruta you might have confused an important fact. There is indeed a neon bulb regulating the 50-80V DC voltage need for gain reduction metering. This has nothing to do with the VU meter light bulb and people have been replacing it with a roughly equal zener diode for decades now.

 

[JohnRoberts]

As usual I post too much about products I am not familiar. FWIW I answer in general terms.

JR

 

[dfuruta]

I just realised, dfuruta you might have confused an important fact. There is indeed a neon bulb regulating the 50-80V DC voltage need for gain reduction metering. This has nothing to do with the VU meter light bulb and people have been replacing it with a roughly equal zener diode for decades now.

Whoops!  You're right, I was thinking about the wrong bulb.

 

[byoung]

My La2a meter lights have both burnt out again and I'm out of replacements. I wanted to try the anti-parallel LED technique. I was thinking 2-3 LED's per meter. They are 2.1v LED's and are rated for 5v reverse voltage. I have a couple of questions, would I need the resistor if using 3 LEDS? How about the diodes? Should any diode work for reverse polarity protection or are there ones better for this scenario? I have both zener diodes TZX7V5D and 1N4001 on hand.