Wiring In Indicator LEDs Under the Tubes

[karl]

I know that this has been covered before, but I need some specifics that I don't think were covered previously.

Is there a specific size of red LED that will fit the center post of the tube socket, per paul's suggestion? If so, what resistor should I use? Are they available from Radio Shack? To which lugs on each tube socket should I connect? I believe lug 6 was mentioned but not another.

Regards,

Karl

[Paul Birkeland]

It's not that easy to retrofit IMO, I think you'd have an easier time starting with this switch:

http://www.digikey.com/product-detail/en/M2043SS1W01/360-1876-ND/1006954

You can wire it up the same as the stock switch, then for the fourth section that's extra, connect one end of your LED dropping resistor between the red wire leaving the 9V bank and going to the center lug on the 9 lug switch, and the other end of said resistor to the center lug of the 4th set of terminals on the new switch.

From the extra "lower" lug on the new switch, wire to your LED's in the sockets (I believe I used T1 size last time I did it with some silicone adhesive), and the grounds of the LED's can be wired to the ground bus or terminal 3.

You can use the stock switch, but you have to feed from the junction of the two 4k resistors at the 9 lug switch, which will be somewhat cumbersome, especially since you need the LED in there.  You might have a better time putting the resistor between ground and the LED, since there is room around the ground buss routed with the RCA jacks.

For the current set resistor, you will want to use as little current as possible, I'd start with 2mA and see how that works (18k -1/4 watt).

-PB

[karl]

PB -

That's what I was looking for.

Thanks,

Karl

[karl]

PB -

What about wiring a single LED in series with the 36V supply?

Regards,

Karl

[Paul Birkeland]

What about wiring a single LED in series with the 36V supply?

You can insert an LED in series with the 36V supply and the audio circuit, but this comes with the trade-off of losing a little bit of your available B+ to do so, as well as having the signal current variations passing directly through the LED.  For something with a very flat Vfd vs. I curve, this isn't that big of a deal, but if you use an LED that has significant voltage variations depending on current, this is not advisable. 

-PB

[Paul Joppa]

The series resistor at 2mA will cut the battery life in half. I'd try the LED between the battery pack positive and the switch - you'll lose a couple volts as PB said, so the end point will come a bit sooner, maybe 3/4 the lifetime before you get down to 24v.

If you have NiCd or NiMH, which have a constant voltage, you can use a red LED in place of the cathode resistor and bypass cap. It will likely even sound better, but check the plate current. The LED is a fixed voltage, while the stock resistor drops 1.2 to 2.0 volts as the alkalines run down. For a 30-v B+ (four 7.2v NiCds) the bias will be about 1.6v; the HLMP-6000 that we use everywhere in C4S boards drops 1.57v usually.

There are several different chemistries for "red" LEDs, all of which will have slightly different voltage drops; a larger chip size will drop slightly less voltage than a small one in the same chemistry. So you may want to experiment...

[badman]

If you have NiCd or NiMH, which have a constant voltage, you can use a red LED in place of the cathode resistor and bypass cap. It will likely even sound better, but check the plate current. The LED is a fixed voltage, while the stock resistor drops 1.2 to 2.0 volts as the alkalines run down. For a 30-v B+ (four 7.2v NiCds) the bias will be about 1.6v; the HLMP-6000 that we use everywhere in C4S boards drops 1.57v usually.

I've had good luck using the low-noise HLMPs in place of cathode resistors and just ordered up a few to do this, my impressions are usually of greater dynamics and openness.  Is the reason for only doing this with constant-voltage batteries that the led will stop conducting (and the tube, with no cathode ground reference) if the alkalines fall below the 1.57?

[Paul Joppa]

The bias with the LED is fixed, but as the high voltage batteries drop, the ideal bias changes, from 2.0 down to 1.2 as I said (not very clearly) originally. So with LED bias, if the voltage deviates from 30v then the operating point is off-optimum. Specifically, the plate current drops rapidly as the supply voltage falls, increasing distortion, reducing maximum output, and reducing available output current.

[badman]

The bias with the LED is fixed, but as the high voltage batteries drop, the ideal bias changes, from 2.0 down to 1.2 as I said (not very clearly) originally. So with LED bias, if the voltage deviates from 30v then the operating point is off-optimum. Specifically, the plate current drops rapidly as the supply voltage falls, increasing distortion, reducing maximum output, and reducing available output current.

Yes- that's what I meant, thanks for confirming.  I have some 9.6V high capacity (and low leakage) NIMH on the way, so a nice flat curve at a little higher voltage should keep the (initially) lower bias in its happy place for this experiment.  The higher voltage should also mean the PJCCS, when installed, has plenty of compliance.