[ARC5] BC-348-R Update

[Ben Hall]

On 8/20/2022 8:23 AM, David Stinson wrote:
> On 8/20/2022 7:45 AM, Ben Hall wrote:
>> 
>> In other news, I think I've got the SLR-M finished.  I've changed
>> the pilot light bias regulation scheme replaced with an LED and
>> zener diode.  
> 
> May we get some details on this, Ben? Don't want to take anything for
> granted. Thanks.

Hi Dave and all,

On the SLR-M, the negative bias voltage for the pair of 25L6 output tubes is derived by a unusual (at least to me) cathode resistor bias scheme.  Cathode current from the pair of 26L6's flows thru R-144 (36 ohms, 1 watt) which is in series with R-154 (3,900 ohms, 2 watts) to the B- bus.  Across R-154 is a 6.3 volt 150 mA rated pilot lamp.

I guess the idea is for the filament in the lamp to be a crude voltage regulator?

This setup had several problems in my opinion:

1)  I only saw about -6 volts of grid bias on the 25L6's.  The manuals seem to suggest -7.5, and I found one audio article where one of the posters had built a P-P 25L6 audio amp bragging that he was running the tubes hard at something like -3 volts bias.  25L6's are not getting any more common, so I don't want to run them hard!  Besides, a pair of 25L6's at like 1/3rd volume is as loud as I'll ever need.

2)  The pilot lamp is quite dim.  On my set, the pilot lamp was clipped to the top of the bracket, so that the bulb could shine down the front of the tuning dial.  I soon figured out that the SLR-M uses a somewhat translucent tuning scale, and that the bulb was likely meant to be clipped to the BOTTOM of the bracket, putting the bulb right behind the tuning scale.  However, with the bulb being dim and the scale being old and not as translucent as it once was, it was super dim.

3)  If the pilot lamp fails (which it likely won't ever, as it's being run so below it's ratings) the negative bias runs away.  I'm not sure if this will or will not hurt anything, but for sure, I can see myself down the road tuning the set on, never seeing the pilot light because it's burned out and I'm thinking I can't see it just because it is so dim, and not getting any output then troubleshooting everything else but a burned out bulb.

So here's what I did:

Years ago, I bought some "warm white" LED's from the e-place that were 10mm in diameter.  I designed and 3D-printed a fake bayonet "socket" that would hold the LED and the LED's dropping resistor.  This way, nothing is modified.  I even indexed the whole mess so it oriented the LED correctly.  ;)  I can't recall the value of the dropping resistor, but it was such that I limited current to something like 5 or 10 mA at 6 volts.

Even at 5 mA, the LED is much brighter (and cooler running) than the filament lamp.  I also "scratched up" the plastic case of the LED with a stainless steel brush to disperse the LED's illumination a little bit better.

I then tacked a 5 watt Zener diode, of value I can't remember, but likely something like whatever standard Zener values are around 7 or 8 volts, across R-154, the 3,900 ohm 2 watt resistor.  I just tacked it on top of the terminals for R-154, so nice and removable.  I did have to experiment a little bit with the Zener value, I think I originally chose 5 volts and that was too low.

I can't find my notes on all this, but I did do the Zener disspation calculations and it's well within 5 watts.  (as a sanity check, it's across R-154, a 3,900 ohm 2 watt resistor...)

The whole thing works pretty nicely.  Bias on the 25L6's is now something nice like -9 volts.  :)

Thanks much and 73,
Ben, KD5BYB