Blue glow 2A3's

[Tom-s]

So. I bought my first few old 2A3's for SII (Finally!  ).
These tubes all test +-50-60% on my EICO tester (--> FAIL!  ). Sovteks test 100+.
In the EICO, they do OK, no shorts, no signs of gas *(that last is a simple test of emissions with grid connected to ground vs floating). Now these exhibit blue glow around the internal elements, a very whitish-blue, when cranking the plate voltage a bit.*

They sounded good (Even while "cold" for the short time minute I had the amp on for these pictures)!

Are these safe to use? All have the same results on the tester. Only one pair tested in SII.
Can i try "burning in" these? Any suggested method?


* Edit. I found my EICO to be very conservative on the tube when testing. When in doubt i stress the tube to see how it behaves.

[Paul Birkeland]

That blue glow doesn't look problematic.  Those tubes probably have some life left in them where they will still sound decent.

[Tom-s]

Thanks PB. Still had sweaty hands during the second power on. 
No arcs, no fire, house still standing. These are quiet and sound good. Upped the volume for a while for bigger grid swings, all is fine.
Will now leave the amp on for the next 6 hours and see.

Next question. 
Would dropping the plate voltage and biasing for 30-35mA have any negative side effects on tube life from these bad testing specimens? I figured this might improve the "safety margin" a bit. 
Essentially going 45, without the tube's (as i don't have any). And yes, won't forget to stay with the normal heater voltage.
_{Why? Was planning on this before I found these tubes.
I hope that upping the choke's capacity would help to bass a bit (with my speakers, that is). Even when the amp is going to achieve 1.5-2W maximum. The lower plate resistance of the 2A3 vs 45 makes a better damping factor with the same output transformer / load (to hopefully work with my speakers for even better bass)}.

Another question. To learn what's at stake here. 
I was most afraid of an arc-over when I saw this blue glow. What would die first if they did? I figured the output transformers wouldn't like it and be first to go? Parafeed or interstage cap next? And with bad luck speakers / tweeters?
As the current flows through the tube, the parafeed capacitor, the output transformer, and back to the tube if i'm correct.

[Paul Birkeland]

When a tube dies a slow death, it's from the filament wearing out.  You just don't get enough emission.  These are very unlikely to cause any damage to your amp when they are exhausted, and you will find that they will start to sound absolutely terrible but will still work.

I would not recommend modifying the operating point. 

[Tom-s]

I would not recommend modifying the operating point.

Thank you PB. Could you please explain. I thought it was perfectly ok to use a 2A3 at 45 OP?
My inspiration was another HIII owner with a Stereomour. https://forum.bottlehead.com/index.php?topic=5388.0
Do these "bad" tube adversely affect such a change in OP? Or vice versa, would the change in OP have a negative effect on the tube (as was the original question)?

[Paul Joppa]

The two main things that wear out a tube are heat and plate voltage. Reducing the current will reduce the temperature (less plate dissipation) which is a good thing. But it won't reduce the plate voltage - in fact, it will slightly increase it as the load on the PSU is reduced. That's a risk.

A better bet for increased longevity is to reduce the plate voltage. This will also reduce the current, without changing the cathode resistor or the distortion. If you drop the plate-cathode voltage to 250v, the current will be about 40mA and you can swap in the higher inductance tap on the plate choke to get the deeper bass. The PSU resistors will get pretty hot, so make sure they are rated for the additional dissipation!

[Paul Birkeland]

A better bet for increased longevity is to reduce the plate voltage. This will also reduce the current, without changing the cathode resistor or the distortion.

This also comes with reduced output power (just over 2W).  For the Stereomour I, it was OK to just run the #45 operating point and take reduced power if you happened to stick a 2A3 in instead of a #45.  The changes in the S2 to run the #45 instead of a 2A3 are a bit more involved, so you will not be in the position where you can swap between those tube easily, hence my recommendation not to change just the operating point.

[Tom-s]

Thanks guys. Helped me a lot!
Made the adaptations to the PS and changed the 130w resistors for 1240 ohm in total.
These are only rated 5w (bought with lower current in mind). So with 2,2w on the 680 ohm ones (placed directly under the vents because of this), i’m very aware i’m pushing the limits here.
Anyhow. Again no fire. 306v on top and 50-51v at the cathode make it close.
With 42-42,5 mAdc through the chokes I think it’s a bit much. Or would this be OK to try?
Every extra 130 ohm would drop +-7 volts on the plate. So could drop 14v extra with the spares.

Edit: this is with 2.5uf parafeed caps.
Edit: am using one 130r resistor to try and match exact OP to keep some headroom when tube rolling. Above measurements were with Sovteks.

[Paul Joppa]

When I tested the chokes, my setup for the full winding ran out of steam at 42mA and the chokes were still had a bit more to give. So 43mA is probably fine.

[Tom-s]

Thanks Paul. I started this all thinking the PC-3 would saturate around 30-35mA based on forum searches. So even with 2A3's i'd had to get as close to that as possible.
Now i'm using 43mA* without any problems whatsoever. Bass changed for the better using the full choke.
Thank you for your help.

*Edit: Yes 43mA, line voltage variation got me to 306V on plate, 52V cathode (even with the extra 130 ohms resistance, so total 1370 ohms resistance vs normal 260 ohms / and the extra resistance from the full choke).