[R-390] Found something interesting about tubes

Barry Williams ba.williams at charter.net
Tue Dec 4 21:39:07 EST 2007


Anybody remember Nolan Lee's R-390As that ran 24/7?




TUBE ROTATION AND REPLACEMENT

Date: Fri., 15 Jan 1999 22:34:22 -0600
From: Nolan Lee <nlee at gs.verio.net>
Subject: Re:  [R-390] MTBF (was tubes and 3TF7 ...)

Every six months, I pull the tubes out of the R-1051B's and install new 
ones. They'll last longer than that but I've got plenty of them. I think 
that I paid .25 a piece for the 6AN5WA's when I bought them and only 
about .70 each or so for the 6BZ6's. At the current rate of consumption 
I think that I have enough for 30 years or so. :-) It costs me about 4 
dollars a year to retube both receivers. Cheap, huh? And, yes I can 
notice a difference in sensitivity before and after retubing. The 
performance difference and the low cost of the tubes is the reason that 
I don't even bother to test them.

The antenna couplers, have a switchable meter on the front panel that 
monitors the ten pairs of 6922's in each one. When I see the numbers 
drop to a level a little below the recommended reading, for one pair, I 
replace that pair. I used to just replace all of them but my supply of 
good quality Amperex 6922's is getting low and they now cost more than 
the coupler is worth thanks to the golden eared audio crowd. When 
they're finally gone, I'll have to start using the Jan 6922's but in the 
mean time, I'll cook Bugle Boy's in the couplers.

I've make little cards up that I keep on each of the receivers. On the 
card is the testing date, the location (V#) in the receiver, the SN of 
the tube tester, and the tested value of the tube expressed as xx/yy, 
where xx is the tester reading and yy is the minimum value of the tube. 
I try to test the tubes about every six months. After a couple of 
decades, you see trends. It's been my experience that for the R390A's 
the tubes that seem to have to be replaced the most often are the 
26Z5W's, and the 0A2WA's. I typically let them run 24/7. I don't trust 
the regulator tubes as far as I can throw them. I hate the damn things. 
Anytime I install new rectifier tubes in an R390A, I select and install 
a new regulator tube. Amazing how many are a couple of volts or so 
"off". I've also had "noise" problems in the receiver that I've traced 
to them. It's amazing the difference in sensitivity the tubes in the RF 
section make as they start to deteriorate. Since it's a gradual thing 
and you don't notice.

About the least trouble that I've had is the 6 or 7 dozen or so tubes in 
the old Tektronix scope and the CA plugin that's usually in it. Over the 
last twelve years or so, I've changed a set of 5 of the 5642 HV 
rectifiers, the 0G3 regulator and a pair of 6CW5's that come to mind. 
Very well engineered piece of equipment. As long as the tubes aren't 
noisy or shorted, they'll usually work fine in it even when their test 
values are well below the normal "worn out" numbers that the tester 
manuals list. I've had MORE transistors fail in the 547 than I have 
tubes fail. Finding matched pairs of those old early 1960's transistors 
can be a real pain, too. No BA collection is complete without a hundred 
plus pound tube type Tektronix scope setting on the bench sucking up a 
thousand watts or so. They're magnificent pieces of gear that are first 
class examples what American industry was capable of producing at one 
time. Quality engineering, craftsmanship, and best quality materials 
available, but mainly, when people took pride in their work.

Tubes, as a whole, are a hell of a lot more reliable than a lot of 
people believe. There were some really suck designed tubes/applications 
in the 1970's, such as those damn compactrons used in televisions and 
used in applications like RF amplifiers that they really weren't really 
up to. Ditto to the tubes in the power supply for the IP-173 series 
panoramic adapters. I've not real impressed with the life of the 6AK6's 
used in the R390A's either. They don't seem to last as long as they 
should. For what it's worth, the RCA's seem to last longer then the  
other brands. The 6DC6's on the RF deck are another one that should be 
replaced every 6 months of 24/7 in the R390A's if you're interested
in peak performance.

On a side note, try to never install a new 5749 in the PTO module of an 
R390A. Always use a good strong used one. The alignment will hold longer.

Just for kicks, measure the sensitivity of your receiver, remove and 
label the tubes on the RF deck and replace them with new tubes. Then 
retest the sensitivity. I'd almost bet money that you'll leave the new 
ones in there. :-)

Oh, about the only piece of gear that I have that gives a time frame for 
tube replacement is my AN/PRM-10 Grid Dip Osc. The 955 acorn tube in the 
head is supposed to be replaced every 300 hours. I was curious about 
this a while back, so I tested the tube then let it run for a couple of 
weeks and sure enough, the tube was flatter than hell. ;-) Which reminds 
me, I need to pickup a few more spares, I only have a couple of them 
left. ;-( Thinking of the GDO, 6X4 rectifier tubes don't impress me much 
either. I've went thru a lot of them in the GDO, the RC and RCL 
bridgers, the old CV-116 when I used it
(it has five of them in it), and the pair of them in the TV-2 tube 
tester. The 6X4 is one of my least favorite tubes.

One factor on "tube life" that's over looked is the application that the 
tube is used in. As an example, lets look at the 6AK6 used in the R390A 
as an example. There's three of them in there if I remember right. Two 
on the audio deck and one on the IF deck. The one used as the 4th IF 
amplifier on the IF deck seems to last the longest, followed by the line 
output amplifier and then the local audio amplifier on the AF deck. Same 
tube, way different "life". When I have to replace a 6AK6 used for local 
audio, I always install the new tube on the IF deck, and take that tube 
and install it in the line audio position, and then place the line audio 
tube in the local audio position. Hmmm, maybe we need a tube rotation 
schedule. <grin> Nolan


More information about the R-390 mailing list