Hey Larry...If what you do(&how u do it) makes you happy more power to
you....That`s why we enjoy our freedoms here in the good ol USA..73,Jim...
----- Original Message -----
From: Larry
To: k7rdx
Cc: Gudguyham@aol.com ; amps@contesting.com
Sent: Sunday, January 28, 2007 10:09 AM
Subject: Re: [Amps] Bad tubes, GS-35B's, and 8877's
We seem to be straying in to philosophy and design methodology here, with
a little side trip into "specmanship".
Let me address some points brought up by Jim and Lou:
I got into the russian tubes late in the game. I didn't think much of them
early on. I still don't. I acquired some and built some amps more out of
curiosity then need. My hobby is amateur radio with a sub-set of that being
the design and construction of "power amplifiers". (also antennas but that's
another story). With over 30 years experience in the design and test of
weapons systems, I think I have a fair idea how to design and test "fail
safe" circuits. It ain't any fun for me. Being retired now and being able to
do what I feel like doing, I refuse to waste what little time I have left
messing with protection circuits. So if a tube requires a "protection
circuit" of any kind, I am not going to use it. 8877's require protection
when used the way I would use them, russian triodes do not. the choice is
clear.
Regarding life time of the tubes: The "hours of operation" number is
misleading. Weapon system components are spec'd differently than, for
instance, broadcast or commercial components. Without knowing the details of
that specification the number is meaningless. The "500 hrs" russian tube may
very well have a longer usable lifetime than a "2000 hrs" Eimac tube, or it
may not. There is no way to tell without comparing the actual specification
and how it was developed. One thing I do know; hardly a week goes by that I
don't run across someone looking for an 8877 to replace the one that blew up
in their "protected" amp. I have yet to run across anyone who has actually
killed a GS-35B while operating. Some years back, Eimac distributed 8877's
with manufacturing flaws. Those tubes had an operating lifetime measured in
minutes. What guarantee do we have they won't do something like that again?
The "good ol' boys" at Eimac are long gone, nothing left but money grubbing
accountants and yuppie managers. You want to depend on them?
Regarding "matched" tubes: I don't have enough GS-35's to form a
conclusion. The two good ones I have happen to be pretty close to one
another. I have a pile of GI-7B's that I got for the 432 amp project. Since
I used two in parallel in a strip line I went through the trouble of testing
and matching them. The variation across the batch was not that bad. In fact,
I've seen more variation in the same sample size batch of 4CX250's. In
actual operation, there was little difference between the best matched pair
and the worst matched pair. I have no idea if this is typical or not or if
it applies to GS-35B's.
Tube arcs: For years I thought "big bangs" were necessary if you wanted to
run power. Explosions and fires were a normal part of operating. Bits of
metal stuck in the wall board, molten balls of copper rolling across the
floor, holes burned through the sides of the chassis, yeah, those were the
days! I had one friend who refused to operate my station unless I provided
flak jacket and helmet. Witness names on request. In spite of all that I
never lost a tube due to a big bang. Lots of chokes, resistors, diodes, fuse
holders, circuit breakers, transformers, etc., but not ever a tube. Once I
figured out how to fuse the anode supply, a lot of the excitement of
operating was lost.
Anode voltage: The GS-35B is spec'd at 6KV max peak and 3 KV operating, at
1000 MHz. Again a weapons system spec and meaningless without the entire
specification detail. For HF and 6 meters, I would feel pretty comfortable
running it at 4 KV maybe even 4.5KV. After that you would be eating into any
safety margins there might be and I wouldn't be happy with it myself. With
4KV under load, the power gain in a HF grounded grid configuration is
getting close to that of an 8877. Yes, there is variation among tubes, but
on the average it's not bad. Of course if you only have 50 watts of drive
available and need all the gain you can get. The thing I've noticed about
the russian triodes is that the output compression point drops quickly as
the anode voltage is dropped and that can give a false impression of low
gain.
Power gain: If low drive is your goal, you should be using grid driven
tetrodes instead of grounded grid triodes. Personally, I got bored designing
and building screen grid supplies and protection circuits and gave up on
tetrodes a while back. I should re-visit tetrodes, but I digress. In a
grounded grid configuration, there are several factors (besides tube Mu)
that contribute to apparent power gain. From what I've seen on the web, most
russian tube amp builders ignore the importance of the cathode circuit and
the output compression point. The efficiency of the cathode network is
extremely important. A half dB of loss in the cathode network translates
into a LOT of watts at the output jack (assuming you are not into output
compression). As I've said before, the russian triodes seem to go into
output compression a lot faster as the anode voltage is dropped than an
Eimac tube. Tuning for max output and measuring power gain is dead bang
WRONG! In order to properly evaluate tubes and circuits (in a linear
amplifier) you need to be measuring the 1 dB compression point and gain at
that point. Also bias point can affect the power gain a lot. Increasing the
idle anode current can sometimes increase gain faster than lowering
efficiency resulting in a lot more output power.
I'm cheap (or efficient depending on your viewpoint). Most of my
amplifiers cost less than 100 bux out of pocket. The first time I went over
that limit was the YC-156, about 50 bux invested in the chassis stuff and
350 in the tube. (power supplies not included in these numbers) My 432 amp
was a bit over 100 bux, 50 bux for a pair of "new" GI-7B's, 20 bux for the
bazillion 4-40 screws, 20 bux for the copper sheet anode line, and a small
pile of cheap "junque box" components and scrap aluminum sheet. The 6 meter
amp was about 40 bux not counting the GS-35B. My preference would be to use
"american" tubes but I don't have a cheap source for them. If I could get
good used 8877 pulls for 50 bux a pop I would use them. If I had to buy a
new tube, it would be a 3CX3000 or 3CX6000. Far better value for the bux
than one of them puny 8877 thingys.
73, Larry
Larry - W7IUV
DN07dg
http://w7iuv.com
k7rdx wrote:
I built my first Russian tube in 1998 (GS-35b) after a ham friend
swapped me a tube at the Puyallup electronics flea market (In the Seattle
area) I was quite impressed at the time with how well it worked on
HF..True,it`s a medium mu tube&requires a lot of bias, but with 100 watt
exciters legal limit is easily done.Now if you do the research you will find
the tube is designed for 500 hrs operation... The Eimac tubes are rated in
the thousands of hrs...But the price is right for the Russian jugs, at least
until the supply dries up(if you get good tubes!.)..I don`t understand why
you fear the 8877 Larry..If you design&build with protective circuitry/grid
trip/plate glitch protection,etc how can you fail? I consider the quality
over the quantity and prices are going down for 8877`s as the surplus supply
goes up too....How many Russian tubes are going to match (Each others)
specs&performance? Not too many from my experiance,you will be hard pressed
to find two GS-35b`s that match each other fairly close,not a real big deal
unless your amp uses a pair of tubes...The guys that run much more than
3.5kv on the Russian tubes? Big chance of tube arcs no amount of "Nichrome"
can cure. Just my opinion from personal experiance, for what it`s worth
73,Jim..
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|