[Amps] GS35B Amps Common Problem / few more comments

Thu Oct 31 09:04:48 EDT 2019

Tony King has comprehensive data on this tube on his webpage.  It is interesting to note that the data sheet actually calls out a degradation of performance of the tube running on a frequency of 60cm.  I assume running 24/7.  The data shows that at a anode V of 2500 volts and anode I of .7A a new tube will make 800 watts output.  Then after 500 hours of use it will drop down to 650 watts.  Although this is for 60cm and not HF frequencies, does a similar degradation take place on HF?  Maybe, and it may be a longer period of time however and going above the 2500v anode voltage, it may add less time to the Po.  I know many hams pushing these tubes to a plate voltage of 4KV and a bit higher to match the performance of an 8877.
Many years ago I was told that the GS-35b was "the poor man's 8877", indeed the $135.00 price for a new one was attractive compared to that of an 8877 so I tried one in the same setting as an 8877 for a comparison.  All things being equal I took an Ameritron Al-82 and I retrofitted in a GS-35b a little tweaking or the input and I had a perfectly working unit with a GS-35b.  Having the pleasure of also having an Al-1500 on hand I made comparisons on band and  power in vs power out.  I was sadly disappointed when I realized to get the same performance from the GS35b vs an 8877 I would need more than the 100 watts I had available for test purposes.  In retrospect, the GS-35b had a bit of an edge over the pair of 3-500's and in comparison one could say that the GS-35b had it over the 3-500's  for all things being equal.  Comparing to the 8877 was no contest.  
The Ameritron amps in question all run plate voltage of 3600v.  Price wise the tube is very attractive and I can see the affection for it however I myself stayed with 8877's and never looked back.  

-----Original Message-----
From: Leigh Turner <invertech at frontierisp.net.au>
To: 'Mike' <k4gmh at arrl.net>
Cc: amps <amps at contesting.com>
Sent: Thu, Oct 31, 2019 3:27 am
Subject: Re: [Amps] GS35B Amps Common Problem / few more comments

Bottles like the 4-1000A thrive on such high plate voltages, but the
internal close-spaced electrode construction of the GS35B (designed for UHF
service) with its oxide coated indirectly heated cathode is not conducive to
robustly withstanding such high plate potentials. As Ian GM3SEK mentioned
these frangible oxide particles can find their way onto the proximate grid
structure and modify tube behaviour, as well as cause potentially damaging
internal arcs.

I forgot to ask, but presume that your filament voltage is precisely 12.6V
and is stable under line and load variations when measured at the socket?
The GS35B does not tolerate either under or over filament voltage and has a
Goldilocks factor sweet-spot for its emission. 

While a buck xfmr or heavy duty Variac placed on the HV xfmr primary winding
are messy, its one of the few practical ways to reduce the P.S. output
voltage if there are no primary or secondary winding taps available. Presume
the secondary feeds a traditional FWB rectifier.

Leigh

VK5KLT

  _____  

From: Mike [mailto:k4gmh at arrl.net] 
Sent: Thursday, 31 October 2019 4:41 PM
To: Leigh Turner
Subject: Re: [Amps] GS35B Amps Common Problem

Hello Leigh,

Yes, your probably right.  The HV xfmr was for a 4-1000 amp.  However, this
is the only HV xfmr I have so gotta use what I got.  

Leigh, do you know any way of reducing the output voltage?  The only couple
of ways I know is to install a buck/boost type of xfmr in the 220 VAC lines
going to the HV xfmr, or inserting a 220 VAC 30 amp. variac in the 220 VAC
lines.  Any ideas on how to reduce the HV supplies HV to 3 KV will be
appreciated.

Thanks for your help, Leigh.  

  73,

  Mike, K4GMH

On Wed, Oct 30, 2019 at 6:42 PM Leigh Turner <invertech at frontierisp.net.au>
wrote:

Hi Mike,

In view of this experience, I'd suggest running the tubes with a lower plate
voltage circa 3 kV max if that can be easily facilitated.

I suspect the premature tube failures experienced now exhibiting unstable
emission / RF characteristics here are somehow related to and likely
instigated by excessive plate voltage....

Whatever internal structural changes that have occurred imparting the wobbly
behaviour must have been sustained cumulatively over time without leaving
any physical evidence in the glitch resistor, etc. from say individual
violent tube internal arc events.....something a fast-blow HV fuse can
mitigate the damaging effects of.

Leigh
VK5KLT

-----Original Message-----
From: Amps [mailto:amps-bounces at contesting.com] On Behalf Of Mike
Sent: Thursday, 31 October 2019 4:54 AM
To: amps at contesting.com
Subject: Re: [Amps] GS35B Amps Common Problem

Hello,

Swapped out the GS35B in one of the amps with NOS GS35B and the amp. works
as it did.  Why did the initial tube fail - don't know as HV supply, bias
system, filament voltage, etc. checked out.

I did see Ian's pictures and the same thing may have happened to the GS35Bs
in my amps.  Again, all the HV supply parts in the common HV power supply
checked out okay including the glitch resistor.

Any ideas on what caused the problem?

One thing I might do is to put a fast blow fuse in the HV circuit although
not sure a grid to plate arc caused the tubes to fail or even if the fuse
can blow fast enough to "save" the tube(s).

  73,
  Mike, K4GMH

On Sun, Oct 27, 2019 at 11:52 PM Mike <k4gmh at arrl.net> wrote:

> Hello,
>
> My two homemade GS35B amps exhibit the same symptoms of a problem and
> can't seem to figure what  is the "common" problem.  This happened
recently
> after several years of operating both amps without a problem in various
> RTTY contest running 1500 Watts output (SO2R).  The symptoms are seen
> initially tuning the amps up.
>
> The symptom(s) occur after the amp is warmed up, filament and blower both
> on for over 5 minutes before high voltage is applied to the amp.  RF is
> applied to the input and plate tuning starts.  The amp output gets to
~1200
> Watts output when the amp input impedance start changing causing the radio
> driving the amp to fold back its output which in turn decreases the amp
> plate and grid current.  The radio's output decrease continues until the
> radio is putting out very little RF.  The amp is correspondingly putting
> out very little RF.  The amp's plate current has gone from ~500 mA to less
> than 100 mA along with the grid current going from 200 mA to less than 50
> mA.  Once this condition is reached, removing the input RF from the amp
and
> immediately reapplying the input RF does not change the last condition of
> amp's input impedance.
>
> Usually a long period without the amp being keyed has to occur before the
> amp's input impedance seems to start returning to normal.  However, once
> RF reapplied, the amp immediately starts to return to the high input
> impedance.
>
> The amp plate voltage goes from ~3400 Volts at 1200 Watts output to ~3600
> Volts at lowest output.  Quiescent plate voltage is ~3700 Volts.  Note: a
> single power supply is used by both amplifiers along with a common blower.
>
> The things check and found okay are the W4ZT bias circuit in each amp, 120
> VAC to each amps filament xfmr, power supply's start up resistors
by-passed
> after HV power supply start up, HV diode string, different radios used,
each

> amp uses a separate antenna, etc.
>
> Again, these amps have operated okay for several years without an issue.
> Then both amps start exhibiting the same symptoms of a common(?) problem
> at the same time.
>
> Appreciate any ideas on where to look next for the problem.
>
>  73,
>  Mike, K4GMH
>
_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps

_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps