The AM Forum
March 29, 2024, 01:32:38 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Hipot Testing of Transmitting Tubes  (Read 5811 times)
0 Members and 1 Guest are viewing this topic.
NW2K
Member

Offline Offline

Posts: 11


« on: June 06, 2019, 01:59:33 PM »

After buying a 3-500z that lit up at only 300v, I'd like to be able to do some tests on transmitting tubes before pressing them into action.  I have a 5kv DC hipot tester and parts to build a 12kv current-limited supply.  I've read some articles from Eimac and some hams, but I have not found a definitive procedure for hipot testing "amateur" vacuum tubes. 

Does anyone here have a general procedure for hipot testing our transmitting tubes? 

Should one test all of the different combinations of tube elements at the corresponding voltages suggested in the datasheet?  Or just measure leakage between anode and the other elements tied together at say 2x plate voltage?

How much leakage is too much and does the answer depend on which elements are being tested?  For example, I've tested several 4-125A between anode and the other elements tied together.  At 5kv, the meter indicated 50-100ua leakage.  Is that leakage a lot?  Thanks.

73, Dean
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #1 on: June 06, 2019, 03:48:50 PM »

Unless you have a spacecharge current from filament/cathode to plate, I don't see how a cold tube Hipot test accomplishes anything.

As far as the 3-500Z historically most shorts occur as filament-to-grid shorts and sometimes those shorts won't show up until after the filament is lit and you have plate current and the filament or the grid sags. I.e, internal heat causes the elements to sag and short.

It seems to me one would need something similar to the PDF file to incrementally test a Triode. Of course for a Tetrode, an additional power supply is needed.


Phil - AC0OB


* Hi Power Tube Testing.pdf (108.59 KB - downloaded 310 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
PA0NVD
Member

Offline Offline

Posts: 615


Nico and Chappie (Chappie is the dog...)


« Reply #2 on: June 06, 2019, 06:18:02 PM »

I do understand your reason Dean. I bought a whole lot of antique radio tubes from the years 20 and 30 and had the same problem.So I made a battery powered tester.
I think that you can make a tester with a powersupply from an old magnetron as HV supply. That is current limited 'And you can add the HV fuse as well.. Put an ammeter in the circuit.
For the filament, no other way than a filament transformer. For the grid, a 0 - 500V negative supply from an small old radio transformer. Start filament and HV supplied and with max negative and see if there is glow / anode current. If so, switch off and end of test.  It may cost you an HV fuse. If not turn down the negative until max anode current. No glow or strange effects, tube seems ok.
You can add one thing more, a resistor and switch in the prim of the filament transformer that lowers the filament approx 15 %. When at max anode current and the when you switch to low filament, the anode current should hardly be lower. When so, the tube may have low emission.
That is similar as I did for the old radio tubes which saved me a lot of disappointments...
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #3 on: June 06, 2019, 09:27:09 PM »

Hi Dean,

We must work with what we have available to us.... but the absolute best way to test tubes is to use a real RF working amplifier that has been tested with a new, good tube for IMD, power output, emission, etc.   These readings will then be compared to the questionable used tube operating data.

I once had a batch of ten used hi-mu triodes designed for linear service.  I tested them in a good working amplifier and took detailed meter, power and IMD readings using 2-tones and an SDR spectrum analyzer.  I was able to find the tubes with the best IMD and power output, though all tubes were close in performance because they were pulls that came out after the same amount of hours.

However, one of the tubes arced over at only 3KV when it was under full RF load, which was good to know since I was able to return it.  Using just DC static tests will leave data blind spots.  RF tests can show things that DC cannot.

IE, any improvised "tube tester" that is missing full stress, real RF operating conditions and careful IMD measurements of a real amplifier is lacking when evaluating used tubes.
For example, especially with external anode tubes where there is no view of the insides, if the grid is missing some of its structure due to an arc-over, this might show up as poorer IMD under RF operating conditions and invisible to DC static tests.  Same goes for plate hot spots where electrons cluster into one area, different from design, affecting linearity.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
NW2K
Member

Offline Offline

Posts: 11


« Reply #4 on: June 07, 2019, 11:13:58 AM »

Thanks to all for your help.  I've already built an apparatus like the one Phil shows and the others suggested and it gives good information.  It includes a supply for tetrodes and has a lot of metering.

I'd also like to be able to properly test for leakage/gas at HV, say 2x operating plate voltage.  This would be a preliminary/screening test.  Right now, I'm not sure of a protocol for the hipot tests.  Which elements to tie together?  What voltages to apply given those tying arrangements?  What leakage is acceptable in each test?  Having a known good/excellent tube is a good idea but not always possible.

I think a cold hipot test would give a quick indication of certain problems and the hipot tester has great metering and current limiting.  If a tube looks good after that fast test, it would then be subject to "hot" testing before being placed in a working amp/transmitter.

On the 3-500z's, it's gas that I'm concerned about.  If the hipot test shows a slightly gassy condition, I can use the above-mentioned apparatus to getter the tube.  Once that is done, hipot again, and if gettering was successful, subject the tube to RF and real operating conditions.

I've only seen a couple descriptions of hipot testing of TX tubes and it mostly had to do with huge bottles.

Thanks!  73, Dean
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #5 on: June 07, 2019, 09:44:08 PM »

Here,

https://www.w8ji.com/gettering_tubes_al572b_al80b.htm

W8JI goes through the process of gettering. The reason for running the plate red is:

Higher power internal anode tubes almost always have a gettering material coated directly on the anode. The getter acts as a sponge, soaking up any gas molecules inside the tube. This is necessary because any gas creates positive ions, reducing tube life. Not only that, even the slightest amount of gas greatly reduces breakdown voltage through a vacuum...The most common gettering material is zirconium. For example, zirconium is used on the outside of graphite or molybdenum anodes in 3-500Z and other tubes. It is the dull gray powdery or grainy texture coating you see on the surface of 3-500Z, 811A, and 572B anodes.

Zirconium getters best at about 1000 degrees C, this is why large metal anode transmitting tubes like the 4-400A, 4-1000A, and 3-500Z must be operated with a dull red to red anode color. Zirconium also releases some gasses and absorbs other gasses at various temperatures. The varying temperature across the length of the anode (and as the anode heats and cools) allows the gettering agent to absorb a wide variety of gasses.

The quickest way to ruin a 3-500Z, or other glass power grid tube, is to never show anode color over a prolonged period of time! Storing a 3-500Z for many years without operation almost guarantees a flash-over will occur at the first application of high peak anode voltage. 



Phil - AC0OB
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
NW2K
Member

Offline Offline

Posts: 11


« Reply #6 on: June 11, 2019, 06:10:40 AM »

Thanks again for your help.  Yes, I know about the gettering process, RF testing and emission/lifetime tests for those tubes, my question concerns hipot testing and proper protocol.  I've been relying on these linked Eimac documents, but as mentioned, they concern certain big tubes:

http://www.arimi.it/wp-content/Eimac/Hi%20Pot.pdf

https://www.cpii.com/docs/related/22/AB21.pdf

For now, I'll continue using the Eimac Application Bulletin 28 as a guide and I need to add some better metering at lower leakage current.  I'll try to report back on what I learn about all of this.  Thanks, 73, Dean
Logged
w4bfs
W4 Beans For Supper
Contributing
Member
*
Offline Offline

Posts: 1433


more inpoot often yields more outpoot


« Reply #7 on: June 11, 2019, 07:57:41 AM »

a sensitive current indicator is a neon bulb .... it ionizes at 80V and a few microamps yields a visible glow in a darkened room
Logged

Beefus

O would some power the gift give us
to see ourselves as others see us.
It would from many blunders free us.         Robert Burns
W1NB
Member

Offline Offline

Posts: 72


« Reply #8 on: June 11, 2019, 07:09:52 PM »

I know nothing of this guy other than watching a few of his videos but it looks like he might be able to assist you: https://youtu.be/ZRkp-CUsPQ4
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.054 seconds with 19 queries.