Bench testing PA Tubes

[Edward Cain]

   I have several types of PA tubes that I don't, as yet, have a working rig to run them in. My question is whether a bench test setup using substantially less than normal operating plate voltage would be of any real value to test for quality of these tubes.
   For example, I saw an ad on eBay for a 4-1000A where the seller presented results of some audio lab test using the plate characteristic curves for 500 V on plate and screen and compared resulting plate current to Eimac specs from the data.
   What do you think?

Ed

[KA0HCP]

http://www.g8wrb.org/data/Eimac/4-1000A.pdfda
Here is the Eimac data sheet.

500v on the screen is standard.  Plate at 500v is way to the left of the chart.   If they aren't running a full curve based on plate voltage then all you have is a single point of comparison.   

Seems pretty shaky.  Just proves that the tube has 'some' output.  I would bid accordingly with the view that the tubes are unproven.

[K1JJ]

Ed,

Using low voltage on the plate, you can certainly test a big PA tube for the usual stuff like fil operation, shorts in the grid or screen, etc. The plate current will vary nicely as the screen and grid current is varied.

However, if the tube has low emission, I don't think you will see signs of this unless you put a normal load on it using a much higher voltage. I would think if emission is low, the tube will perform fine at low voltage and low power, but fall apart with say 3kv at 500 ma for a 4X1.

If you had a lot of them to test, then a simple test setup that could handle the voltage with a bank of power resistors for the plate load would do. Using a DC or audio control signal should work fine in this case.

I once had a batch of used PA tube pulls and built a fixture for them using a blower, fil xfmr, etc.  I used it mostly to getter the gas by leaving the fils on for a coupla days. (they were indirectly heated fils)  But the real test came when they were placed in the 75M rig. I found all 10 had similar but varying parameters.  One arced over and got sent back. The others I simply matched up as pairs that had similar parameters. They worked out FB at full power. The pre-testing was worth the effort.


T

[The Slab Bacon]

Ed,
    I have often considered making a tester for large power tubes, but I am missing that all too hard to find part- a round tuit. I have plenty of square ones, but no round ones...... It would probably be a worthwhile piece of test equipment to have.

However........... I do have a way to somewhat test them. I usually put them into something that uses them and fire it up. Check for something close to what I would expect for outpoot. Then momentarily (very quickly) swing the tuning off resonance and see what the plate current peaks out at. The trick is NOT to dilly dally and do it as quickly as possible so as not to damage the tube. This will gomewhat give you an idea what it is capable of and a feel for how "good" the tube is.

If it wont pull much plate current off resonance, it is most likely only good for a decoration or paper weight.................

[Rob K2CU]

IF you have a 500V supply that can put out up to an Amp, then for a 4-1000, the only thing you won't be sure of is arc over at HV. A close look at the curves suggest that at zero grid 1 bias, you should pull 600 mA plate and about 75 mA screen at 500V plate and screen. Providing -75V of grid 1 bias should come close to shut off.  You could implement steps of say, 25V....-100, -75, -50, -25, 0, +25. IF emission is good you will be able to verify it. Just be sure to let her rip for only short times to minimize dissipation. This will also verify grid 1 control, as JJ said. Good idea to getter cook it first.

[KM1H]

Test them as a GG triode with 1000V or so on the plate and vary the bias until the anode is nice and red. Record bias and Ip and draw your own curve; if the Ip poops out before youre drawing its ratings then its a tired tube even at DC. This test tells you basic DC tube emission condition as well as regetters at a low enough voltage that wont sustain an arc.

You can also vary filament voltage from the bogey down to 5% low to get a feeling for life expectancy. There should be no drop at least the first 3%.

BTW, use plenty of air as ALL the power is going to heat.

Carl
 

[Edward Cain]

Thanks to all who commented.

What I gather is that, if I can supply a large amount of plate current, even at low plate voltages, I can get a fair idea of emission. What I obviously can't determine is whether there will be arcing at rated plate voltage.

Well, this will give me much more info than testing continuity of filaments.

Thanks again to all.

[Jim, W5JO]

If you plan to check more than one type you might want to build this.

http://www.realhamradio.com/High_Voltage_Breakdown_Tester.htm

Following the check this device will perform, plug the tube into its circuit and you will know how much emission it has.  If your tube fails the first test, don't do the second.

[The Slab Bacon]

I got lucky a few years ago and found a surplus Hi-Pot tester at a hamfest for $50. No one else showed any interest in it so I bought it. Well worth every penny!!

Saved me the time and hassle of building one!! And has earned it's keep testing transformer insulation alone. 

[WA1GFZ]

also good if your heart stalls to charge up a cap and reset the da machine

[KM1H]

I have a commercial 5KV hi-pot from the ESS back shelf junk pile ($25 a decade + ago) and a 13KV homebrew. I never plug in known problem type tubes and zap at full HV without going thru the testing.

While a gassy tube will fail even at 2KV regettering at LV and retesting may eventually result in a useable tube.

Ive salvaged many 3-500's ( and others including 4X1's) this way and while they may not test at 12KV they certainly work fine at 2500-3500VDC even as AM linears