How NOT to test a 4CX250B.

NopeMeterSo I found a new in box Amperex 4CX250B and idly did a Google search and this came up and the NOPE METER pegged. HARD. Broke the needle. I’m gonna need a new NOPE METER now. PREPARE FOR HORROR. Please note that this tube is being run at a plate voltage of 800 volts DC and hipot tested at 2500…!!

Say, did the ENGINEER pass the 2500 volt hipot test?? Forget coffee, that’d wake him up.

scarytubetest1
Interlocks? Shielding against safety hazards? Naaaah just get really up close and personal with that PLATE TRANSFORMER.
scarytubetest2
Just… reach right over that high voltage right there, WHAT COULD POSSIBLY GO WRONG??

These came from an eBay auction – auction description behind the <lj-cut>. Oh look I just dated myself. But hey look, unlike the dude who used this test setup, I’m still alive 😉

Nooow JUST SAYING — the most common application I see the 4CX250B used in is as a VHF transmitter Intermediate Power Amplifer [IPA] tube. It’s usually not subjected to much abuse like that and in many transmitters it’s asked to loaf along at as little as 50 watts! For more of a tribute to the wonders of the 4CX250B, check out the post on engineeringradio.us. Please touch a grounded shorting stick to your computer before continuing.

4cx250B testing

 

 

The present auction is for a pair as pictured used and tested 4cx250B transmitting tetrodes.  The tubes have been matched for plate current and gain as per the outlined test procedure.   Briefly stated these tubes have been procured  from a medical facility that  uses them in  equipment in large numbers and they are either replaced on a scheduled basis for preventative maintenance or otherwise have been removed if they were found to be defective.     We  have  sold  several of these already through past ebay auctions from the same lot and  some have worked well and others returned as non working.     I have recently been in contact with a local business that has very good tube testing capabilities and have decided to now offer these tubes fully tested in hopes that the prospective bidders  will now feel more secure with installing tubes  that should now  perform as they should and  also not pose any risk of damage to their equipment.  Below is a detailed description written by the Mr Silk of Silk Electronics:

 

I am Robert Silk,  sole proprietor of  Silk Electronics also based in Salt Lake City.   I have been in business selling tubes for 25 years and as a former defense electronics electrical engineer  I have taken my design experience into my basement lab assembling an elaborate test facility that is capable of testing tubes that can  not be tested by small portable testers that are typically used by most tube sellers.   I welcome questions or comments regarding my equipment and I can be reached at  fnord!

Referring to the first picture,  The heart of the test facility is based around a Weston 686 tube analyzer that I have extensively modified and improved on in many aspects.     The tube element voltages are measured by means of a Fluke dvm sitting above the Weston and external regulated power supplies that float above ground level can be placed in series as needed with the Weston which greatly increases its capabilities.    The filament supply is also external to the analyzer and can supply up to 60 amps and as seen to the left of this is an HP 200CD oscillator which provides an ultra stable low distortion sine wave as the grid signal.     The Weston analyzer itself provides the negative grid bias and also measures grid current by means of a zero center 15 ua fs meter.   The transconductance is measured on the Weston very accurately on  the center fan shaped meter with accuracy mainly limited to my ability to resolve the meter reading.

The plate supply consists of the pictured HP 712C 0-600V 200 ma supply sitting above the analyzer and is also in series with a Kepco supply ( outside the picture) bringing the total plate voltage to a bit over 800VDC.   The screen supply is to the right of the analyzer and is an HP 711A loaded with 7500 ohms because the screen current is typically negative on this tube so it is required for proper regulation…

Now regarding the testing of 4cx250B tubes as one picture shows I have built a test socket incorporating ceramic capacitors across the elements and rf choke coils in the filament and plate leads which were required to prevent oscillations .   A high speed  28VDC turbine fan  keeps the tube cool while under test.   Also visible in the main picture is a variable 20kv high voltage supply that is  adjusted down to 2.5 kv which I use for  a non-destructive high pot test.   This was done by heating the filament with the nominal 6.0 VAC for at least a minute and with the screen at zero volts and the grid at –42.0 VDC ( the grid voltage used in the rest of the testing) a neon bulb in series with the HV supply and  a 800k ohm resistor  will light if the tube under test has more than a milliamp of leakage from the plate to any other element.     I rejected any tube that failed this test at 2500V which is higher than the tube is rated to operate at.

 

Referring to the picture of the Eimac 4X250B plate curves which is the electrical equivalent to the 4CX250B my goal was to test the tube with it dissipating roughly half its rated 250 Watts which then would heat the tube enough to show up any grid current problems and also provide useful gain and plate current results that could be compared to what would be expected from the plate curves.   This curve is for the screen voltage  at 350V and as can be seen at 800V on the plate to get 200 ma of plate current we need to apply –42 V to the grid.  Also the book states that the transconductance at 200 ma is 12,000 so this was an obvious useful point to test the tube.    To verify that the test procedure was valid  I tested a  new tube and obtained reading that correlated with the expected results.

I have recorded the measured grid current,  the transconductance and plate current ,   and screen current.   Ideally the grid current should be zero but a bit usually exists however I rejected any tube showing  more than 15 ua  and if I suspected a problem I allowed more warm up time to observe the tube as grid current often only shows up when the tube has been hot for some time.

Now the question arose at what point should  the measured plate current and gain be acceptable as it may show signs that the tube’s output may be dropping off from its nominal performance.   To test this I set up a circuit using my HP 214B pulse generator which can provide a pulse of up to 100V into 50 ohms.  The goal was to pulse power test the peak emissions of the tube with a 10% duty  cycle pulse.   The output of the pulse generator is in series with an adjustable dc power supply such that the base line can be adjusted to bias the tube to cut-off which is defined as 1 ma plate current.       As can be seen at 800V on the plate with the grid at zero volts the plate current is a bit above 1 ampere and this test is a good correlate of the tube’s ability to put out full power.   As seen a 6 uf capacitor  is then connected between the plate and cathode to supply the 1 ms pulse while the plate supply recharges the capacitor.       A Fluke   DVM in series with the plate circuit measures plate current.  It follows then that if for example 100 ma average plate current is measured the peak plate current is exactly 10 times this or 1 amp because the duty cycle of the pulse is 10% .   The tube can be safely tested in this manner because the average plate dissipation is kept below its maximum ratings of 250 Watts.

It is interesting to note that although the tube under test may test at say 70% of its nominal gain and plate current at the aforementioned test point the peak emissions is still typically close to or even at 100% of its nominal value hence I am using the gain and plate current measurements mainly as an aid in selecting plate current matched pairs and accepting any tube that has a peak  emissions  of above one ampere.    A drop off in gain from nominal represents a requirement for a greater drive signal for a given output however I believe that the peak emissions is a more important measurement.

 

In summary of the 50 tubes I have tested so far I rejected roughly 1/3 of the tubes as per the above testing procedure .  Accordingly I feel the testing has adequately been performed and I hope that the prospective bidder will now bid with confidence.

 

2 thoughts on “How NOT to test a 4CX250B.”

  1. It is playing in the vacuum tube, such as 4CX250B and 4X150A / D.
    4cx250b dedicated tube tester is good if the circuit diagram but,,,
    I use a simple tube tester in the self-made.

    It is your great Shack.

    Imai-cho 1221-2 Gojo, Nara Prefecture, Japan
    Name Matsuda Tomotaka

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.