>>>Joe, Skip, Rich, ET. al....,
>>> Please 'critique' the following and see if I am wrong in any of my
statements.
>>> If I am wrong I humbly apologize to K8CU and anyone reading this post,
but if I am correct, I hope you will support my recommended change to his
published circuit.
>>>Regards,
>>>Dennis O.
>>>12
---------------------------------------------------
>>>I just read the post quoted below, then went to the K8CU website and
looked at the HV breakdown tester referred to in that post; and after taking a
close look at the schematic shown on the site, I have some doubts as to the
accuracy of the tester as it was applied in this particular case, and also in
the
example given on that site referring to the testing of some Russian (GS35B, and
GS31B) and American (4CX250B and 4CX800A) RF power Vacuum tubes.
>>>The writer of the original post made the following statement:
Another GS-35B had 3kv as upper limit, but after cooking it, a 8 kv limit was
reached before a leakage 50 microamps occurred!
>>>This indicates to me that he was making his tests by cranking up the
voltage until approximately 50 ua of leakage current showed on the meter of the
high voltage breakdown tester.
>>> He also stated that:
One of the GU-43's had a 2200 volt limit, but after letting it cook for a
while, it can 'take' 6 kv before leaking.
>>> and:
A copule of GI7Bs had about 1500 v, but after ritual, I think at end, I had
approx 3900 volts on both...
>>>Here's what I see as the problem:
>>> The schematic I looked at would have a 'significant' error factor with 50
ua of leakage current applied to the Device Under Test, (D. U. T.).
>>> This error factor would increase linearly with the leakage current from
an error of ZERO at 0 ua to an error of a little over 2500 volts at 50ua.
>>> That would mean that that GS-35B Joe had tested must have had a 500 volt
'upper limit,' instead of a 3,000 volt 'upper limit' with 50 ua of leakage;
and those G-17B's would NOT be testing good at 3900VDC but rather at only
1400VDC!
>>> In the example given on the K8CU website, he stated that he had applied
4KV at 25ua to the tubes being tested, meaning that the actual voltage applied
to the devices under test was 'significantly' less; or: [4KV - (25ua X 50
Megohms)] = only 2750 volts!
>>> The problem is due to the extremely high series resistance in the leakage
path where the 50 ua meter is located, and can easily be 'fixed' with a very
minor, (*and cost saving because you don't have to purchase as many
resistors), change to the circuit.
>>> This is the modification I suggest:
>>> Simply reduce the series resistance to the voltmeter by 50 megohms and
connect it to the high (most positive) side of the bridge rectifier that is
used
across the leakage current meter.
>>> Now that leakage current voltage drop in the 50 meg series resistance is
accurately accounted for in the voltmeter reading and much more accurately
represents the voltage applied to the device being tested.
>>> * (and now you would only need 29 of those 5.1 meg resistors instead of
39 of them).
>>> PS. I also do not understand the need for the diode bridge across the
meter if there is always a DC voltage applied to the D. U. T. from the tester,
but I'm sure there is some reason for it that is probably way beyond my
ability to understand.
-------------------------------The Original
Email-----------------------------
Hello Jeff, Rich, Skipp Et. al
I have had some experience in this....
After buying a few different russian tubes....(GU-43B, GS-35B, GU-84B,
GI-6B/7B).
I've made some tests:
Another GS-35B had 3kv as upper limit, but after cooking it, a 8 kv limit was
reached before a leakage 50 microamps occured!
One of the GU-43's had a 2200 volt limit, but after letting it cook for a
while,
it can 'take' 6 kv before leaking.
A copule of GI7Bs had about 1500 v, but after ritual, I think at end, I had
approx 3900 volts on both...
The gadget for these measurements was a K8CU High voltage breakdown tester.
http://www.realhamradio.com/High_Voltage_Breakdown_Tester.htm
-Clipped here again-----
Regards
/Joe
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|