-----Original Message-----
From: Tom Rauch <w8ji@contesting.com>
To: amps <amps@contesting.com>; Steve Thompson <rfamps@ic24.net>
To: <amps@contesting.com>
Date: 07 September 2000 22:11
Subject: Re: [AMPS] Blown TL922A... What to do?
>
>> characteristic remains the same. There is a single series resonance in
the
>> low/mid vhf region, and the impedance stays low either side of it. When I
>> next get the bench clear (might be a while), I'll run some plots. Please
>> say if there are any particular configurations you (or anyone else) would
>> like me to try and simulate to measure.
>>
>> Steve
>
>Hi Steve,
>
>What I found useful and educational is the measure the
>impedances along the path from anode to the switch contacts, and
>calculate the amount of current the tube must sink in order to
>develop the voltages necessary to cause the switch to arc at
>say...100 to 200 MHz.
>
>The 3-500Z, in the fiberglass socket with groundplane on it, has
>minimum S21 (or S12) loss at about 180-200MHz, indicating the
>grid is "ungrounded" at that frequency. This would be near the
>frequency where the tube oscillates in a parasitic mode.
Has anyone been able to measure the frequency of oscillation in, for
example, TL922? What dip frequencies are found with a GDO in a 922? In other
amps?
>
>As a matter of fact if you remove the suppression you can measure
>the oscillation in some cases where the tube takes off, and it is up
>there in that area.
>
>What I found was a steady stepping down of VHF voltage through
>the distributed impedances in the tank ans associated wiring at
>VHF. Measurements indicated it is IMPOSSIBLE for the tube to
>drive the tank at VHF and cause a switch arc.
>
>Of course you nee a real amp to measure. What amplifier do you
>have available? If you have one that I have, we can go through the
>measurements and compare results on here.
I don't. I've never mastered a straight key, so I've not been able to pass
the code tests to get on hf over here. A degree and 25 years designing rf
for a living doesn't count. I'm hoping to sidestep the system by taking
Extra class exams next month.
>
>It should be very simple, with the right test equipment, to prove the
>possibility of switch contact failures from VHF parasitics.
>
>I maintain contact failures are caused by problems at or near the
>resonant frequency of the tank, where driving impedance matches
>or nearly matches the tube's optimum operating loadline. It is easy
>to get several times the operating voltage of the tube in the tank if
>the tank is improperly loaded.
>
>Do you agree or disagree with the above statement?
Without hands on experience of real amplifiers, I'm treading very carefully.
I chipped in because I'm lucky enough to have the gear to measure impedance
'properly', and those who haven't might not realise what happens in hf sized
layouts as the frequency goes up. I was also hoping to gain an insight into
how Rich visualises these things, as he and I use language that's
sufficiently different to cause problems in discussion.
For sure, if you have a low impedance at a given point and frequency, it is
difficult to generate high voltage there, and if you mistune the output
network you can get a high impedance and lots of volts at the tune C. That's
simple fact.
There's a few types of amp that crop up regularly here with problems, such
as TL922. Anyone near Newbury got one I can poke around inside?
Steve
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|