[TMC] PAL-350 Trouble

[JF Samuels - K2CIB]

With your credentials I am sure you will solve it soon.  I agree that it 
is best to ask if others have encountered the same problems before 
delving in too deep.  It does seem to be in the output section rather 
than the drive.  Do you have a spectrum analyzer? They could be 
doubling.  I use a bird meter for power output to a 51 ohm load, "T" 
pick off and an IFR 1200S to look at the signal.  If it was a 
neutralization problem, it could be creating power at a frequency not 
handled by the current output network and dissipating it in the coils.  
It might be worth bypassing the screen grids with external caps.

During my short career at TMC I worked on that amp, the GPT-750 and 
larger ones, receivers, audio equipment, the PMO (I built the 
test/adjustment device for it), and worked with the engineers on 
designing a follow-on to the GPR-90.

Good luck!

73, John

On 8/5/2018 10:03 AM, Meir Ben-Dror WF2U wrote:
>
> John,
>
> Thanks for the input.
>
> I bought the PAL-350/PSP-350  together with the SBE-3 and the PMO. 
> They were set up,  and working as a secondary station in the owner's 
> shack (I had a schedule on 40 m with the owner who used the setup to 
> demonstrate it a couple of days before loading them in his vehicle) 
> prior to transporting them to Dayton, and from there to my QTH. The 
> units were in individual desktop cabinets. I installed the components 
> in a rack I already had and interconnected as they were.
>
>   * I have the printed manuals for everything, plus I downloaded the
>     GPT-350 system manual to confirm I interconnected everything
>     correctly, from John Poulton's website (I know him personally and
>     got the other PAL-350A/PSP-350A I referred to from him, but that
>     was out of his storage, in unknown working condition).
>
> When I set up the system, after verifying correct operation of the PMO 
> (I have a working VOX-5 also) and the SBE-3, I first made sure the 
> bias for the idle current is set up correctly, and I redid the 
> procedure according the manual.
>
> The test was on the 80 m frequency of the net I check into weekly 
> (3715). I did the amplifier test using the initial settings according 
> to the manual, and the tuning was very close to the initial setting.
>
> I have to add that I'm an experienced ham, using tube equipment since 
> the 1960's and collecting/restoring/using only tube equipment, both 
> military and commercial surplus, with some classic ham gear in 
> between. I also happen to be an EE, now retired, with the bulk of my 
> professional career spent in designing high power, wideband tube RF  
> (distributed amplifiers from 200 W to 5 KW output, covering the range 
> of 10 kHz to 200 MHz, mainly for RFI testing and MRI research) as well 
> as solid state amplifiers, including a commercial 1 kW solid state amp 
> covering 1.5 to 30 MHz for communications.
>
> I always like to gather information on equipment I'm working on from 
> others experienced with the same equipment, to save time and to find 
> out whether there is a known common known point of failure in the 
> particular equipment I'm not yet familiar with. I enjoy operating more 
> nowadays than spending my limited  time analyzing everything starting 
> at the component level...
>
> Back to the amp: If the drivers had a problem, I wouldn't be able to 
> drive it to the maximum plate current from the SBE-3. I'd expect 
> around 50% efficiency from an AB1 amplifier. By the way, the amp 
> doesn't have the SSB/CW switch for changing the bias for higher 
> efficiency for CW,  like the later models. The other, -A version amp I 
> mentioned does have the switch.
>
>
>   * The ADLC is turned down.
>   *  I haven't tried it on another frequency.
>   * I'll look for the initials on both amplifiers.
>   * The 20% efficiency I'm experiencing is common to both amplifiers
>     (using the first amp's power supply). Next test is the other amp
>     with its own supply.
>   * Today I'll test it with another power meter, but I got the result
>     on two already. I expect and accept +/- 20% discrepancy between
>     the meters. I could invest in a real calibrated power meter and
>     not rely on the SWR/power meters I have in antenna tuners and
>     standalone - on the other hand I'm not a fanatic to squeeze every
>     watt out of a device, because I think in dB when looking at power.
>   * if the screen voltage was way too low, I wouldn't be able to drive
>     it to full Ip with the available drive. This rules out also very
>     leaky screen bypass caps that are built in the sockets.
>
> Back to the shack this morning for a little weight lifting to put the 
> PSP-350A supply on the bench to test the -A amp with it. Tomorrow I 
> have a chiropractor's appointment anyway...
>
> Again, thanks for your ideas, and if you're interested, I'll follow up 
> with further reports.
>
> 73, Meir WF2U
>
> Landrum, SC
>
>
> On 8/5/2018 1:51 AM, JF Samuels wrote:
>> Meir, here are my random thoughts:
>>
>>   * Use a scope with a diode probe to measure the amplification of
>>     the three stages to see if you are losing RF.
>>   * Substitute tubes, one at a time.
>>   * Electrolytics of this age are probably not working well, if at
>>     all.  Replace them.
>>   * Look at any circuit which could bias off the tubes because of
>>     negative feedback.
>>   * The rating is 200W CW output.  At 1900V this means that you would
>>     need only 105 ma / .65 efficiency = 162 ma plate current.
>>   * Does the output power change with frequency?
>>   * If you have 340 ma @ 1900V, where is the other 400W going? Faulty
>>     output pi network?  Bad match to tube impedance?
>>   * Here is the manual:
>>     http://www.tmchistory.org/tmc_manuals/manuals_db/pal-350/tm_pal-350b_8_20_62.pdf
>>
>>   * Are my initials on the chassis: JS?  I worked on many of these in
>>     final test in Mamaroneck in 1958.
>>
>> John Samuels, K2CIB
>>
>>
>>
>