[TMC] TMC PAL-1K Low Output Power Issues
triodes at optonline.net
triodes at optonline.net
Thu Jan 26 11:52:17 EST 2023
Hi John,
Thanks again for your guidance!
I did uncover one thing today. Someone had changed L222 in the cathode
circuit of the 6146 (see image. It's the large adjustable pi-wound choke
with the green winding). This choke plus the other components in the
6146 cathode appear to be a part of the RF negative feedback circuit,
and this is where the sampled output from the PL-172A is injected into
the driver stage. L222 is 750 uH @ 100 Ma, and I have no idea as to what
value choke replaced it, but it is probably a similar value, judging
from it's size, etc. My guess is if the replacement is greater than 750
uH, it should not present a problem.
I'm thinking of replacing it. I have 1000 uH and 600 uH available. I
doubt if the value is that critical....what do you think?
With regard to the plate tuning, loading, and coarse loading settings,
here are a few of the settings I had recorded when I last fired her up:
3-4 Mhz Band: F=3850 Khz, Plate Tune: 076. Loading: 032. Coarse Load: 1
6-8 Mhz Band: F= 5000 Khz. Plate Tuning: 063. Loading: 034. Coarse Load:
2
6-8 Mhz Band: F= 4500 Khz. Plate Tuning: 052. Loading: 034. Coarse Load:
2
Note that I can tune the PL-172A to resonance, as evidenced by a dip in
Ip, and as I tune off resonance, the output power goes up and eventually
the overload breaker trips. This is true on all bands, and at resonance,
the output power is around 50 watts on all bands, and going
substantially off resonance will increase the output power to around 150
watts. It is when going beyond that when the breaker trips.
I noted your earlier comments about tuning this amplifier only to
resonance, and balancing the the Ip and Isg.
73,
Bruce, W2XR
------ Original Message ------
From: jvendely at cfl.rr.com
To: triodes at optonline.net; tmc at mailman.qth.net
Sent: Thursday, January 26, 2023 11:01 AM
Subject: Re: [TMC] TMC PAL-1K Low Output Power Issues
Hi Bruce,
Please see my comments below, set off by <<...>>
On 1/25/2023 8:55 PM, triodes at optonline.net wrote:
> Hi John,
>
> Thanks for the reply!
>
> I can't tell you off-hand what the grid bias value is for a
> resting/quiescent Ip of 175 Ma, but it is with the grid bias
> adjustment pot providing the minimum possible bias voltage.
<<With a stock PS-4A power supply, setting the grid bias adjust to
minimum voltage should result in very high plate current, far higher
than 175 mA, so low cathode emission is certainly possible.>>
>
> Here are the operating parameters I have recorded for this
amplifier:
>
> Ep: Approx. 3000 VDC
> Ip, with Grid Drive: 400 Ma
> Isg: 15 ma
> Esg: +500 VDC
> Grid Drive: 14 on the Multimeter. Recall that the multimeter does
not
> read the grid drive in Ma, but rather it is a relative reading based
> upon the rectified RF drive voltage as applied to the PL-172A grid.
> Note that the TMC-recommended reading for drive to the PL-172A is
10,
> so I am driving the PL-172A harder than recommended. At a drive
level
> of 10, the RF output power drops to around 50 watts on all bands.
> RF Output Power: Approx. 150 watts, on all bands
<<You don't say what the control grid bias voltage was, so I'd check
that. Also, what frequency were you tuning on, and what driver tuning,
plate tuning, and plate loading control settings resulted in the above
currents and output power?>>
>
> These parameters and the RF output power are pretty consistent on
all
> of the bands within the RFC-1.
>
> That is interesting about the driver stages, and the possibility
that
> they may be outputting a 2nd harmonic of the excitation frequency.
But
> would this be true on all of the bands, based upon the consistency
of
> the relationship of DC input to RF output/poor efficiency on all
> bands? I don't think so.
<<Yes, it would not be possible to tune the driver to a harmonic on
all frequencies, but it is possible at some frequencies.>>
>
> There may be an issue with the driver stage. The 2-4 Mhz band
requires
> that the 1st Amplifier Tuning cap be fully counter-clockwise
(position
> 0) to obtain drive to the driver stage, when tuning the amplifier on
> 3850 Khz. This is indicative of the 2-4 Mhz driver band either
> requiring realignment, or a defective component, etc. The 4-8 Mhz
band
> within the driver stage is completely dead, but I can get adequate
> drive from the 8-16 Mhz band. All of the other driver bands operate
> correctly. My plan was to attend to the 2-4 and 4-8 driver band
issues
> once I got the low RF output issue addressed.
<<Since there are obviously multiple faults in the driver circuits
which you've got to correct eventually, I'd be inclined to fix these
problems and realign the all the driver bands before moving on. You've
already done a good test of the output network which gave reasonable
results, so a good guess is the problem is less likely to be in that
area.>>
>
> Can you also provide any guidance as to the possibility that the
plate
> blocking capacitors may be defective, and they could be responsible
> for the gross inefficiency of the output stage, etc.?
<<I suppose anything is possible, but the plate blocking capacitors
failing open would be a very unusual failure mode for this type of
capacitor. If they were shorted, there would have been fireworks and an
immediate plate breaker trip as HV appeared at the RF output connector.
If the capacitors are not discolored or leaking oil, I'd bet they're
OK. Your output network test also adds weight to this.>>
>
> Any more suggestions and guidance that you can recommend here?
<<The bottom line is we have multiple problems here--a weak PL-172, at
least two driver stage faults and possible misalignment, and perhaps
another yet-undiscovered problem. It's a little hard for me to
troubleshoot much further without being able to see and operate the
amplifier. But double check the grid bias voltage as the bias adjust
pot is varied, and let me know what the driver tuning and output tune
and load control settings were when you experienced the poor
efficiency. There could be a clue there...>>
73,
John K9WT
>
>
>
>
>
> ------ Original Message ------
> From: jvendely at cfl.rr.com
> To: triodes at optonline.net; tmc at mailman.qth.net
> Sent: Wednesday, January 25, 2023 8:04 PM
> Subject: Re: [TMC] TMC PAL-1K Low Output Power Issues
>
> Sorry, I completely missed your comment that you're getting
> high plate current but little RF output, which puts a somewhat
> different twist on things. Still, the fact that you can't get idle
> current over 175 mA at any reasonable grid bias suggests low
> emission. -110V Eg1 should produce about 220mA. I would start with
a
> DC emission test as I described earlier, and go from there. What
grid
> bias produces 175 mA plate current in your amp? What plate and
screen
> currents do you get when applying RF and getting the high input
> power/low output power condition? Tubes with 10-15% emission drop
may
> still get you close to full output, though with much higher input
> drive and poor linearity.
>
> That horribly low plate efficiency sounds more like a grossly
> mistuned condition. Have you aligned the driver stages? Don't
assume
> they're correct, as these amps may be badly screwdrivered if
> previously owned by a ham. I've seen some real butchery.
>
> In some cases, it is possible to mistune the output of the driver
> section to the second harmonic of the input stages, essentially
> turning the driver into a frequency doubler. This will make it
> impossible to tune the PA section. This could happen more easily if
> the driver is badly out of alignment. If you haven't done so,
> carefully go through the driver stages alignment procedure and make
> sure things are tracking properly. Before you begin, measure the
> input resistors in the 1st RF driver, as these have often been
> overheated and are out of tolerance.
>
> The RFC amp which serves as the IPA for the GPT-10K and 40K is
almost
> identical to the RFD amplifier in the PAL-1K, but there are some
> differences. The driver stages are somewhat different, and the
output
> inductor taps are in different locations than on the PAL-1K, as the
> amp only tunes to 28 Mc. So, follow driver alignment procedure in
the
> GPT-10K manual.
>
> 73,
>
> John K9WT
>
> On 1/25/2023 5:21 PM, triodes at optonline.net wrote:
> > HI John,
> >
> > Wow, your reply is incredibly helpful, and I thank you for taking
the
> > time for providing such a detailed and concise reply. I'm sure
that
> > other users of the RFC-1 and RFD-1 will find your guidance
extremely
> > helpful as well, and are equally appreciative!
> >
> > It is good to see, based upon your response, that the likely
culprit
> > here is the PL-172A.
> >
> > Do you think there is also the possibility, as another poster has
> > suggested, that perhaps the two 500 pF plate blocking capacitors
may
> > be bad? Not shorted, but open? My feeling is this is probably not
the
> > case, as I doubt that the test I had performed earlier, that
being
> > terminating the PL-172A plate to ground with an 1800 ohm
resistor, and
> > looking into the RF output port of the amplifier with my network
> > analyzer, would have yielded the kind of positive results I was
> > seeing, assuming the blocking capacitor was bad.
> >
> > Assuming the PL-172A is indeed weak, where is all of the RF
output
> > going, with roughly 1400 watts of DC input? Will a weak PL-172A
result
> > in such a gross loss of efficiency, that in reality, the tube is
only
> > delivering around 150 watts to the input of the pi-network, with
1400
> > watts of plate input?
> >
> > My PAL-1K "clone" consists of the PS-5 HV PSU, the PS-4A
intermediate
> > voltage PSU, and the RFC-1 PA unit. Unlike the "real" PAL-1K, I
am
> > using the RFC-1, instead of the RFD-1. The RFC-1 was given to me
a
> > number of years ago in beautiful cosmetic condition, and it was
pulled
> > out of a TMC GPT-40K at coastal station WSL, here on Long Island.
It's
> > been a long road, but I want to put that RFC-1 back on the air,
and
> > actively use it.
> >
> > That said, I had to create a custom wiring harness to correctly
match
> > the interfaces on the PS-4A, to that of the RFC-1. Of course when
> > using the RFD-1, this does not need to be done, and the captive
> > pendant cable on the RFD-1 connects directly to the mating
> > MS-connector on the PS-4A.
> >
> > Thanks again, John. Very much appreciated!
> >
> > 73,
> >
> > Bruce, W2XR
> >
> >
> >
> > ------ Original Message ------
> > From: jvendely at cfl.rr.com
> > To: tmc at mailman.qth.net
> > Sent: Wednesday, January 25, 2023 4:05 PM
> > Subject: Re: [TMC] TMC PAL-1K Low Output Power Issues
> >
> > Hi Bruce,
> >
> > I have several TMC transmitters which have various versions of
this
> > power amplifier that I've been working with over the last 40
years.
> > They are excellent, stable, trouble-free amplifiers. From your
> > description, and assuming all element voltages are correct at the
tube
> > socket, I agree that you very likely just have a bad output
tube. If
> > at the nominal grid bias of about -110V you cannot obtain the
correct
> > idle plate current of about 220 mA, the tube probably has low
cathode
> > emission. Your amplifier is a perfectly useful tube tester. A
> > simple emission test can be done by adjusting grid bias (no RF
drive)
> > and checking plate current against a few points on the published
> > characteristic curves. You should be able to adjust up near max
> > dissipation if the tube is good. Do it fairly quickly, and keep
> > screen and plate dissipation within limits.
> >
> > Relatively few original Penta Labs PL-172s (with the glass
seals) are
> > still good today. Many have become gassy, others have developed
> > internal "barnacling", which can cause arcs when HV or RF drive
are
> > applied which can damage the amplifier. The ceramic seal 8295A
is
> > much less prone to gas problems than the glass seal PL-172 but
many,
> > especially the oldest, are no longer useable.
> >
> > Arcing between anode and screen grid may vaporize R222, the 12
ohm
> > resistor in the screen grid circuit. It's very important to make
sure
> > shunt and multiplier resistors in the metering circuits are OK so
the
> > meters read accurately.
> >
> > Before attempting operation with a new tube, I would recommend
> > running it overnight on filament and grid bias (with normal
cooling
> > air, of course), and afterward, do a comprehensive hi-pot test
for
> > inter-element leakage current, if you can. Problems not evident
with
> > simple ohmeter tests often show up this way, and in some cases,
the
> > tube can be "debarnacled" to the point where it functions
properly.
> > Parallel all grids, cathode, and filament terminals, and measure
> > leakage current between anode and the paralleled elements. Test
with
> > anode voltage up to at least twice the DC plate voltage. Leakage
> > should be somewhere under 50 microamps or so.
> >
> > Unfortunately, good PL-172s and 8295As are getting harder to
find.
> > Tubes advertised as NOS frequently are not, and even those that
are
> > NOS may be defective, so it's all a bit of a crap-shoot.
> >
> > Correct neutralization is of course important, but it is not at
all
> > critical on this amplifier. Even if grossly maladjusted, it will
> > absolutely not cause the problems you're having. If
neutralization
> > is correct, plate and screen currents should dip at very nearly
the
> > same plate tuning setting, and there should be little to no
> > interaction between the grid tuning and plate tuning adjustments.
> >
> > Improper tuning and loading of these amplifiers is a very common
> > problem. All you need do is follow the tuning procedure in the
> > manual, but here are a few brief pointers. When tuning up,
always
> > keep the multimeter switch in the Isg position, and
simultaneously
> > monitor both plate and screen currents. Many PL-172/8295A tubes
were
> > trashed by improper tuning/loading adjustments resulting in
excessive
> > screen current. This warps the screen grid, and the old arcing
> > problems begin. Keep screen dissipation within limits (35W) at
all
> > times.
> >
> > Unlike many amateur amplifiers with tetrodes, which are tuned
for max
> > screen current and min plate current, the PAL-1K is tuned for
MINIMUM
> > plate and screen currents, and these should nearly coincide.
High
> > (and very touchy) screen current indicates an underloaded
condition,
> > and will result in erratic operation, poor linearity, frequent
breaker
> > trips, and possibly damage to the tube. As loading is
incrementally
> > increased and tuning is re-dipped, screen current will decrease
and
> > become less touchy as you approach a properly loaded condition.
At
> > proper load, plate and screen dips should be smooth and
uncritical.
> > If necessary, it's better to run slightly high plate current to
get
> > the screen current in-range. The PL-172 is a radial beam
pentode, and
> > contrary to popular belief, you will not see negative screen
current
> > on this amplifier. If the screen current meter goes negative,
> > something is seriously wrong.
> >
> > With proper tune and load procedure and a good tube, you should
very
> > easily obtain 1 kW CW at plate currents of 400-500 mA, and well
under
> > 100 mW drive, depending on frequency. Below 14 Mc, you may only
need
> > 20-30 mW drive to get full output. Screen current should not
exceed
> > 35-40 mA. For best linearity, you may need heavier plate
loading,
> > running at the high end of the aforementioned plate current
range, but
> > it shouldn't exceed around 550 mA. You must load the amplifier
to
> > full output on single tone before operating SSB, even if you
intend to
> > operate at reduced output. With a two-tone signal at 1 kW PEP
3rd
> > order IM should approach -40 dB relative to either tone.
> >
> > Hope this helps, good luck, and let us know how things go. It
will
> > be good to hear another of these transmitters on the air...
> >
> > 73,
> >
> > John K9WT
> >
> >
> >
> >
> > On 1/24/2023 11:15 PM, triodes at optonline.net wrote:
> > >
> > > Hello, Fellow TMC Enthusiasts!
> > > I have been spending a lot of time debugging this TMC PAL-1K
> > > amplifier subsystem I have here. As some of you know, the
PS-4A and
> > > the PS-5 units were basket cases when I got them, but the
final
> > > amplifier was in surprisingly good shape, and I figured the
power
> > > amplifier unit would be largely plug and play, but that has
> proven to
> > > not be the case. I have made a lot of progress, and I am now
> down to
> > > one remaining issue; the final amplifier is suffering from
very
> poor
> > > efficiency on all bands. With 1400 watts DC input, I am only
> getting
> > > out around 100 to 150 watts, where it should normally be on
the
> order
> > > of 700 to 850 watts CW output.
> > > The fact that this is true on all bands indicates that the
coarse
> > > loading padding capacitors at the output-side of the
pi-network are
> > > good. If they were shorted or open, that would be a hell of a
> job to
> > > get in there to remove them. Like many of the components in
the
> RFC-1
> > > or RFD-1 1 KW Power Amplifiers, these amplifiers were
literally
> built
> > > around these parts, with very little consideration given to
future
> > > serviceability, etc.
> > > I think the issue is a weak PL-172A output tube, or the final
> stage is
> > > perhaps over-neutralized. PL-172As are becoming unobtainium,
and I
> > > have one (hopefully) good tube on the shelf, but I'm reluctant
to
> > > install it unless I know what is really going on here.
> > > Over-neutralization is of course easily dealt with in either
the
> RFC-1
> > > or RFD-1 amplifiers; TMC made the neutralization process very
> simple.
> > > Remove the hole plug on the front panel, and neutralize the
final
> > > stage per the procedure in the tech manual.
> > > I think the PL-172A is weak, as I cannot get the resting
cathode
> > > current above 175 Ma with 3000 VDC on the plate, and the TMC
> spec is
> > > 220 Ma idling current with (of course) no signal applied to
the
> grid.
> > > That is usually indicative of a tube suffering from weak
emission.
> > > There can't be too many issues causing this low efficiency in
the
> > > final amplifier stage. The voltages on the plate, screen,
grid, and
> > > heater are all correct, and the tube is getting plenty of
drive.
> Part
> > > of the problem is I don't have any way of properly testing the
> > > PL-172A, and I certainly don't want a repeat of what happened
> the last
> > > time, fireworks and all, when I installed a PL-172A that came
> with the
> > > RFC-1 marked, "Used, Tests Fine". That tube may have tested
fine 40
> > > years ago, but it most probably became gassy sitting on the
shelf
> > > unused for all of these years.
> > > I know I'll eventually get there, but I figured I would
solicit
> some
> > > thoughts and suggestions from any of you who have been down
this
> > > PAL-1K (or similar TMC equipment) road before.
> > > 73,
> > > Bruce, W2XR
> > > ______________________________________________________________
> > > TMC mailing list
> > > Home: http://mailman.qth.net/mailman/listinfo/tmc
> > > Help: http://mailman.qth.net/mmfaq.htm
> > > Post: mailto:TMC at mailman.qth.net
> > >
> > > This list hosted by: http://www.qsl.net
> > > Please help support this email list:
http://www.qsl.net/donate.html
> > ______________________________________________________________
> > TMC mailing list
> > Home: http://mailman.qth.net/mailman/listinfo/tmc
> > Help: http://mailman.qth.net/mmfaq.htm
> > Post: mailto:TMC at mailman.qth.net
> >
> > This list hosted by: http://www.qsl.net
> > Please help support this email list:
http://www.qsl.net/donate.html
> >
> >
> >
> >
>
>
>
>
More information about the TMC
mailing list