[AMPS] More SB-220 amusement(!)
Pete Smith
n4zr@contesting.com
Sat, 20 Sep 1997 15:44:24 -0700
Rich Measures wrote:
>
> >At 09:58 AM 9/18/97 -0800, you wrote:
>
> NOTE -- Although the e-mail input from Mr. Smith was not posted to the
> [AMPS] group, I'm guessing that other amplifier owners might like to be
> included in the ongoing analysis and repair process. Thus, I am posting
> this reply to the group.
>
> >>>On Thu, 18 Sep 1997 10:39:52 Pete Smith <n4zr@contesting.com> writes:
> >>>>
> ....... .....
> >> Check the tube with a high-Pot. If you don't see over 6 kV breakdown
> >>between the grid and the cold filament, the tube may have been damaged by
> >>a VHF parasitic.
> >>
> >Easy to say, but how many people have a hi-pot tester.
>
> I don't have a clue how many do. However, high-pots are easy to build.
> An article on building one is available on my my Web site. Repairing or
> building an amplifier without a high-pot. is difficult at best. C and H
> surplus on Colorado Blvd. in Pasadena, CA reportedly has some 4500Vrms
> low current 50/60 Hz transformers in stock.
>
> >Not me, for one.
> >In the history of this amp with me, there has not been a single, sudden
> >transitory event that plausibly resulted from a parasitic.
>
> What seems plausible is worthless compared to what one measures.
> If the 0.82 ohm grid current meter shunt resistor opened, there are three
> possibilities:
Rich, it wasn't the grid shunt, but one of the plate metering resistors
-- I would have traced the backside of the Harbach board, but am
currently in NC attending the birth of a grandchild, and can't say for
sure which one because it was incinerated. But the grid side seems
untouched.
> . 1. Anode short: - During the 90s, Eimac produced some 3-500Zs with
> weak spotwelds on the anode cooler assembly. If some of the weak
> spotwelds let go, the anode cooler tilts and touches the grounded grid.
> Since the only dc path between ground and the negative side of the anode
> supply is through the 0.82 ohm resistor, it can burn out.
There is no visible tilt.
> . 2. Gas :. - With anode V applied, the inside of the tube will glow
> with a blue color. It may be a good idea to connect an electric heater
> in series with the electric mains for this test since, with a large
> amount of gas present, something is needed to limit the mains-I. If you
> see whitish deposits inside the envelope, a large amount of oxygen gas
> was present when the filament was operated.
No blue, no white.
> . 3. VHF Parasites. - Measure the R of the VHF parasitic suppressor
> resistors. If the R is substantively higher than the marked value, a
> parasite is indicated. Using a high-pot -- provided the vacuum checks
> good -- if the grid/filament breakdown potential is lower than normal, a
> bent filament is indicated.. An alternate test is to use a C meter to
> measure the grid to filament C. If the C is greater than 10 pF, a bent
> filament is indicated.
> [NOTE - Bent filaments can be caused by other than a VHF oscillation. If
> the 3-500Z has been operated horizontally for more than c. 4k hours of
> filament run time, the filament might be able touch the grid.]
> ... ... ..
> >>I would change the 220 from V-cutoff bias to R-cutoff bias. In the stock
> >>config. a shorted tube will destroy the unfused, filament transformer.
> >
> >Sorry, I don't understand what this implies, ...
>
> On standby, +110V is applied to the fil. CT to cut off the tubes. If a
> grid/fil. short occurs, it places a short on the +110V supply, which
> crispy-critters/smokes the unfused transformer that supplies the ac
> voltage to the supply.
>
> >...or how you would go about it.
>
> In it's stock configuration, +110V is wired to the relay coil and to one
> of bias section contacts on the relay. Remove the wire that brings
> +110V to the bias contact. (the bias contacts are the middle section of
> dpdt contacts). Check to see if one side of the coil is still wired to
> the
> +110V supply.
>
> >The popping was very muted, and clearly from the metering board. There was
> >no color or other transient stuff happening with the tubes. The anodes
> >still look entirely normal.
> >
> Which, if any, resistor(s) on the metering circuit board have changed R?
> The 1.0 and 0.82 ohm units are of special interest.
I'll have to check when I get home, but I believe it was probably the
1.0 ohm (I think, from memory) in the plate metering circuit. does this
change your diagnosis?
>
> >... and things operated just fine thereafter through an entire CW
> >sprint. Doesn't sound very plausible that it was killed by a parasitic and
> >then recovered for that duration.
> >
> Intermittent parasites don't necessarily prevent an amplifier from
> subsequently functioning normally.
> >>
> >>>...and maybe the meter also unless there were protection diodes
> >>>across it.
> >
> >There are.
> >
> good
>
> >>>Other things probably blown are one or more of the bias diodes on the
> >>>Harbach board.
> >
> >What's the cause?
>
> All of the grid current and all of the anode current passes through the
> bias diodes. Bias diodes can short during an instanteous current surge in
> grid or anode current.
>
> >If I replace the metering resistors and the resting
> >current looksw OK, does that rule out the diodes?
>
> Yes. If the bias diodes short, the ZSAC/'resting current' would increase
> to several hundered mA because there is zero volts of operate bias. //
> I do not normally replace a resistor unless the measured resistance is
> out of tolerance.
>
> >I've read and re-read those articles -- and much of the debate here and
> >formerly on rec.radio. It just doesn't sound like what you describe, Rich.
>
> Full blown, big-bang ones get most of the attention, Pete, but there are
> the also-rans.
>
> > I'm agnostic on VHF parasitics, but they are clearly not the only thing
> >that goes wrong in HF amps.
>
> Clearly. However, the naysayers have tried to cloud the matter with
> furious, spurious arguments -- for a reason that isn't terribly difficult
> to decipher. One of the naysayers even accused me of having said that
> 'parasitics bend grids'. ROFL, L. The whole biz reminds me of the
> goings on during Richard Millhouse Nixon's final months in the
> Whitehouse.
> - Amplifiers can fail from a number of causes. Parasitics are just one.
> To the best of my knowledge, Kenwood (Trio) is the only manufacturer who
> will currently say anything about parasitics. In the old days, Kenwood
> talked about "normal arcing" on page 14 of the 922 owner's manual. Now
> Kenwood says that the TL-922 does not have any more parasitic-arcing than
> other brands. However, the 922 is out of production -- and all 922s in
> the field are safely out of warranty. .
>
> >This just doesn't sound like it. There were
> >no tuning anomalies (except the gradual one I described), ... ...
>
> The gradual element is a bit of a puzzler.
Yeah, to me too. If it had been only the plate current reading, I would
have said it was just the shunt failing, but the RF output was also
declining, so I gotta look further.
Thanks for your help.
73, Pete N4ZR
>
> >...no sudden
> >phenomena, and no arcing.
> >
> No one can know the likely cause until you conduct the measurements.
> >>>
> >>>Probably time to think about replacing the tubes!
> >>>
> >>Not unless the grid/filament breakdown voltage is abnormal.
> >
> >Any way to test this without a hi-pot tester?
> >
> yes. With a C-meter. See above.
> >>>
> >>>>Any advice greatly appreciated.
> >>>
> >>>Here is some free advice. When something STARTS to happen, shut down
> >>>immediately to investigate. The SB-220 is a very forgiving amp and often
> >>>gives warning signs....just heed them.
> >
> >But how do you characterize the situation if you don't test and observe?
> >
> - If you want to put the problem to rest, you gotta test.
> Rich---
>
> R. L. Measures, 805-386-3734, AG6K
>
> --
> FAQ on WWW: http://www.contesting.com/ampfaq.html
> Submissions: amps@contesting.com
> Administrative requests: amps-REQUEST@contesting.com
> Problems: owner-amps@contesting.com
> Search: http://www.contesting.com/km9p/search.htm
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm