[Amps] Source of Mica Caps?
Carl
km1h at jeremy.mv.com
Tue Jun 4 09:48:44 EDT 2013
>
> -----Original Message-----
> From: Carl
> Sent: Monday, June 03, 2013 11:26 AM
> To: Jim Thomson ; amps at contesting.com
> Subject: Re: [Amps] Source of Mica Caps?
>
> I have 2 TL-922's on the bench as I type. Both had a tube arc that opened
> a
> grid choke and prevented any further damage with their original 1989
> Eimacs.
> Since those are among the troublesome date codes a replacement tube will
> get
> them back to their owners while directly grounding could have cost a lot
> of
> money.
>
> ### whoa. Tube arc from anode to grid ?? A good 3-500Z will hi pot
> test to well
> over 8 kv. Why would it arc from anode to grid in the 1st place ?
** Whoa nothing. That is a typical gas arc path. Guess its time for you to
ask Rich or Tom how amps work (-;
If it
> arcs from
> anode to grid, the tube is probably toast. Put a 50 ohm, 50 watt wire
> wound in series
> with the B+ lead. Put a high speed HV fuse in series with the 50 ohm, 50
> watt wire wound
> glitch resistor. Put a 2nd HV fast fuse between one leg of HV AC
> sec...and input to FWB.
> Put a rear panel 3agc fuse holder, with a super fast grid fuse. Wire
> grid fuse, so it is
> between chassis and negative terminal of grid meter....or grid shunt.
> Toss the 6 x grid caps
> and both chokes away. Bond the 6 grid pins to the chassis with wide cu
> strap.
** An almost total waste of time and money for the average ham. Replace the
"VHF" choke in the TL-922 and SB-220 with a glitch resistor and go back to
operating. The grid choke is the grid fuse as long it is sized properly.
>
> ## If tube arcs from anode to grid, the 50 ohm glitch resistor will
> LIMIT the fault current
> to a low value. 3000 v divided by 50 ohms = 60 A of fault current.
> The 60A of fault current
> will then blow open the fast HV fuse in under 2 msecs. That has been
> verified on the bench with
> test gear, scope etc. So LIMIT fault current, then INTERRUPT the fault
> current. This same scheme
> not only works on 3-500Z amps, it also works on any other GG tube like
> GS35B, 3x3 and 3x6 tubes etc.
** Plus all Ameritron and other brand direct grounded 3-500 amps which
sustain lots of expensive damage when a tube lets loose.
>
> ## draw too much dc grid current, and 3agc grid fuse blows open. No
> path for dc grid current means
> the amp can't be driven! Power output of amp drops to zero watts. SWR
> between xcvr and input of
> GG amp rises to infinity....and xcvr shuts down asap due to high swr.
> If either HV fuse blows open, or the
> dc grid fuse, simply replace and you are back in business.
>
> I'll let amp owners decide which choice they want and hopefully not be
> swayed by a lot of published nonsense from a tiny minority. If an amp
> becomes unstable replace the parasitic suppressors with something
> resembling
> the up to 40+ year originals and not voodoo nichrome crap. Learn how to
> diagnose and fix a problem and not a symptom.
>
> ## The reason the CC resistors increase in value over time is cuz they
> are subjected
> to a ton of heat from the HOT anode ! The heat will travel down the
> lead...and cook the
> crap outa the 2 watt CC. Then you have the additional extra heat from
> the normal suppressor
> action, since most suppressors will run hot on 10m..and also 6m anyways.
> Using something
> more than 2 watts will allow some leeway and headroom in the design of
> the 10 or 6m
> suppressor.
>
>
>
> Ive been converting amps to 6M since the mid 60's and have gone thru
> several
> suppressor designs; if an old one failed then replace the resistor.
> Grounding the grids "may" work and they "may" also take off with new
> tubes;
> my experience is it "will" happen at times especially with Chinese tubes.
> I
> leaned to the side of stability and consider suppressor resistors as
> consumables just as spark plugs and modern resistors last longer just as
> do
> ignition system components.
>
> ## what various 6m suppressor designs are you talking about now....or is
> this
> just more trade secret stuff ?
>
** Yup. Ive detailed some over the years and got tired of other for profit
types copying them.
> Stop constantly whining about the SB-220 choke voltage drop. It affects
> nothing and if it bothers you that much put an electrolytic across the
> zener.
>
> ## 150 ma of grid current ..per tube is a lot. 300 ma in total on my
> L4B amps.
> Your SB-220 uses grid chokes that have a 25 ohm DC resistance. .150 x 25
> ohms = 3.75 vdc
> of V drop across each grid choke. You only have a 5 volt zener in there
> for bias to begin with.
> On peaks, you will now have 5 + 3.75 = 8.75V of bias on each tube. I call
> that lousy bias regulation.
> BTW, putting a lytic across the 5 V zener will buy you nothing. The
> zemer is already regulated!
** Try to understand that the extra 3.75V is only at peaks which acts as an
AGC control. The zener is far from an ideal regulator as the undersized 10W
versions drift with heat. Amp Supply went with 50W.
>
> ## You claim the V drop across the 25 ohm resistor provides NFB. And
> that bonding the grids to the chassis
> will reduce the NFB to zero, and simultaneously degrade IMD....and also
> increase gain. Heath claims the NFB
> comes about by the V divider action from the tubes grid to cathode C +
> grid caps. Which is it ??
** They dragged that in total from Collins as used in the 30L1 without
understanding why. Drake copied it also in the L-4/4B.
While it provided some level of RF NFB it was far from consistent across the
bands.
If you want to properly ground the grids for RF use a total of .02uF or so
split at the 3 grid pins and use a resistor for the NFB and fuse; use
whatever resistor value floats your boat. Its more work than many want to go
thru.
>
> ## the drake amps use grid chokes that have LESS than 1 ohm DC
> resistance. I measured .91 ohm on 8 grid chokes
> from 4 different drake L4B amps. .150 x .91 = .1365 vdc V drop across
> each drake grid choke. Drake amps run ZERO bias, so the
> bias will fluctuate between zero and a max of .1365 V. These days, I
> bond the grids to chassis, but use regulated bias.
> Like 3-10 x 6A10 diodes, switchable in 3 steps using a DPDT-center off
> toggle on rear apron.
> Plus a huge value lytic across the entire string. Bias doesn't budge on
> ssb/cw.
> I used calibrated watt meters on both the input and output of each
> amp....and did
> a before and after grid grounding with wide cu strap test. 20 - 25 watt
> reduction in drive power when the grids are directly
> bonded to the chassis ! So much for the 25 ohm dc resistance of the
> heath grid chokes providing NFB, they don't. The heath
> chokes + grid caps don't improve linearity either.
** All that shows is that the caps are doing almost all the work at zero
bias and the chokes are along for the ride. If the caps are actually
consistent across the bands then so would be the drive reduction. The Drake
engineer who designed that amp said he measured a 3dB IMD improvement but
the band wasnt specified.
Ive measured 2-3dB on a stock SB-220 but not sure of the band, likely 20M.
On 6M it was the same within measurement error and 3dB is likely the most
that can be squeezed out at that plate voltage. I did not measure in the CW
position that I can remember anyway.
> ## with a 20-25 watt redux in drive requirements, the XCVR will run a lot
> cleaner, IMD wise. ARRL extended lab results for the kenwood
> 870, the xcvr was tested at 50-85-100 watt pep out for the 2 tone imd
> test..on 20M. The imd @ 100w pep out was lousy, -30 db pep.
> it really cleans up when run at 85 w out....and really clean at 50 w out.
> With a 20-25 w redux in drive requirements, the same 870 would be
> operating
> at 75-80 w pep out.....with lower imd. Lower xcvr imd means lower imd
> from the kw linear amp that is being driven by the 870. GIGO concept.
** On the TS-940 around 75-80W is the sweet spot and is why my LK-500ZC only
runs 1200W. OTOH how many hams ever reduce power and also run the
compressor? Net result is zero to a lot more IMD.
** Just listen to any AL-80A, they splatter wide at 100W and the owners care
less.
Carl
KM1H
> Jim VE7RF
>
>
>
>
> Carl
> KM1H
>
>
>
> ----- Original Message -----
> From: "Jim Thomson" <jim.thom at telus.net>
> To: <amps at contesting.com>
> Sent: Monday, June 03, 2013 12:42 PM
> Subject: [Amps] Source of Mica Caps?
>
>
>> Date: Sun, 2 Jun 2013 19:58:58 -0400
>> From: "Carl" <km1h at jeremy.mv.com>
>> To: "Jim W7RY" <w7ry at centurytel.net>, "Randall K Martin"
>> <rkmassoc at comcast.net>, <amps at contesting.com>
>> Subject: Re: [Amps] Source of Mica Caps?
>>
>>
>> I still dont understand the claim made by a few that increasing gain and
>> power output increases stability when hard grounding the grids.
>>
>> I would take various web sites with a grain of salt and try to separate
>> the
>> good from bad.
>>
>> The 115pf caps were chosen by Heath at the end of the SB-220 run and a
>> bulletin suggested changing them. The SB-221 and HL-2200 use the 115pf
>> but
>> Ive replaced with whatever 110 and 120pf I come across in bulk. Ive yet
>> to
>> have instability in a well built and updated SB-220, even on 6M, with the
>> Heath design.
>>
>> Carl
>> KM1H
>>
>> ## TL-922 are rock solid when oem suppressors used and grids directly
>> grounded, and chokes tossed.
>> Ditto with drake L4B..and just about every other amp that floats the
>> grids with caps... including SB-220.
>>
>> ## Sb-220 with grids directly grounded to chassis works superb on 6m.
>> I have seen KM1H SB-220
>> 6m conversions, where the grids were grounded to chassis...after the
>> fact by the owner..works great.
>>
>> ## The 25 ohm DC resistance of the chokes used in the SB-220 isnt doing
>> you any good..just more
>> unwanted V drop. The DC resistance of the chokes used in the Drake
>> L4B is less than 1 ohm.
>> I still tossed the chokes and caps in all 4 of my L4B amps.
>>
>> Jim VE7RF
>>
>> .
>>
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
>>
>>
>> -----
>> No virus found in this message.
>> Checked by AVG - www.avg.com
>> Version: 10.0.1432 / Virus Database: 3184/5879 - Release Date: 06/03/13
>>
>
>
> -----
> No virus found in this message.
> Checked by AVG - www.avg.com
> Version: 10.0.1432 / Virus Database: 3184/5880 - Release Date: 06/03/13
>
More information about the Amps
mailing list