开云体育

R L Measures wrote:
<snip>

### Not quite. The GRID RING is bonded to the chassis !! This
is not a case like Rich using 1/2 watt resistor's as "grid fuses" [
which WILL leave the grid floating]

Tony -- Have you ever blown a grid fusing R and observed what happens?

nope, no grid fuse if the grid is bonded to the chassis directly ;)
I was referring to B- floating off ground but if it isn't, then there isn't an issue.

73, Tony W4ZT

R L Measures wrote:

I suppose you require the same deal on
just balancing TWO tubes ???? [2 x fil chokes + 2 x
blocker's]............ even the 2 x tube senario sucks.

With multi-tube indirectly-heated cathodes, adding cathode RF-NFB is a piece of cake, but with directly-heated cathodes it's a bucket of snakes.

Amen!! So true so true.

I remember building a 2 x 4-1000 amp... with two of

Smart move. If one needs more suds, go for One tetrode with handles.

Thats why I built a 8171! At one point Rich also gave me some help,
now it runs like a charm!

73 Jim SM2EKM

On the "other" reflector a statement was made about TPTG oscillators
and I posted a follow up question about the time of the Great
Migration to this reflector so I never did get an answer. I'd like to
ask it again here.

The 'Gentleman From Georgia' (you know who) made the comment that a
TPTG oscillator will not oscillate if the grid tank is tuned higher in
frequency than the plate tank. Under that condition, apparently the
phase of the fed-back voltage is wrong and oscillation can not occur.

Since this is essentially what a VHF parasitic oscillator is, it
caught my attention.

Is this statement true? If so, then it would seem that all one has to
do to achieve VHF stability is to make the two resonances meet the
above requirements. Can it be that simple?

TPTG oscillators - the intentional kind - have not been used in ham
circles for decades so I wouldn't be surprised if knowledge of them
has been pretty much lost.

All comments welcome.

73, Bill W6WRT

On Sep 29, 2006, at 4:16 PM, pentalab wrote:

--- In ham_amplifiers@..., R L Measures <r@...> wrote:


RICH SEZ.. The V-drops across the cathode resistors is what helps to
equalize the cathode currents in the 4, 811As. However, 4,
individual bifilar filament chokes and 4. DC blocker/coupling caps
are needed.

### Well... that sucks !!

That's the awful reality of adding cathode RF-NFB with thoriated tungsten tubes in parallel

I suppose you require the same deal on
just balancing TWO tubes ???? [2 x fil chokes + 2 x
blocker's]............ even the 2 x tube senario sucks.

With multi-tube indirectly-heated cathodes, adding cathode RF-NFB is a piece of cake, but with directly-heated cathodes it's a bucket of snakes.

I remember building a 2 x 4-1000 amp... with two of
everything in it.... fil xfmrs... fil chokes... 2 x fil variacs...
separate adjustable bias....... and 3 x plate + 3 x grid current
shunts..... Tube 1.... Tube 2.... Tube 1+2. Then cross ur
fingers... and hope the drive split's into two equal parts...
ditto with airflow.

After that... I said I'd never build / design another 2+ tube
linear again..... I didn't.

Smart move. If one needs more suds, go for One tetrode with handles.

Just trying to trbl shoot a 4 x tube
amp is not fun. Still, I did have a 4x 811A linear.... and
the tubes never gave me any problems........ the rest of that
Dentron amp was pure junk.

Amen, Jim. Dennis Had is reportedly now in the AF amp business -- which is where he pretty obviously should have gone in the first place.

later.... Jim VE7RF








Yahoo! Groups Links

R. L. Measures, AG6K, r@somis,org, 805-386-3734

On Sep 29, 2006, at 4:53 PM, pentalab wrote:

--- In ham_amplifiers@..., R L Measures <r@...> wrote:

RICH SEZ...Now the bad news: Without W8JI, the "recognized
amplifier expert", Charles Thomas Rauch, Jr., and his statements
such as the laws of AC circuit analysis do not apply to VHF
suppressors; and VHF resonant circuits can't ring, but HF ones can,
things will definitely be less entertaining.

### this is one soap opera I don't wanna ever hear again!

Why read it more than once? Before I started the grate parasite debate with Tom I was warned by a Ham in Manhattan that he wins debates in the eyes of his groupies by never backing down, even on issues that, to a RF-savvy person, make him look the clown. In my opinion, this is probably something he picked up from his father, Charles Thomas Rauch, Sr.
- editorial - If I go to my grave with a bunch of newbies believing Tom's statement that Ni-Cr alloys have reverse RF skin-effect, VHF resonances can't ring, et cetera, I will not be resting in peace - and that's for damn sure. .

### The best way to solve parasitic problems in a TL-922 SB-
220/221 is to REMOVE all 6 x 200 pf bylass caps + 2 x RF
chokes from sockets... and toss em !! Strap all 6 x grid
pins to chassis ..with wide strap..... end of problem.

The grid resonance in a stock TL-922 is in the high 80MHz range. The grid resonance with the grid pins strapped to ground is typically about 1MHz lower. How does this make the amplifier more stable at the 120MHz resonance in the anode circuit?

Rich, AG6K

On Sep 29, 2006, at 6:05 PM, pentalab wrote:

--- In ham_amplifiers@..., Tony King - W4ZT <w4zt-
060920@...> wrote:

pentalab wrote:

Fellows

I have several fuses installed in most of these linear's.

Like a fast 750 ma 3agc style grid fuse... located between
chassis.. and negative terminal of 0-1 Amp grid meter.

TONY SEZ... I'd protect this with diodes and leave that fuse out.

It wouldn't be a good idea for the cathode to ground circuit to
open and essentially float the grid.

### Not quite. The GRID RING is bonded to the chassis !! This
is not a case like Rich using 1/2 watt resistor's as "grid fuses" [
which WILL leave the grid floating]

Tony -- Have you ever blown a grid fusing R and observed what happens?

On Sep 30, 2006, at 12:16 AM, pentalab wrote:

--- In ham_amplifiers@..., "badgerscreek" <qrp73@...>
wrote:

So how do you come up with the inductance and resistance value for

the suppressor? Maybe you can give us some insights how you go about

designing a suppressor, certainly one size would not fit all? Can

you place a picture of your 3cx6000 suppressor on here?

### I think u can just see it in the pix. I posted a ton of pix
on "photo's".... just click on my VE7RF photo album..... it's got
the Coaxial dynamics 25 kw slug [for a 7/8" line section no
less.] on the front of the photo album.

### I do em by trial and error. On the 3000A7, I used the one
turn chrome plated suppressor from the 8K ultra. [it has NO
resistor]. Last year, Henry radio tells me .."the chrome did
nothing.. so we went back to silver plate strap". Now they are
telling me.. the LAST bunch of 8-K ultra's that left the factory
[b4 they lost their lease].. had NO suppressor's !! Turns out
the "parasitic problem" was a plate choke problem all along !!

The 3cx3000A7 has 0.6 pF of feedback C. That's 6x more than the 8877 and 4x more than a 3-500Z. At its max-f rating of 110MHz, the
3cx3000A7 has an output to input feedback path with 2500-ohms of XC.

I figured as much... since their 2 x piece plate choke uses 20 uh
+ 180 uh. [the 8-k uses a 3000A7] In MOST cases... on a HB
3000A7... u can just toss the suppressor.

... especially if you just love to repair amplifiers.

### On the 6000A7.... I was clueless. The 11-m boyz are using
250 W globars !!! I looked at JA6TAY's stuff... and he uses
flat strap... wound directly onto globars. You can only get 3
turns of 3/4" wide strap on a 5" long globar anyway.... so I
found that [50 ohms]... worked.. rock stable. Next trick... we
will remove it alltogether.. and replace with 1" wide flat strap..
and see what happens.

### I'm convinced... 98% of stability is how well the control grid
is grounded. A 3-500Z has only ONE internal grid pin....

It's a 360? cone shaped collet.

which
then splits into 3 x external pins... which bought em nothing.

If the 3-pins are grounded directly, the resonance at the grid is in
the 80MHz range. If the 3 grid pins are grounded through 200pF caps,
the grid resonance is in the 80MHz range. In my experiences it is better to measure resonance with a dipmeter than to guess at it.

These big triodes have huge 4.25" diam [almost 13"
circumference] grid rings on a 3000/6000. The grid ring on a YC-
156 is even bigger.... 5.25" diam [almost 16" circumference] !

### W8JI once modified a small GG metal triode... improved the grid
grnding... and built the worlds only [as far as I know] 160-6m
linear.... with NO parasitic suppressor !!

So why does the AL-1500 - that he apparently designed - have a
reputation for sudden 8877 failures? As I see it, an amplifier that
appears to be stable during an hour or so of testing is not
guaranteed to be stable in the long run.

### In any event...in yrs gone by.. I always start with a 50 ohm
globar.... and keep adding turns[ 1 1/2... 2...2 1/2....3] till the
parasitic goes bye bye. If using a real small globar... and u run
outa room.... reduce the width of the strap.

Smooth move. Reducing the width of the strap increases RF-
resistance, which lowers the L/R suppressor's VHF-Q, which in turn
lowers VHF gain and reduces the chance of VHF oscillation. Another
way of increasing RF-resistance is to choose a conductor that is more
resistive than copper. Example -- The 8169 / 4cx3000A amplifier at:

appears to have no parasitic suppressor whatsoever, however the anode
to blocking-cap connector strap, as well as the connection to the
Bruene-bridge neutralizing cap is made out of a nickle-chromium
alloy. According to (W6HW) - a friend who worked for Collins Radio
Co. - the 30S-1 amplifier used the same technique.
- Murphy was right: Things may be more complicated than they look.

What other mods have you made to the L4B? You still running the

stock power supply?

#### Sorta... same plate xfmr. I added external step start
which steps starts the HV + the fil xfmr. [25 ohm-100 w in ONE
leg]. I also disconnected the 2 x 50k- 100 w bleeders [ + the 7 k-5
watt resistor in series with em...used for the +130 V RX cut off
bias]. 99% of the heat from the top of the L4B/L7 HV supply
is from the 70 w CCS from those bleeders !

But it's a great coffee or soup-warmer.

Rich, AG6K

On Sep 30, 2006, at 12:22 AM, pentalab wrote:

--- In ham_amplifiers@..., "craxd" <craxd1@...> wrote:

All,

Hsu posted this at the other group. To be made in China, it looks a
lot like an AL-811 if you ask me. BTW, did anyone ever ask Hsu to

the

group here?



Best,

Will

### Will.... I'm gonna nonminate you as... "project
manager". ...so you can round up all these .."stragglers".... and
bring em over ...to the dark side.

Try posting this on AMPS to see if Tom is staying awake (he censors it).

Jim VE7RF







Yahoo! Groups Links

On Sep 29, 2006, at 4:56 PM, Tony King - W4ZT wrote:

pentalab wrote:

Fellows

I have several fuses installed in most of these linear's.

Like a fast 750 ma 3agc style grid fuse... located between
chassis.. and negative terminal of 0-1 Amp grid meter.

I'd protect this with diodes and leave that fuse out. It wouldn't be a
good idea for the cathode to ground circuit to open and essentially
float the grid.

When a high-R exists in the cathode circuit, as soon as a little current begins to flow, the tube cuts off. My modified SB-220 and TL-922 do this, and if a tube ever happens to short, nothing is damaged.
- note - In a stock SB-220 or 922, a shorted tube can destroy the filament transformer in short order.

Also have a 3 A cathode fuse [also a 3agc] in the center tap of
the fil xfmr circuit [ along with the 60 x 6A10 6-A , rotary
switched bias diodes. ] This 3 A cathode fuse has a 1k-25
watt ressitor across it.

1k is a little low, use something like 15K to 20K across the fuse. When
the fuse blows you will have an increase in bias that will cut the tube
off and will develop a voltage proportional to that needed to cut it off.

With 3-500Zs, 1k-ohm will limit anode current to about 25mA when the fuse opens.
- editorial - A 250v-rated fuse in a circuit operated by 3000v is not good engineering practice.

Question is... what's an easy way to obtain a visual
indication /led/neon/etc.... to let us know said fuses are
blown ???

If the cutoff bias is high enough, a neon bulb with a series resistor
could be used as an indicator.

It takes c. 90v to light neon. A LED takes c. 1v. Also, with 1k=ohm and 3-500s, 25mA means the fuse is open, so one does not need an idiot light. Also, when the grid-I meter deflects backward, a tube has a filament that is shorting to the grounded-grid.

... ... ...

For some added protection... we are thinking of an adjustable
spark gap between load cap ...and chassis..... set to fire at a
V higher than the normal low swr V.... and a point well below the
rating of the 5 kv vac load cap. I think Rich may have done
this ?

Correct

I don't want the expensive ceramic vac load cap to ever
internally arc.

Spark gaps are good!

However, adding a rugged low-ohm low-L resistor in series with the spark gap helps to limit peak-I during a glitch.

Sri for the drivel....... does anybody actually read any of my
posts....... or is this all old news ???

Later... Jim VE7RF

I enjoy reading your posts Jim!

Best thing I have seen is to use one of the triode control boards and
use the protection circuitry they provide. You will find there's hardly
any failure that will can not be detected and reacted to almost
instantly. You might have to make a few adaptations to meet the need for
QRO++ but the principle is the same. This is especially good because it
can trip the amp off line so you don't have high drive going into a
malfunctioning amp.

73, Tony W4ZT



Yahoo! Groups Links

--- In ham_amplifiers@..., "craxd" <craxd1@...> wrote:

All,

Hsu posted this at the other group. To be made in China, it looks a
lot like an AL-811 if you ask me. BTW, did anyone ever ask Hsu to

the

group here?



Best,

Will

### Will.... I'm gonna nonminate you as... "project
manager". ...so you can round up all these .."stragglers".... and
bring em over ...to the dark side.

Jim VE7RF

--- In ham_amplifiers@..., "badgerscreek" <qrp73@...>
wrote:

So how do you come up with the inductance and resistance value for

the suppressor? Maybe you can give us some insights how you go about

designing a suppressor, certainly one size would not fit all? Can

you place a picture of your 3cx6000 suppressor on here?

### I think u can just see it in the pix. I posted a ton of pix
on "photo's".... just click on my VE7RF photo album..... it's got
the Coaxial dynamics 25 kw slug [for a 7/8" line section no
less.] on the front of the photo album.

### I do em by trial and error. On the 3000A7, I used the one
turn chrome plated suppressor from the 8K ultra. [it has NO
resistor]. Last year, Henry radio tells me .."the chrome did
nothing.. so we went back to silver plate strap". Now they are
telling me.. the LAST bunch of 8-K ultra's that left the factory
[b4 they lost their lease].. had NO suppressor's !! Turns out
the "parasitic problem" was a plate choke problem all along !!
I figured as much... since their 2 x piece plate choke uses 20 uh
+ 180 uh. [the 8-k uses a 3000A7] In MOST cases... on a HB
3000A7... u can just toss the suppressor.

### On the 6000A7.... I was clueless. The 11-m boyz are using
250 W globars !!! I looked at JA6TAY's stuff... and he uses
flat strap... wound directly onto globars. You can only get 3
turns of 3/4" wide strap on a 5" long globar anyway.... so I
found that [50 ohms]... worked.. rock stable. Next trick... we
will remove it alltogether.. and replace with 1" wide flat strap..
and see what happens.

### I'm convinced... 98% of stability is how well the control grid
is grounded. A 3-500Z has only ONE internal grid pin.... which
then splits into 3 x external pins... which bought em nothing.
These big triodes have huge 4.25" diam [almost 13"
circumference] grid rings on a 3000/6000. The grid ring on a YC-
156 is even bigger.... 5.25" diam [almost 16" circumference] !

### W8JI once modified a small GG metal triode... improved the grid
grnding... and built the worlds only [as far as I know] 160-6m
linear.... with NO parasitic suppressor !!

### In any event...in yrs gone by.. I always start with a 50 ohm
globar.... and keep adding turns[ 1 1/2... 2...2 1/2....3] till the
parasitic goes bye bye. If using a real small globar... and u run
outa room.... reduce the width of the strap.

What other mods have you made to the L4B? You still running the

stock power supply?

#### Sorta... same plate xfmr. I added external step start
which steps starts the HV + the fil xfmr. [25 ohm-100 w in ONE
leg]. I also disconnected the 2 x 50k- 100 w bleeders [ + the 7 k-5
watt resistor in series with em...used for the +130 V RX cut off
bias]. 99% of the heat from the top of the L4B/L7 HV supply
is from the 70 w CCS from those bleeders ! After disconnecting
the series strap between em [ I physically left em in]... I was
expecting the no load HV to skyrocket... it didn't move up at
all... even a needle width !

## also replaced the 8 x 100 K 2 watt carbon comp Equalizer
resistors on the lytics. [they were STILL 100 k each..since
1977 !].... with 3 watt 100 k from Rich.

### One diode finally ate itself.. so replaced em all with
1N5408's [ I buy 1N5408's and it's big brother, the 6A10 (1 kv-
6A) by the hundreds].

### also added 160m to one of the L4B's. Easy 160m mod... easier
than a SB-220. The stock Bifilar is wound on 5//8" rod.... and
measures 28 uh... plenty for 160m [ the 220 only has a 10 uh
bifilar] I used the rest of Rich's SB-220 160m mod. The tuned
input consists of 1600 pf arco compression trimmers... padded
with silver mica's... and I think.. a T-50 coil... works slick.
Some 100 pf HEC doorknobs pad the tune... and some more pad the
load. The STOCK L4B plate choke measures 154 uh... plenty for
160m.... just add some 4700 pf disc ceramics at the base.

## I was goona use 3 x T-225-2B's for the 160m tank coil. [BEWARE
the 2-B's are 1" thick... the 2-A's are only 1/2" thick]
Laminated 3 of em = 2.25" OD x 3" long... and heavy. Dumped this
idea.. as Marv, wa6cw worked out the losses at a whopping 57
watts.. whether 1-2-3 cores ussed !! Found some airdux.... then
found some more airdux... ame ga wire... but CLOSER spaced between
turns. The airdux coil took up LESS room than the torroids.. and
zero heat... no losses !

### since the L4B's have a built in wattmeter.. I converted em to
PEP reading [which does the grid current a well !!]

### also, the L4B's are designed to handle 2 kw on
BYPASS...BEWARE.. the ONLY way to calibrate these wattmeter's is
to use another amp in front of it.. in series to provide 1 kw !!

### since that's the case.... plus I hook up all 3 of my L4B's in
SERIES.. nose to tail... tuned up on different bands... when I added
the RJ1-A qsk mod.... I didn't use Rich's idea of a reed relay
on the input... but opted for a 2nd RJ1-A instead.

### Note... GIGAVAC sells their EQ of a RJ1A + the next size
up... + a dpdt ceramic vac relay cheap.. new., etc, plus more
relays.. to hams who can provide proof of their license, part of
their "ham program".




One mod which I made which corrects the very lame blower,

### It's only a 1550 rpm unit. A 1800-2000 rpm would be ideal...
more air... with out too much more noise.

is just reducing the chimney size slighly, now you can actually

feel air coming out of them! Its a 90mm frosted Glass Chimney made
by Camping Gaz. The Coleman Chimneys are too big and the cover wont

close properly.

### say what ?? The stock chimney's are 4" tall x 4" diam. My
top cover closes just fine. The present day, current Coleman
chimney that's closest to it is I think either bigger diam.. OR
taller. Are ur's smaller diam than stock.... do they still fit
the clip retainers??? Or or they just shorter ?? At least they
used the correct anode connector's.. with vertical fins..... I don't
know what the hell Eimac was thinking of.. with their infamous HR-
6 / HR-8 anode connector's !!


I did a A/B chimney tests on the tubes and the

narrower one runs cooler. I tried it on both tubes, my opinion is
that the Stock drake glass chimneys are a bit big for the lame

blower

### The top cab really restricts air too. Remove it.. and loads of
air.

### The L4B runs NO bias !! Idles at 230 ma @ 2650 V = 600
watts...= 185 deg C top cab lid temp = plane nuts !

### The fix is... install 10 x 1N5408's on a perfboard... in the CT
of the fil xfmr. Idle drops to 100 ma on ssb [2650v]... and just
40 ma on Cw [1900v] IMD is superb on either voltage. Now...
here's the kicker.... since the Eimac 3-500Z is only Mu=130...
and my Eimac 3-400z's have a higher MU=200.... all the 200 Mu tubes
will idle LOWER anyway.... so when switching tubes around.... I made
the new bias diodes adjustable.... by installing a minature SPDT..
CENTER OFF toggle on rear apron. That one toggle will give 3 x
positions of bias.... all 10 diodes... or just 3 of em... or 6 of
em... slick ..and works good. Install at least a 1000-2000 uf
lytic across the ends of the 10 x diodes... for superb bias reg.

I do have a Dayton blower that will mount to the back, but i hate
noise and i dont want another Henry Vacuum Cleaner in the shack!
Thats why the TL922 and SB220 are such good amps for the money and

the noise factor. certainly 1kw is enough unless you running a attic

mounted Isotron!

### agreed. I did have 3 x Isotrons on my patio on the 4th floor
of my old condo..... worse ant ever.. a real struggle. We did have
208 /120 v single phase power in the condo... and I was gona modify
one L4B so the fil xfmr ran on 120 V... but the plate xfmr ran on
208 V.. which would of course give slightly lower plate V. I
ended up moving.. and the 3 x isotrons were sold..... If I had of
stayed much longer.... I woulda vapourized em in the end.

### 1 kw is plenty... esp with summer time heat. I design the QRO
stuff as a lark... just to see if it can be done.. and could be
built.. and made to be stable.... then boil 8 gals of oil with it !

### One big metal triode is just as easy to build, in fact
easier... than 3 x 4CX-800A7's....in a shoe box.

### Yrs ago... I looked at the price of a 8877.... and thought... a
pair of 8877's would be plane nuts. A 8877 is not much less than
a new 3000A7. The 3000A7 can be rebuilt over and over... bomb
proof.. esp with it's 225 W CCS grid. Then it escalated from
there, you end up using the same vac tune/load caps etc....
slightly bigger box.... and HV supply.

### what I find really intriquing with HB stuff is I start with a
blank sheet of paper.... then start designing.... and not around
what's currently in my junk box either. Shop carefully on the
internet /Fair radio/ Allen bond, here, etc.... and you can minimize
costs. What I end up with 100% of the time, blows away anything

from Henry or Alpha, QRO, etc. Build it urself.. and you can

make any, or all components as heavy duty as you want. Use any
style analog/digital meter you want... any size KVA plate xfmr, nice
cab's etc.. u get the idea.

### I also like to experiment... try new concepts... and really push
the envelope. I like to see how much is "too much"... and find
these points of "diminishing returns" everyone talks about.

### I fool with AM once in a while.... and we are allowed 750 w
of CXR.... measured at the FEEDPOINT of the ant ! That equates
to 3 kw out pep [100% mod].... and 3750 w pep [125% mod] and 4500
w pep [150% mod]. A 8877 won't do any 3 kw out pep... never
mind 4500w pep. A single 3000A7 will meet these needs.

### I also fool extensively with ESSB. Some of the fellows are
experimenting with Class A linears for ESSB. Now Class A is
only aprx 25% eff.... so u need LOADS of anode diss.

### Our legal limit here in Canada is 2250 w pep on SSB [again
measured at the feedpoint] Again, a 8877 wouldn't do the job.
Just .5db of feedline loss = 10% loss..... so 2.5 kw from the amp
is needed... just to get 2250 w at the ant !

### Now imagine running ESSB Class A.. with 25% eff.. to get
2500 w out.... you need 10 kw dc input = 7.5 kw of anode
diss ! Start to get the picture ?

### That's when I got into the variable bias scheme.... run zero
or almost no bias on a 3000 A7 or esp a 6000A7... and high plate
V.... and watch the idle power go through the roof. You could
almost load test a HV supply that way.

### It's always "one big, never ending work in progress " for
me. Just trying to add 160m or warc bands is a test of skill...
and nerves.


### It might be a labour of love........Or a love of labour !!

Later... Jim VE7RF

Greg

Later...... Jim VE7RF

--- In ham_amplifiers@..., Tony King - W4ZT <w4zt-
060920@...> wrote:

TONY SEZ.. And that still takes us to electronic protection...

### My concern with electronic protection is.... if it
malfunctions... when u need it ! I have seen these high energy
diverter's.. [used to crow bar HV supplies, used in commercial
large HV supplies like John Lyles describes at times]
malfunction.... then u have the charred remains of a huge diode
stack !

### Having said that... it would be simple to incorporate two
of everything... redundant protection schemes... and
incorporate "test switches" on em..... so u can actually test em
on site... without having to drive ur expensive tube into 'too
much plate + grid current'

### I have looked at both triode boards in the past. Some of it

would need extensive mods, etc.

But the price is nominal for them... use the pieces you need. Take

a little time looking at Paul Hewitt's board (WD7S). He has done
some nice work and all his trimmers are multi turn pots.

## agreed. He did have one UNIQUE twist. His board would sense
the HV [ I think from the HV multiplier resistor's] and if NO
hv present.... would inhibit the T/R relay's......very slick.

### IF my HV fuses blow... the amp is still online, being
driven.... and a split second later, the grid fuse blows. His
scheme would keep it shut down... until the HV was XXX volts.


TONY SEZ .. Paul does do some neat sensing and comparing input and
output. worth looking at!

### I'm gonna check his site again. I really don't like the SS
relays in the 240 V primary. To work right... they should be zero
cross Voltage on turn on.... and zero CURRENT cross on shut
down. The 80 A rated ones I have seen require aprx a 100+
square inches of aluminium plate [very thick].. as the heatsink.
They also have leakage across em. In Canada, they can't be used as
the sole disconnect... they gotta be supplemented by at least a
fuse. My electrician buddy got zapped last yr from one... from the
leakage.

### Like u say.... the SS relays/ mech relays in the primary are a
poor way to go, to shut off ... "follow on energy".... and totally
useless to eliminate the B+ energy stored in a bank of lytics.

### Come glitch time, the idea of opening up the primary 240 V
with a SS device freaks me out... ditto with a contactor... even a
big one. Notice on the backs of 240 V breakers .. like the
P+B "controlled magnetic hydraulic breakers" [ these things are
500% better/faster than any standard heat activated breaker] they
all have slots on the rear. My buddy sez they are... 'arc chutes'
designed to divert the arc out the back.. and away from the
contacts. Most breakers will have a Max KA interupt rating on
em... like 100 ka, most are designed to handle a dead short....
and assume the supply line from the street WILL pump out a huge
amount of fault current.

### Even the fast breakers in the 240 primary are not fast enough
during a glitch. To eliminate the "follow on energy" from the
xfmr + diodes + caps..... I used the pair of sandfilled HV
fuses. 7900V /50 ohm glitch R = 160 A 160A flowing
through a 3 A rated HV sandfilled fuse blows extremely fast...
and quenches the arc asap... as the sand turns to glass!! The
other reason the fuses are filled with sand is to void 95% of the
air inside the fuse to begin with. [little or no air left to ionize]


### The 100 A breaker in the 240 V line always remains intact...
the faster fuses always beat it to the punch.

Love reading your technical details!

### Tnx.... none of it is rocket science... just a lot of
experimenting over the yrs... and bumbling through things. I still
contend.. most of this stuff coulda been done at least 30 yrs ago.

### My buddy phones me last night in an uproar... his 15 kw
linear is putting out zero watts... and input swr has
risen........ turns out it was the grid fuse gone open...... from
the day b4.... when he just about vapourized his 15 m ant !

### QRO is one thing..... ALL this stuff downstream is another.
His 160m GP has a 30 ohm Z..[his mfj shows 1.7:1 swr right at
the feed point.] He wants to use a 4:1 balun on it... using a T-
200 core !! ["to match it"] The simple solution was to use an L
network... minus the cap of course. A simple 4.9 uh coil
hooked directly across the feedpoint coax resulted in a flat swr.
F-12 calls it a hairpin... Hy-gain calls it a beta match. It works
on yagi's..... and now it also works on verticals + Groundplanes....
simple... one component. Now we remove the temp 12 ga 4.9 uh
coil.... and replace it with a permanent 1/4" OD copper tubing
coil....nothing to blow up.... end of problem..... end of story.

Later... Jim VE7RF

73, Tony W4ZT

All,

Hsu posted this at the other group. To be made in China, it looks a
lot like an AL-811 if you ask me. BTW, did anyone ever ask Hsu to the
group here?



Best,

Will

pentalab wrote:
<snip>

### Not quite. The GRID RING is bonded to the chassis !! This
is not a case like Rich using 1/2 watt resistor's as "grid fuses" [
which WILL leave the grid floating]

agreed

### same deal on a YC-156 and also a socketed 3000/6000.

yeah... and how sweet it is!

### I use several 6 A diodes between chassis and B-... [and also across all the meter's] . When the grid fuse blows the cathode can't float more than +/- .7 V.

That's good and also the 1K safety resistor makes it all good ;)

Also have a 3 A cathode fuse [also a 3agc] in the center tap of
the fil xfmr circuit [ along with the 60 x 6A10 6-A , rotary
switched bias diodes. ] This 3 A cathode fuse has a 1k- 25
watt ressitor across it.

TONY SEZ a little low, use something like 15K to 20K across the

fuse. When the fuse blows you will have an increase in bias that will cut the tube off and will develop a voltage proportional to that needed to cut it off.
### partially agreed. Trbl is... you can STILL drive it.. and get
say 40% output [Class C] I tried a 10-100k across the grid fuse as well. IF the grid fuse blows.... the bias developed will do exactly the same thing as if the cathode fuse blows. [cathode fuse with 1k to 100 kw across it] u still get 40% output.

And that still takes us to electronic protection...

### You can remove the resistor across the grid fuse with NO problem at all. Works better too. With no path for dc grid current... u get zero watts out. Still don't know where the 800+
watts of drive has gone ??

Question is... what's an easy way to obtain a visual indication /led/neon/etc.... to let us know said fuses are blown
???

TONY SEZ if the cutoff bias is high enough, a neon bulb with a

series resistor could be used as an indicator.
### I'm gonna measure the cutoff bias on the existing 100 k 2 w mf
in the vac t/r set up.... on RX. and measure it.

continued below...

I enjoy reading your posts Jim!

## tnx. I didn't know whether to take my marbles home or not.LOL

Best thing I have seen is to use one of the triode control boards and
use the protection circuitry they provide. You will find there's hardly
any failure that will can not be detected and reacted to almost instantly. You might have to make a few adaptations to meet the need for
QRO++ but the principle is the same. This is especially good because it
can trip the amp off line so you don't have high drive going into a
malfunctioning amp.

### I have looked at both triode boards in the past. Some of it would need extensive mods, etc.

But the price is nominal for them... use the pieces you need. Take a little time looking at Paul Hewitt's board (WD7S). He has done some nice work and all his trimmers are multi turn pots.

As far as a grid over

current device goes... the ones depicted in the handbooks work very
well... albeit they all add varying bias.... cuz of the grid current flowing through a resistor [ vdrop used to trigger the 2n222
etc... then a 4pdt latching relay.... hotswitching amp offline,
latching to itself... turning on a led etc. ] I have seen and built
Orr's plate over current device same deal. [ just don't use the
vac relay.. used to open off the HV ]
### During a "glitch".... it's a whole nuther ball game. They all use series diodes... such that during normal operation... slightly excess grid/plate current will simply kick amp offline. During a
glitch.... with huge currents flowing... the safety diodes on the plate/grid protection devices will turn on..limiting voltage to a
safe value.. so the grid/plate overcurrent protection device doesn't
get fried.
## IF plate grid current are slightly over XXX.... the fast fuses
take a long time to blow. During a glitch.... the fuse blow REAL
fast.. and protect the tubes grid, etc. The eletronic stuff...
while fast... still has to activate a mech relay. U can get small
mech relays that will op in <2 msecs however.
### I'm going to try and incorporate BOTH fast fuses and electronic grid/plate current protection... I'll let u know.

I think you're thinking in the right direction. Paul uses solid state relays in primary leads but I'd think they would be difficult to find and very expensive for really big stuff. Of course they can open the circuit quicker than anything else. Only problem is, with that huge capacitor bank, speed on the primary might not be the saving grace.

### also a 2nd high reflected power kick out device b4 the big amp would help.... help to kick the IPA + xcvr off line... IF the
grid fuse in the big amp blew.

Paul does do some neat sensing and comparing input and output. worth looking at!

Love reading your technical details!

73, Tony W4ZT

--- In ham_amplifiers@..., FRANCIS CARCIA <carcia@...>
wrote:

I put a 50K 10 watt resistor across my fuse in the fil. C.T. so

the bias is at cut off if it ever blows.

### agreed. Basicly anything from 1 k to 1 meg is fine. When
the CATHODE fuse blows..... the amp can still be driven a bit....
so a 2-10 w resistor will work fine.

### IF u use a CATHODE fuse....make damn sure u stick a
resistor across the fuse holder... like u described.

### I once had a 4-1000.... and while re-installing it...
forgot to hook up the B- to the pos junctions of the plate +
grid meter's........ a real disaster !! The cathode will try and
assume full plate V... and the bypass caps at the cold ends of
the fil choke starting snapping away !!

### also... in schemes like the L4B... where they use +130 vdc
[and +90 vdc on cw position] to cut off the tube[s] on RX is
bad news. When the 3pdt t/r relay is in 'mid air'... the
cathode is floating for a split second ! I rewired it with a
simple 100 k-2 w mf in the center tap.... and use the same
contacts... to just short out the resistor on TX.

gotta run.... Jim VE7RF

--- In ham_amplifiers@..., Tony King - W4ZT <w4zt-
060920@...> wrote:

pentalab wrote:

Fellows

I have several fuses installed in most of these linear's.

Like a fast 750 ma 3agc style grid fuse... located between
chassis.. and negative terminal of 0-1 Amp grid meter.

TONY SEZ... I'd protect this with diodes and leave that fuse out.

It wouldn't be a good idea for the cathode to ground circuit to
open and essentially float the grid.

### Not quite. The GRID RING is bonded to the chassis !! This
is not a case like Rich using 1/2 watt resistor's as "grid fuses" [
which WILL leave the grid floating]

### same deal on a YC-156 and also a socketed 3000/6000.

### I use several 6 A diodes between chassis and B-... [and
also across all the meter's] . When the grid fuse blows the
cathode can't float more than +/- .7 V.

Also have a 3 A cathode fuse [also a 3agc] in the center tap

of

the fil xfmr circuit [ along with the 60 x 6A10 6-A ,

rotary

switched bias diodes. ] This 3 A cathode fuse has a 1k-

25

watt ressitor across it.

TONY SEZ a little low, use something like 15K to 20K across the

fuse. When the fuse blows you will have an increase in bias that
will cut the tube off and will develop a voltage proportional to
that needed to cut it off.

### partially agreed. Trbl is... you can STILL drive it.. and get
say 40% output [Class C] I tried a 10-100k across the grid
fuse as well. IF the grid fuse blows.... the bias developed will
do exactly the same thing as if the cathode fuse blows. [cathode
fuse with 1k to 100 kw across it] u still get 40% output.

### You can remove the resistor across the grid fuse with NO
problem at all. Works better too. With no path for dc grid
current... u get zero watts out. Still don't know where the
800+ watts of drive has gone ??

Question is... what's an easy way to obtain a visual
indication /led/neon/etc.... to let us know said fuses are
blown ???

TONY SEZ if the cutoff bias is high enough, a neon bulb with a

series resistor could be used as an indicator.

### I'm gonna measure the cutoff bias on the existing 100 k 2 w
mf in the vac t/r set up.... on RX. and measure it.

continued below...

I enjoy reading your posts Jim!

## tnx. I didn't know whether to take my marbles home or not.LOL

Best thing I have seen is to use one of the triode control boards

and

use the protection circuitry they provide. You will find there's

hardly

any failure that will can not be detected and reacted to almost
instantly. You might have to make a few adaptations to meet the

need for

QRO++ but the principle is the same. This is especially good

because it

can trip the amp off line so you don't have high drive going into

a

malfunctioning amp.

### I have looked at both triode boards in the past. Some of it
would need extensive mods, etc. As far as a grid over
current device goes... the ones depicted in the handbooks work
very well... albeit they all add varying bias.... cuz of the grid
current flowing through a resistor [ vdrop used to trigger the
2n222 etc... then a 4pdt latching relay.... hotswitching amp
offline, latching to itself... turning on a led etc. ] I have seen
and built Orr's plate over current device same deal. [ just
don't use the vac relay.. used to open off the HV ]

### During a "glitch".... it's a whole nuther ball game. They all
use series diodes... such that during normal operation... slightly
excess grid/plate current will simply kick amp offline. During a
glitch.... with huge currents flowing... the safety diodes on the
plate/grid protection devices will turn on..limiting voltage to
a safe value.. so the grid/plate overcurrent protection device
doesn't get fried.

## IF plate grid current are slightly over XXX.... the fast
fuses take a long time to blow. During a glitch.... the fuse
blow REAL fast.. and protect the tubes grid, etc. The eletronic
stuff... while fast... still has to activate a mech relay. U can
get small mech relays that will op in <2 msecs however.

### I'm going to try and incorporate BOTH fast fuses and
electronic grid/plate current protection... I'll let u know.

### also a 2nd high reflected power kick out device b4 the big
amp would help.... help to kick the IPA + xcvr off line... IF
the grid fuse in the big amp blew.

Take Care........ Jim VE7RF

73, Tony W4ZT

I put a 50K 10 watt resistor across my fuse in the fil. C.T. so the bias is at cut off if it ever blows.

Tony King - W4ZT wrote:

pentalab wrote:

Fellows
I have several fuses installed in most of these linear's. Like a fast 750 ma 3agc style grid fuse... located between chassis.. and negative terminal of 0-1 Amp grid meter.

I'd protect this with diodes and leave that fuse out. It wouldn't be a good idea for the cathode to ground circuit to open and essentially float the grid.

Also have a 3 A cathode fuse [also a 3agc] in the center tap of the fil xfmr circuit [ along with the 60 x 6A10 6-A , rotary switched bias diodes. ] This 3 A cathode fuse has a 1k-25 watt ressitor across it.

1k is a little low, use something like 15K to 20K across the fuse. When the fuse blows you will have an increase in bias that will cut the tube off and will develop a voltage proportional to that needed to cut it off.

Question is... what's an easy way to obtain a visual indication /led/neon/etc.... to let us know said fuses are blown ???

If the cutoff bias is high enough, a neon bulb with a series resistor could be used as an indicator.

continued below...

Most of the "visual indicating" type fuse holders I have seen... all require use of the line voltage being fed into em... ie: 120 vac, etc. I was thinking.. perhaps a 1-2 ohm resistor across each fuse holder... such that the v drop across the resistor [with fuse open] will trigger a led/etc. [this would amount to installing a typ grid/cathode electronic overcurrent device..... except it would only work IF the grid/cathode fuse opens]. Origionaly, we installed a 100 K- 2w-MF across the grid fuse.... believing if the grid fuse blew.... the resulting bias developed.. would cutoff the tube... and tube couldn't be driven. In actual practise... with the typ globs of drive... you can still drive the tube... albeit.. in class C ! So we subsequently removed the 100K resistor. Now,,, if the grid fuse opens... zero output... input swr rises from flat.. to 2:1. ..... No DC grid current. Here's the real concern. With the grid fuse blown open.... there is NO return for DC grid current... fine so far. But what happens to the drive RF [800 ++ watts worth] ?? Are we burning up the cathode.... and /or the grid ..or both ???? In all cases these RF decks/ power supplies are remote located 30-
50' away from the station. The station can however... read all 4 bird line sections... and also continuously monitor the PEP refelcted pwr [part of the high swr shut down circuit]. On a related note.... IF either HV fuse blow open..[1st one in one leg of sec of plate xfmr... the 2nd in the B+].... and with tube driven... ALL the electrons flow to the control grid... instead of the anode... and you guessed it... the grid current will skyrocket.. [read it will blow the grid fuse asap]
We have seen cases whereby the HV fuse + grid fuse will blow open... cuz of an antenna problem... and high swr [trips on high reflected pep power] circuit has NOT tripped the amp offline. We have also seen cases, where the high reflected power circuit trips 1st.... and trips both linears offline.. [which also applies -10 vdc to the ALC of the xcvr. ]
All depends sometimes... whether it's an arcing ant/badly installed connector. [resulting in an INSTANTANEOUS wide open/dead short... infintite swr] OR just a high swr... between 2:1 and 5:1
In some cases the B+ sandfilled fuse blows..... some times BOTH sandfilled fuses blow. For some added protection... we are thinking of an adjustable spark gap between load cap ...and chassis..... set to fire at a V higher than the normal low swr V.... and a point well below the rating of the 5 kv vac load cap. I think Rich may have done this ? I don't want the expensive ceramic vac load cap to ever internally arc.

Spark gaps are good!

Sri for the drivel....... does anybody actually read any of my posts....... or is this all old news ???
Later... Jim VE7RF

I enjoy reading your posts Jim!

Best thing I have seen is to use one of the triode control boards and use the protection circuitry they provide. You will find there's hardly any failure that will can not be detected and reacted to almost instantly. You might have to make a few adaptations to meet the need for QRO++ but the principle is the same. This is especially good because it can trip the amp off line so you don't have high drive going into a malfunctioning amp.

73, Tony W4ZT

--- In ham_amplifiers@..., R L Measures <r@...> wrote:

RICH SEZ...Now the bad news: Without W8JI, the "recognized

amplifier expert", Charles Thomas Rauch, Jr., and his statements
such as the laws of AC circuit analysis do not apply to VHF
suppressors; and VHF resonant circuits can't ring, but HF ones can,
things will definitely be less entertaining.


### this is one soap opera I don't wanna ever hear again!

### The best way to solve parasitic problems in a TL-922 SB-
220/221 is to REMOVE all 6 x 200 pf bylass caps + 2 x RF
chokes from sockets... and toss em !! Strap all 6 x grid
pins to chassis ..with wide strap..... end of problem.

### The origional RL Drake L4 [1963] had the 6 x caps + 2 x
chokes crap in it.... Heath just copied it in 1969.

### Those caps... + the stray C from cathode to grid was
supposed to make a "V divider" and provide "NFB" [it
didn't... it made the IMD WORSE]

### I mentioned the "grnding ALL 6 x grid pins to the chassis"
[what a novel idea for a GG amp !] to some friends who had SB-
220/221's.... TL-922's, Henry's etc.

### They all did just that. The fellow with the SB-221
reported his severe TVI into his neighbour's satellite dish
setup.... VANISHED. He also reported his power output on 10m
shot WAY up. [it had lousy eff on 10m].

### The fellow with the TL-922 reported that he could now remove
the after market nichrome suppressor's... and insert the stock
Kenwood suppressor's back in... and it remained ROCK STABLE ever
since!

### They ALL reported that their drive requirements DROPPED
between 20 and 25 watts. [that includes my 3 x L4B's... + my
buddy's Henry 3-500Z amps]

### Dropping the drive power from a Kenwood 870 maakes a BIG
difference. The IMD on a Kenwood 870 DROPS like a rock.. when
you lower the output from 100w to 85w.... and it's way down
at 50 w. [the ARRL lab did IMD tests on a 870 on 20m.. at 3 x
diff pwr levels... 100-85-50w... in their "extended review" ]

### For bigger tubes... GLOBARS wound with wide flat strap are
the real answer....rock stable... and the SP Type globars will
handle 350 deg C all day long !

### Even on smaller tubes.... you can use 10-20-25 w globars...
and flat strap/wire. The trick is to just use enough turns to
suppress the parasitic... and no more......... and no, the globars
won't burn up on 10m.

### Of course... to do this right... add some HV fuses,
cathode fuse, grid fuse... and a really good glitch R. I use a
50 ohm - 50 W wirewound right inside the L4B's.... mounted
directly to the Millen Red HV connector... on the inside...
mounted vertically.

Later...... Jim VE7RF