开云体育

Hee hee.

Thanks Bob, nice to start the day with a nice laugh.

73 Jim SM2EKM
--------------

Robert B. Bonner wrote:

On Nov 7, 2006, at 1:34 PM, Peter Voelpel wrote:

Peter,

It might be a good idea to contact SM2CEW



If it costs already a lot of time and money I would invest 100 more Euros
(for new Parts)
in a well regulated screen supply.
Otherwise I am shure you are asking for trouble with an output >4KW.
The data for the screen in HF class B is 800V at 100-120mA, where will that
derive from?

73
Peter

________________________________

From: ham_amplifiers@... [mailto:ham_amplifiers@...]
On Behalf Of Peter
My PSU being about ready, I await comments before whether deciding
going on with my G2DAF design or changing to stabilized g2 / g1
suplies ( the G2DAF option being the more attractive to me as far as
construction is concerned ).

In my experiences, the G2DAF design is c. 11db cleaner with a 2-tone modulation test than it is with voice modulation. With voice modulation, total IMD is c. 21db below pep.

My project costs me much time, efforts and money and I therefore need
all available information before deciding what to do.





Yahoo! Groups Links

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

Yeah who needs fuses?

Current limit with the power company.

Hello? Yes this is NUKE STATION #6, yes this is SM2EKM I'm coming on line,
pull the rods out.

--- In ham_amplifiers@..., "Robert B. Bonner"
<rbonner@...> wrote:

--- In ham_amplifiers@..., "Robert B. Bonner"
<rbonner@> wrote:

Jim,

## A 500 ohm 10-25 watt resistor across it is a throwback to
Orr's books.["low Z across cathode fuse"] We already tried that

on

the 3x6 amp.. and other amps. We removed the 800 ohm 10 watt

R...

and replaced it with a 100 K 3 watt MOF. IF Cathode fuse
blows... the V-drop across the 100 k resistor will bias the tubes
to cut off ASAP. Dissipation across the 100 K R is virtually zero
watts.

***You would be surprised how little it takes to bias off an amp.

Currently

on one project I have a 3000 ohm 10 watt resistor biasing the tube

during

standby. If you lose a fuse do to some reason other than a HV arc

the 500

Ohm resistor does a sort of controlled slow shut down. 100,000 is

way up

there, to high in my opinion. You already used a 1000 ohm, and

mine is

playing with a 3000 ohm, and Hank uses 10K to bias the tube in

just about

everything.

### Anything from 500 ohms to 2 megohms works just fine. Rich's
electronic bias scheme used 1 meg. Depends on what tube u are
trying to cut off... and plate V. A 3x3 needs 45v to get it down
to just 1 ma idle current. [zsac] A lot of eimac specs will say
XXX V will produce 1 ma of zsac. IMO... somebody found a 25 k-
25 watt in their junk box... used it in a linear for cut off
bias.... on RX.... then it got copied into the next 25 hand books.

### IMO a 100 k or any value of MOF will be more reliable than a
wire wound.

However I have a 5 OHM 100 watt in the B+

You should also have a 100-200 Ohm 25 watt B- lead to ground in
the power supply and the 1 ohm grid metering resistor in the head.

### The grid meter in my case is a separate meter 0-1000Ma... shunt
is built into it... abt .82 ohm I think. Can't stand
multimeter's. I want to be able to see all the metering all the
time.. all at once... including fil V and current, plate V, grid
I.. plate I.. plate xfmr pri v and I... etc.

The 100-200 ohm resistor is for WHEN the fuse blows, anything over

10X the meter resistor doesn't effect the meter calibration. A 100
Ohm keeps the B- close to ground if the fuse blows. That's what its
for.

### That 100-200 ohm resistor from B- to chassis just keeps B-
from straying to far. IF ur cathode fuse BLOWS in the CT of a
thoriated tungsten fil tube like 3-500Z or a 3x3.... the B- is
STILL at chassis potential... via the grid shunt !!! A better
method is to just use RVS connected diodes between chassis and B-
.... then B- can never float more than +/- .7 V Dunno abt oxide
tubes like a 8877... with a separate cathode.

### IF u got a B+ to chassis short.... and u didn't have diodes
across ur meter's... and only a 200 ohm r between chassis and B-
...... u would smoke both meter's.... IF the meter's blew wide
open.... the only other alternate path is through ur 200 ohm
resistor from chassis to B-. IF the meter's were protected,
the B+ would come out of the chassis.. thru both meter'... back
to B-.... shorting out the B+ supply.

*** That way things don't "run wild" if the fuse opens. Yes

without a resistor or if that resistor opened up there could be a
safety issue.

### agreed. IMO the MOF resistor would be less likely to open up
than a wire wound. For redundacy... a PAIR of 100 k 2-3 w mof's
across the fuse... or a pair of wirewounds would be the ultimate
answer... problem solved.

There are many ways to design and or build amplifiers all are just

fine and values can be variable within reason.

### Partially agreed. some designs are flawed... some are just
outright dangereous. Some older ARRL designs put the HV meter
between B+ and chassis... instead of B+ and B- ... no R between
B- and chassis.. and no meter protection. IF grid shunt ever
opened up... or no connection via multi meter switch... HV meter
would drop to zero.... when in fact.. HV is still present... and
lethal.

I use 1 ohm grid shunt. A 100 Ohm in the PS B- to chassis,

Usually 10K to bias the tube during standby and have used as little
as 3000. A pair of 20 OHM step start resistors and 100K bleeder
across the power supply.

### You only need step start resistance in one leg of the 240 v
line... and only one spst relay/ contactor to short it out. Only
thing of interest in a HV step start is loop resistance in the
primary. Having said that... I tried 50 ohms... and also 25
ohms. 50 ohms will limit 240 v inrush to 4.8 A 25 ohms will
limit 240 v inrush to 9.6 A I found with 50 ohms... that plate
V would not rise as high as 25 ohms.... and when shunted, the
secondary surge was WAY higher. That was with 6800Vdc supply...+
100 uf filter cap.... and 14 second delay. A 0-50 A ameter in one
leg of the 240 V primary will give one a real eye opener...esp that
sec surge.

### 25 ohms works great.. and any relay/contactor will handle
9.6A. Plate V starts at zero.... just sails right uo.... then
slowly keeps climbing... then it jumps that last little bit. Energy
in joules goes to the square of the voltage. A cap that only
has 1/2 V on it.... is only 1/4 charged up..... u want the V up as
high as u can get... b4 shunting... and a 25 ohm resitor works
better than a 50 ohm. [for a 6.8 kv + 100 uf filter]

I have always placed 500 across the cathode fuse, however the bias

size would also work. The 500 will drop amp gain to nothing almost
instantly and you've still got power applied at that point. That RF
needs to go somewhere.

I haven't experimented with my expensive tubes by removing the

cathode fuse and applying power... I've only had 1 cathode fuse
POP in 35 years, and that was during power up for some reason.

### Try it... tube shuts down.... works great.

I have never used a 100K for a standby resistor. I would

personally prefer to just cut off the tube rather than bury it in
cutoff.

### Anything from 600 ohms to 2 megs works fine for the RX cut
off... and no, u are not burying it into cutoff. IF it actually
was really toatally 100% cut off.... and no current flow... u
would also have NO v drop... and NO cut off bias developed. Fact
is...it reaches equilibrium.... it's never quite cut off..... just
enough current flows to almost cut it off ... but not quite. EG: to
get 45 v of cut off bias.... you will have .45ma of current flowing
through the 100 K resistor. [diss =.02 W]

#### Haven't tried anything above 2 meg.... I'm sure even 10 meg
would work.... u just don't EVER want a wide open.... so a PAIR
of resistors... either MOF OR wire wounds would be better... then
u have redundancy.

later... Jim VE7RF

I have two linear power amplifier units for sale:
4XSD1427(THX15-C) in parallel push-pull in each unit, with driver stage and large heat sink.RF output power( each unit):400W, freqence range:1.8-30MHz, it was used in a 400W solid state SSB transmitter, not for RF power generator.I have test all transitors,they are all OK.You can add a combiner to build a 800 or 1kW linear amplifier.
I ask $400 for two unit, include shipping cost( surface mail).
73! Hsu

Peter,

It might be a good idea to contact SM2CEW



If it costs already a lot of time and money I would invest 100 more Euros
(for new Parts)
in a well regulated screen supply.
Otherwise I am shure you are asking for trouble with an output >4KW.
The data for the screen in HF class B is 800V at 100-120mA, where will that
derive from?

73
Peter

________________________________

From: ham_amplifiers@... [mailto:ham_amplifiers@...]
On Behalf Of Peter
My PSU being about ready, I await comments before whether deciding
going on with my G2DAF design or changing to stabilized g2 / g1
suplies ( the G2DAF option being the more attractive to me as far as
construction is concerned ).

My project costs me much time, efforts and money and I therefore need
all available information before deciding what to do.

Good afternoon List !

I'm new to the list.
Today I got a tip G2DAF experts might be around here.

From various sides I got incompatible advices on the G2DAF concept
for a HF linear ( SSB speech ) PA with a Philips QBL 5/3500 ( @ Va 6
Kv and @ Ia 1.1 A , very heavily cooled... ).

It seems the contrary opinions are all about the IM3 distortion ( and
to a lesser extent the efficiency ).

I'd be grateful for unbiased reviews or ( even better ) real life
experiences on this subject.

Maybe even one of you did build yourself a QBL 5/3500 HF linear PA on
basis of the G2DAF concept ?

My PSU being about ready, I await comments before whether deciding
going on with my G2DAF design or changing to stabilized g2 / g1
suplies ( the G2DAF option being the more attractive to me as far as
construction is concerned ).

My project costs me much time, efforts and money and I therefore need
all available information before deciding what to do.

Thanks for input.
Peter, PE1E.

Bob,

I think I ditched the idea with a fuse. What the heck,
if my 8171 box doesnt have a fuse why bother with this
toybox.

73 SM2EKM
-------------
Robert B. Bonner wrote:

I see. Well its getting strange. As I wrote before,
in the manual I have someone had written 15 ohms on
the schematic, there is no parts list in the manual.

However I have two L-4Bs here and I see now that
both has 150 ohm resistors.

Oh well, its a playtoy amplifier anyway.

73 Jim SM2EKM
-------------------


Robert B. Bonner wrote:

The L4B manual out of my file shows 3 - 15 OHM 2 Watt resistors as the
Parasitic Supressor

BOB DD

--- In ham_amplifiers@..., "g0vdz" <g0vdz@...> wrote:

Dear All, I have a Yaesu FL 2100z. I have had it for about 6 months
and all though I do not use it very much it is nice for contests.
I went onto cqww and decided to enter a topband entry. When i

fired

up the liner, it would only put out a marginal signal. I.E. 25

watts

in 40 watts out.
I have lifted the lid and the RFC or anode choke looks a bit the
worst for wear. Could this be the cause of the problem. The other
bands seem reansonable. Has anyone had this problem and if so what

was

the problem
Many thanks to all Nigel G0VDZ

Nigel,

I also have a FL2100z, just got it going again after a mishap in
shipping some years back(wonderful posties, AMP a tad heavy I guess)
Anyway what sort of input Ia was the beast drawing on 160 M ? Was it
getting enuff drive, u say 25 Watts in, maybe the input Pi filter is
US or the switch contacts. If it was drawing the normal (or near
normal Ia , i.e. 300mA ish) then got to be in the Plate CCTS
somewhere. I did try to fire mine up on 160M and got some orrible
frying noises from inside. Found out that one of the sleeved heavy
wires from the plate side to the padded C's was arcing thru the
insulation to the body of the band change switch, were it was bent
around the switch body, pulled it away from that and no Frying noises
anymore. Just need a new pair of 572b's as the previous owner cooked
em ever so slightly on a qrg just a tad LF of 10 Meters Ha Ha....

Hope that the above gives your some ideas, perhaps you'll can run the
beast in a dummy load and try some more tests.

73 Derek G8ECI

On Nov 6, 2006, at 5:40 PM, pentalab wrote:

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

On Nov 6, 2006, at 7:03 AM, Jan Erik Holm wrote:

I?m working on fixing up a L-4B, it has fried resistors
in the supressors. They use 3 x 15 ohms but I have none
in my junkbox however I have plenty of 10 ohm R?s, will
that do? 3 x 10 ohm R?s?
It doesn?t seem to matter too much, I notice that the L-7
use 2 x 47 ohm R?s and the Henry 3K 2 x 150 ohm R?s, it
seems to be all over the place.

RICH SEZ.... Hello, Jim. == In a VHF parasitic suppressor, when

one sees obliteration, it usually means VHF oscillation.

### No it doesn't.

Jim == The suppressor resistors are paralleled with c. 80nH of L.
What frequency would be required to develop enough V-drop across 80nH
to fry the resistors?

With his 3 x paralled 15 ohm resistor's... it
means he has 5 ohms in total.... and the wrong value to start with !


The goals of a VHF

parasitic suppressor are to build one with a low VHF-Q that won't
melt-down at 29MHz. Typical factory-stock suppressors have a Q of

c.

5 @ 100MHz. However, VHF suppressors can be built with a Q of

1.5.

This Q-decrease decreases VHF gain with the tradeoff of about 2%
power decrease at 29MHz.

### which means the nichrome has to diss 26 watts.

Please show your calculations.

The STOCK L4B
suppressor's are bullet proof....

So why did the resistors fry to a crisp?

and even more bullet proof, after
u grnd the grids directly to chassis.... like several hundred other
owner's have done. Rich... how many suppressor kits have you
actually sold to L4B owners ??? This is truthfully the very 1st
time in 30 years I have ever heard of anyone frying the suppressor
in a L4B.... or a L4 for that matter. I don't even think it can be
done. Now with just a 5 ohm suppressor... that's a diff story.

I?m following VE7RF?s advice to ground grid pins with
wide straps.

RICH SEZ.. please check grid resonance with a dipmeter before and

after and get back to us with the results.

### Yes, do that.... it might change all of 1 mhz... so what ?

A grounded grid is not grounded at the grid resonance. IOW, the input
is no longer shielded from the output by the grid at this frequency.
I see it as no coincidence that 3-500Zs typically oscillate near
their grid resonance.

At
least it's now more stable...and 25 watts less drive required...

At 28MHz, how does 600pF of grid bypass C dissipate 25w?

and
BETTER IMD..... since the original "NFB" was flawed. The IMD
improves even more... since the xcvr is much cleaner with 20-25
watts less output.

Not always.

- note - at and above a grounded grid's resonant frequency, the

grid no longer shields the cathode/input from the anode/output.

### That's nice... and totally irrelevant in this case. When the 6
x grid pins are bonded to chassis with short, wide straps.. the
shielding effectiveness of the grid is at absolute MAX.

Even where the grid is a parallel resonant circuit? Do parallel
resonant circuits act like an open or a short, Jim?

Suppose I have to put in a fuse in the B-
return to be safe. Also I?m swapping out the .82 R in
the PS for a better glitch R,

good idea. With a sturdy 10?- 20? glitch-R, why use a HV fuse?

### Uh... so u don't have to rely on opening off the ultra SLOW
breakers in the L4PS.... of which mine have never opened in 30
yrs... since the stock .82 ohm 1 w R always beats it to the
punch. You remove the HV fuse....

Real fuses are designed to extinguish the metal vapor arc.

then install a glitch R.....
the glitch R and 25 uf filter cap in the HV supply will form an
RC network.... dragging out the DURATION of the fault current.

Limiting the peak magnitude of the fault current seems to be what
avoids damaging a tube.

All
u end up doing is stressing the stock oem diodes in the HV
supply. That surge rating for diodes is a 1/2 cycle rating... 8.3
Msecs Takes a helluva lot longer than that to open up a slo blo
primary breaker in the 240 line.

Jim == The turns ratio of the HV transformer is c. 1 to 5. The pri. R is likely c. 1-ohm and the sec. R is c. 13-ohms. How many sec.
amperes would flow in such a transformer with the sec. shorted?

also I put diodes across
the I meter.

good move, Jim

### Install RVS connected 1N5408's or better yet. 6A10's..
across BOTH meter's.

Anything else I should think about?

RICH SEZ... lower Q suppressors,

##### Mother Mary... save us all.. from Parasitic hell !!

If the Blessed perpetual Virgin can demonstrably reduce the Q of a
VHF suppressor on a HP Z-meter, I'm going back to church.

step-start, high speed switching. 8160
final amplifier?

### 8160 [3x 10] requires a 100 A fil xfmr... which means the fil
xfmr HAS to be mounted AND insulated from the chassis....

Correctomundo.

and a
bifilar used in the fil xfmr 240 v PRIMARY. Since the fil xfmr is
now in the UNDERSIDE of the RF deck [it's entirely hot with drive
RF]...

it just added a TON of un needed extra weight.

Porcine shampoo, I just picked up a Dahl 7.5v/100a xfmr with one hand
and my guess is c. 17-lbs

.. enough such
that it tips the scales.. and one guy has extreme difficulty getting
the RF deck outa the rack.

One needs no rack when one has PEM nuts.

### with a 3x 6.... fil xfmr can be 6' away... in lower portion of
Rack.. and cabled with 2 ga flexible wire.. to RF deck. A 80 A
bifilar is a snap.... using paralled 8 ga magnet wire. You can see
pix of that assy on the photo page.

Later... Jim VE7RF

73 Jim SM2EKM




Yahoo! Groups Links

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

Yahoo! Groups Links

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

On Nov 6, 2006, at 4:36 PM, pentalab wrote:

--- In ham_amplifiers@..., Jan Erik Holm <sm2ekm@...>
wrote:

I?m working on fixing up a L-4B, it has fried resistors
in the supressors. They use 3 x 15 ohms but I have none
in my junkbox however I have plenty of 10 ohm R?s, will
that do? 3 x 10 ohm R?s?
It doesn?t seem to matter too much, I notice that the L-7
use 2 x 47 ohm R?s

### Whoa.. stop right there. I Don't have the L4B manual handy...
but I'm positive they use 2 x 47 ohm 2 Watt carbon resistors.. and
silver plated strap..... should be same or similar to the L7. I'll
check all 4 x of my L4B's later tonight.

A timely question: Why did the resistors fry?

and the Henry 3K 2 x 150 ohm R?s, it

seems to be all over the place.

I?m following VE7RF?s advice to ground grid pins with
wide straps.

#### Originally when I tried this stunt.. I used just one strap ,
from just ONE grid pin PER socket.. [2 x straps in total]. Worked
good... except 10m input swr went nuts. ...



Suppose I have to put in a fuse in the B-

return to be safe.

### IF u want a high speed grid fuse.... install a 3agc bayonet
type fuse holder on rear panel. Wire between grid shunt and
chassis. [stock, one end of grid shunt goes to chassis... just open
this point up] Use a 250-350 ma fast fuse.



Also I?m swapping out the .82 R in

the PS for a better glitch R,

#### Whoa. The .82 R 1 watt resistor in the B+ ... IS the HV
fuse ..

It's not a HV fuse because a metal vapor arc forms between the
resistor's two copper wires as the fuse disintegrates. Current stops
after the arc extinguishes - and that only happens after the filter
caps have mostly discharged.
Real HV fuses are lengthy and they are filled with silica sand to
extinguish the metal vapor arc. However, since a HV fuse does not
limit initial peak current, one needs a sturdy glitch-R.

NOT a glitch R. LEAVE the .82 ohm IN !

That's an order!

### Here's where u install a glitch R in a L4B. ...

...

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

I must put this right. 15 ohms was wrong, sorry!

Explanation:

Since I have no manual I have found one on internet,
on the schematic someone has scribbled with a pen
3 x 15 ohms for the resistors, thats why I did say
15 ohms.

Last night I took them apart and resistors are 150
ohms and no nothing else.

Also the resistors arent fried, when inside the coil
they just looked fried to me however its just the
paint that got fried. I did measure them and they
all measure around 180 ohms, a tad high maybe however
at 20% tolerance they are ok.

What Jim writes below about L-4B suppressors is still correct.

I think I just put the crap back again.

Im sorry about the error.

73 Jim SM2EKM
-------------------------------------------------------------------

pentalab wrote:

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

On Nov 6, 2006, at 7:03 AM, Jan Erik Holm wrote:

I?m working on fixing up a L-4B, it has fried resistors
in the supressors. They use 3 x 15 ohms but I have none
in my junkbox however I have plenty of 10 ohm R?s, will
that do? 3 x 10 ohm R?s?
It doesn?t seem to matter too much, I notice that the L-7
use 2 x 47 ohm R?s and the Henry 3K 2 x 150 ohm R?s, it
seems to be all over the place.

RICH SEZ.... Hello, Jim. == In a VHF parasitic suppressor, when

one sees obliteration, it usually means VHF oscillation.

### No it doesn't. With his 3 x paralled 15 ohm resistor's... it
means he has 5 ohms in total.... and the wrong value to start with !


The goals of a VHF

parasitic suppressor are to build one with a low VHF-Q that won't
melt-down at 29MHz. Typical factory-stock suppressors have a Q of

c.

5 @ 100MHz. However, VHF suppressors can be built with a Q of

1.5.

This Q-decrease decreases VHF gain with the tradeoff of about 2%
power decrease at 29MHz.

### which means the nichrome has to diss 26 watts. The STOCK L4B
suppressor's are bullet proof.... and even more bullet proof, after
u grnd the grids directly to chassis.... like several hundred other
owner's have done. Rich... how many suppressor kits have you
actually sold to L4B owners ??? This is truthfully the very 1st
time in 30 years I have ever heard of anyone frying the suppressor
in a L4B.... or a L4 for that matter. I don't even think it can be
done. Now with just a 5 ohm suppressor... that's a diff story.

I?m following VE7RF?s advice to ground grid pins with
wide straps.

RICH SEZ.. please check grid resonance with a dipmeter before and

after and get back to us with the results.

### Yes, do that.... it might change all of 1 mhz... so what ? At
least it's now more stable...and 25 watts less drive required... and
BETTER IMD..... since the original "NFB" was flawed. The IMD
improves even more... since the xcvr is much cleaner with 20-25
watts less output.

- note - at and above a grounded grid's resonant frequency, the

grid no longer shields the cathode/input from the anode/output.

### That's nice... and totally irrelevant in this case. When the 6
x grid pins are bonded to chassis with short, wide straps.. the
shielding effectiveness of the grid is at absolute MAX.

Suppose I have to put in a fuse in the B-
return to be safe. Also I?m swapping out the .82 R in
the PS for a better glitch R,

good idea. With a sturdy 10?- 20? glitch-R, why use a HV fuse?

### Uh... so u don't have to rely on opening off the ultra SLOW
breakers in the L4PS.... of which mine have never opened in 30
yrs... since the stock .82 ohm 1 w R always beats it to the
punch. You remove the HV fuse.... then install a glitch R.....
the glitch R and 25 uf filter cap in the HV supply will form an
RC network.... dragging out the DURATION of the fault current. All
u end up doing is stressing the stock oem diodes in the HV
supply. That surge rating for diodes is a 1/2 cycle rating... 8.3
Msecs Takes a helluva lot longer than that to open up a slo blo
primary breaker in the 240 line.

also I put diodes across
the I meter.

good move, Jim

### Install RVS connected 1N5408's or better yet. 6A10's..
across BOTH meter's.

Anything else I should think about?

RICH SEZ... lower Q suppressors,

##### Mother Mary... save us all.. from Parasitic hell !!




step-start, high speed switching. 8160
final amplifier?

### 8160 [3x 10] requires a 100 A fil xfmr... which means the fil
xfmr HAS to be mounted AND insulated from the chassis.... and a
bifilar used in the fil xfmr 240 v PRIMARY. Since the fil xfmr is
now in the UNDERSIDE of the RF deck [it's entirely hot with drive
RF]... it just added a TON of un needed extra weight... enough such
that it tips the scales.. and one guy has extreme difficulty getting
the RF deck outa the rack.

### with a 3x 6.... fil xfmr can be 6' away... in lower portion of
Rack.. and cabled with 2 ga flexible wire.. to RF deck. A 80 A
bifilar is a snap.... using paralled 8 ga magnet wire. You can see
pix of that assy on the photo page.

Later... Jim VE7RF

73 Jim SM2EKM




Yahoo! Groups Links

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

--- In ham_amplifiers@..., "Robert B. Bonner"
<rbonner@...> wrote:

Jim,

If you think you need a B- fuse, put a 500 ohm 10-25 watt resistor

across the fuse incase it opens.

### IF u install a fuse in the B-.... it's technically a "cathode
fuse"... and a 3agc fuse holder will suffice. Use a 750 ma fast
blow fuse.

## A 500 ohm 10-25 watt resistor across it is a throwback to
Orr's books.["low Z across cathode fuse"] We already tried that on
the 3x6 amp.. and other amps. We removed the 800 ohm 10 watt R...
and replaced it with a 100 K 3 watt MOF. IF Cathode fuse
blows... the V-drop across the 100 k resistor will bias the tubes
to cut off ASAP. Dissipation across the 100 K R is virtually zero
watts.



You should also have a 100-200 Ohm 25 watt B- lead to ground in
the power supply and the 1 ohm grid metering resistor in the head.

### What u just said doesn't make sense at all. Are u saying to
install the 100-200 R between B- and chassis ??? or in series
with B- ??? Both are bad idea's and more throwback's to Orr's
books. IF u install a 100-200 ohm R between B- and chassis....
and say u get a B+ to chassis arc.... fault current could travel
up the 100-200 ohm R..... which makes it now a way too high in
value...."glitch R" ... but only if you fried the 1 ohm grid
shunt..1st.

### If it's just to... "hold the B- close to chassis potential...
and not screw up the metering"... well , the B- is ALREADY at
chassis potentail... via the grid shunt. A better way to do all
this is to just use RVS connected 1N5408's BETWEEN chassis and B-

### This will ensure the B- can never wander more than +/- .7 V
Also, THEN... if u ever get a B+ to chassis arc... fault current
is up one side of the RVS connected diodes... then DIRECTLY to B-
completeing the loop... shorting the HV supply... and the 50 ohm
glitch will limit the Fault current to 53 A.. while the .82 ohm HV
fuse blows asap. [I have NEVER tripped the built in breakers in the
L4PS HV supply. ]





That way things don't "run wild" if the fuse opens.

### Do it with a 100 K across the cathode fuse... and nothing
happens.... just auto shuts down. BUT the CATHODE fuse HAS to
have A resistor across it at all times.... you don't want this
part of the circuit to ever open up... otherwise the cathode will
try and assume full plate V... and stuff like .01 uf bypass caps
between chassis and cold end of fil choke will start exploding.
make sure all the wiring from B-... all the way through to the CT of
the fil xfmr is rock solid.

My large projects have 5 ohm 100 Watt glitch resistors in the PS.

#### 5 ohms is not big enough value to limit current. 4 kv /5 ohms
= 800 A of fault current. 2 A of normal plate current [say a 2 x
8877 amp] thru a 5 ohm glitch will diss just 20 watts.

What's a parasitic suppressor? OH no...

### The stock suppressor in the L4B is fine as is... runs fine
with RTTY /FM/A0/AM on 10M too.

Later... Jim VE7RF

BOB DD

On Nov 6, 2006, at 3:02 PM, ad4hk2004 wrote:

no square in the equation...

C = k * ( area/distance)

As I recall, this is not the equation.

k is the factor for insulator dielectric factor
() is the ratio of the plate area divided by the distance between plates

denny

--- In ham_amplifiers@..., jmltinc@... wrote:

So, is the equation wrong? Jim's answer implies an inverse square

equation.

-J

Yahoo! Groups Links

R L Measures, AG6K, 805.386.3734
r@..., , rlm@..., www.somis.org

--- In ham_amplifiers@..., Jan Erik Holm <sm2ekm@...>
wrote:

I?m working on fixing up a L-4B, it has fried resistors
in the supressors. They use 3 x 15 ohms but I have none
in my junkbox however I have plenty of 10 ohm R?s, will
that do? 3 x 10 ohm R?s?
It doesn?t seem to matter too much, I notice that the L-7
use 2 x 47 ohm R?s

### Whoa.. stop right there. I Don't have the L4B manual handy...
but I'm positive they use 2 x 47 ohm 2 Watt carbon resistors.. and
silver plated strap..... should be same or similar to the L7. I'll
check all 4 x of my L4B's later tonight.



and the Henry 3K 2 x 150 ohm R?s, it

seems to be all over the place.

I?m following VE7RF?s advice to ground grid pins with
wide straps.

#### Originally when I tried this stunt.. I used just one strap ,
from just ONE grid pin PER socket.. [2 x straps in total]. Worked
good... except 10m input swr went nuts. The original idea.. was I
left in the 3 x 200pf caps + choke per socket... so if experiment
bombed... no big deal.. just remove the 2 x straps.

### The 10m high input swr fix was easy.... just REMOVE all 6 x
200pf caps and 2 x chokes... and replace with 3 X wide straps PER
socket...... and keep the coax from xcvr to L4B short. Check and
retweak input swr on all bands if needed. U should be able to get
dead flat input swr. I tweak with top lid off.. on low plate V
cw mode... 600 w into a dummy load... on each band.




Suppose I have to put in a fuse in the B-

return to be safe.

### IF u want a high speed grid fuse.... install a 3agc bayonet
type fuse holder on rear panel. Wire between grid shunt and
chassis. [stock, one end of grid shunt goes to chassis... just open
this point up] Use a 250-350 ma fast fuse.



Also I?m swapping out the .82 R in

the PS for a better glitch R,

#### Whoa. The .82 R 1 watt resistor in the B+ ... IS the HV
fuse .. NOT a glitch R. LEAVE the .82 ohm IN !

### Here's where u install a glitch R in a L4B. Remove the
entire HV interlock assy. It straddles both the top of the
chassis.. and below the chassis. Notice that when u activate the
interlock in slow motion... that the B- lead attached to the HV
interlock below the chassis... grnds out to the chassis.... a split
second before the B+ [attached to top side of chassis
interlock] grnds out to chassis. This was done so B+ will short
[via chassis] DIRECTLY to the B-... blowing the .82 ohm HV fuse.

### IF the HV interlock was just wired to short out the B+ to
chassis.... the path would be through the grid shunt through the
grid meter [if multimeter switched to read Grid I].. then through
the plate current meter... then back to B-... frying both meters and
grid shunt.

### Install REVERSE connected 1N5408's directly between
chassis... and B- Then u will NEVER have the above problem IF
HV shorts to chassis for any reason.

### With the HV interlock removed... we now have room to easily
install the glitch R. Use a 50 ohm 50 watt wire wound
[dale/clarostat/IRC/Ohmite, etc... they all have identical physical
specs]... and mount one end directly to the Millen HV connector.
This glitch R will stand vertical. Top end of resistor will
easily self support itself.. with only the bottom end connected. I
used a internal tooth lockwasher/split ring lockwasher +
flatwasher.. on the Red Millen HV connector.

### Run new 5 kv Belden/similar test prod wire from base of
existing plate choke to TOP of new 50 ohm 50 w glitch R.

### Now... it's bullet proof.



also I put diodes across

the I meter.

### Put REVERSE connected 1N5408's across BOTH meters. [4 in
total.]

Anything else I should think about?

### Yes.... use a separate outboard box.. like a standard
electrical box 8" x 8" x 3-4".... and make a step start delay
for the 240 V line. I used a DPST 30 A P+B relay... to turn on
the 240 V. A 2nd relay..... SPST 30 A P+B relay shorts out
the 25 ohm ...50-100 watt step start resistor. Step start
resistor is simply installed in ONE leg [either leg] of the 240 v
line ..AFTER the 1st relay. Install MOV's across all relay
contacts. My 25 ohm 100 w ww is on standoffs. A better R is to
use a METAL finned dale, etc, R... and bolt directly to sidewall of
box.

### I hardwired the incoming [fused or breaker] 240 V line directly
to the 1st relay [dpst]. Also the L4PS is hardwired to output
side of the step start external box.

### I ran a control cable up to small box on the desk.. with TWO
toggle switches in it. [ 1 x mov across each toggle] Throw
switch 1... and relay 1 operates........ wait 2-4 seconds... throw
switch 2... 2nd relay activates. I used a 1A.. 3gac fuse holder on
the step start box... to feed the control cable. You can get
fancy and use a 2-5 second timer if u want. The built in rocker
switches on the L4B RF deck are flaky at best... I just leave em
both on all the time... and use my control box. [tiny mini box]to
turn the amp on/off.

### This step start will step start BOTH the fils AND the HV
supply.. all in one shot.

### Remove the connection BETWEEN the 2 x 50k-50 watt wire
wound's in the HV supply. Those two bleeder's dissipate a total of
75 watts... or 37.5 watts each. .... which is why the HV supply runs
so hot. The eq resistor's across each of the 8 x 200 uf lytics
ate already 100 K 2 watt units..... and all 8 of em are mounted
several inches away from the caps. I replaced all of mine with
Rich's 100 k 3 watt MOF resistors. Funny thing was... after 25
yrs... the original 100 K carbon resistors were still 100 k ohms.

### Now, you would think the no load HV would junp up with out
the 75 watt load from the series 50 K bleeders.... but it doesn't
budge an inch !! Now here's part 2. In series with the old
pair of 50k bleeders was a 7 k 5 watt.. BELOW the chassis.. In the
outboard HV supply. The v drop across this 7 k is +130v... used to
bias the tubes off on RX. .... dumb idea... since the +130v will
eventually burn the center contacts off the 3PDT T/R relay.
Also... when in mid air... the cathode will try and rise to full
plate V. Now, since u disconnected the 2 x WW HV bleeders.. u
no longer have +130 available. Fix is... wire a 100 K 2 watt MOF
in the CT of the Fil xfmr [yellow wire, if I remember correctly]...
and rewire center contacts of 3PDT T/R relay, so u just short
out this 100 k on TX. Bullet proof. ... and now zero heat from
the outboard supply. Note... drake used eq resistor's across each
lytic... AND a pair of series 50 k ww across the B+ and B-

Want more? Replace the 8 x lytics with ones that are 2-3
times the UF. Install ONE RVS connected 1N5408 across each
lytic. As is... when sucking 800 ma of plate current on key
down... u can easuliy see the 3+ % ripple on any monitor scope.
I'd replace all the diodes in the doubler with 1N5408's.

## Change the +27 vdc 1/2 wave rectifier for the T/R relay to a
DOUBLER..... results in +63 vdc. I installed an internal 3agc fuse
holder... in one leg of the sec that feeds the doubler.. with I
think.. a 1 A fuse. Install a 2 watt drop resistor [forget the
value.. think it was 900 ohms] in one leg of T/R relay coil. This
won't give u qsk....;but super fast vox... as it's now sped up..
and u can speed up any open frame mech relay.. the same way u can
with vac relays. Install a rvs connected 1N5408 diode across the
T/R relay coil. Don't install any bleeder across the 2 x lytics
in the new +63 vdc relay supply.... u want max OCV.

## If u want fancy.. and qsk... use a RJ1A on output side.

### also... install 10 x 1N5408's in the CT on perf board.... in
series with the 100 K 2-3 watt MOF RX cut off resistor from
above. Install a 1000-5000 uf 50 V lytic across the ends of the
bias diode string. Drake ran it with no bias.. and idle was 230
ma. With 10 x diodes it's now 100 ma on ssb.. and 40 ma on cw. NO
increase in imd on ssb on either voltage. [eimac 3-500Z.. MU=130]

### The 1000D and 1000MP-MKV handle the +63 vdc just fine. I also
key em with a P+B ODC-3 opto isolator.... +13.8 vdc input @ 12
ma.... output side rated for 200 V @ 1 A CCS. Output side of opto
has built in back emf protection... but use a diode across T/R
relay coil anyway.

### I have 4 of these amps. I can't get em into parasitic osc...
doesn't mnatter what I do.. including changing tubes to brand X.
Have u got the stock 1900/2650 V power supply. That tune cap
will arc at 4 kv. Want more power... get a bigger amp.

### grnd the grids directly.. and drive power drops 22 watts....
which gets negated a bit... if u install the 10 x bias diodes.

Later... Jim VE7RF

73 Jim SM2EKM