I have had my L7 amplifier for about a year now, and it was working very well. I recently read and article about grounding the grids on the 3-500 tubes. I had not checked that with my L7. Today I had a chance to check and found that my tubes were wired just as Drake made them. With the RF chokes and the 1kv caps. Well I has some 16 gauge buss wire and decided to do the recommended grounding of the grids. Was not the most easy task, but taking my time about 2 hours or so.? I removed the RF chokes and caps and wired the grids pin 2,3, and 4 to ground with the buss wire. Firing up the L7 very gingerly and driving it with my TR7 at about 60 watts, I get about 600 watts out on CW and about 1200-1300 peak on SSB. So I guess the grounded grid is the way to go as it reduces the drive power and increases output power without the increase in plate current. I will keep this for now. I saved all the parts so it something happens, I can revert but it is working I think better will leave it.? Just some note on my L7, as it is a work horse.

On 3/29/24 17:37, K6OXN Bob wrote:

I have had my L7 amplifier for about a year now, and it was working
very well. I recently read and article about grounding the grids on
the 3-500 tubes. I had not checked that with my L7. Today I had a
chance to check and found that my tubes were wired just as Drake made
them. With the RF chokes and the 1kv caps. Well I has some 16 gauge
buss wire and decided to do the recommended grounding of the grids.
Was not the most easy task, but taking my time about 2 hours or so.? I
removed the RF chokes and caps and wired the grids pin 2,3, and 4 to
ground with the buss wire. Firing up the L7 very gingerly and driving
it with my TR7 at about 60 watts, I get about 600 watts out on CW and
about 1200-1300 peak on SSB. So I guess the grounded grid is the way
to go as it reduces the drive power and increases output power without
the increase in plate current. I will keep this for now. I saved all
the parts so it something happens, I can revert but it is working I
think better will leave it.? Just some note on my L7, as it is a work
horse.

When I grounded the grids on my L-4B; I used wide strips of copper cut
from a sheet of copper.? I don't recall the thickness of the sheet, but
it's one I got at a hobby shop.

73

Stan
KM4HQE

It is a PITA on an L7 due to the tight spaces. I just did a second one. Works nice.

73

-Jim
NU0C

On Fri, 29 Mar 2024 15:37:30 -0700
"K6OXN Bob" <bobk6oxn@...> wrote:

I have had my L7 amplifier for about a year now, and it was working very well. I recently read and article about grounding the grids on the 3-500 tubes. I had not checked that with my L7. Today I had a chance to check and found that my tubes were wired just as Drake made them. With the RF chokes and the 1kv caps. Well I has some 16 gauge buss wire and decided to do the recommended grounding of the grids. Was not the most easy task, but taking my time about 2 hours or so.? I removed the RF chokes and caps and wired the grids pin 2,3, and 4 to ground with the buss wire. Firing up the L7 very gingerly and driving it with my TR7 at about 60 watts, I get about 600 watts out on CW and about 1200-1300 peak on SSB. So I guess the grounded grid is the way to go as it reduces the drive power and increases output power without the increase in plate current. I will keep this for now. I saved all the parts so it something happens, I can revert but it is working I think better will leave it.? Just some note on my L7, as it is a work horse.

There was just some discussion about this, I think in this list. Drake's idea was that the indirect grounding provided some negative feed back which reduced the distortion. That may be desirable even at the price of some reduction of power.

--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998

Yes, I was part of the discussion. I also posted a link to W8JI's engineering analysis that shows why the indirectly grounded grid circuit was a bad idea. The idea did not originate with Drake and I have noted that amplifier manufacturers seem to have abandoned the idea more recently. Some never used it.

73

-Jim
NU0C

On Fri, 29 Mar 2024 20:46:27 -0700
"Richard Knoppow" <1oldlens1@...> wrote:

There was just some discussion about this, I think in this list.
Drake's idea was that the indirect grounding provided some negative feed
back which reduced the distortion. That may be desirable even at the
price of some reduction of power.

On 3/29/2024 6:53 PM, Jim Shorney wrote:

It is a PITA on an L7 due to the tight spaces. I just did a second one. Works nice.

73

-Jim
NU0C

--

73

-Jim
NU0C

On Sat, Mar 30, 2024 at 04:46 AM, Richard Knoppow wrote:

<There was just some discussion about this, I think in this list. Drake's idea was that the indirect grounding provided some negative feed back which reduced the distortion. That may be desirable even at the price of some reduction of power.

toggle quoted message Show quoted text

?

On 3/29/2024 6:53 PM, Jim Shorney wrote:

It is a PITA on an L7 due to the tight spaces. I just did a second one. Works nice.
73
-Jim
NU0C

--
<Richard Knoppow
<Los Angeles
<WB6KBL
<SKCC 19998

Drake, heath, Henry, Kenwood, etc, all did the same thing.? It was a stupid idea, right outa the gate...and doesn't work anyway.? ?IF you want say 10 db of IMD reduction, you require 11 db of NFB...so not quite a 1:1 ratio.?
And? 3 db of NFB means the gain just dropped by 3db...so now you require 200 watts? to drive the amp.... for a lousy 2.8 db of IMD reduction.? With 6 db of NFB,? it will require 400 watts to drive the amp....with a 5.9 db imd reduction.?

Problem is, there is only one way to properly introduce NFB to any GG amplifier....and that's by inserting a low value resistor between the coupling cap and the cathode. Typ values are 10-20 ohms.? Problem with that is, the input Z of the cathode has now increased by 10-20 ohms, so now all your PI tuned input values are way outa whack.? 2nd issue is, the 10-20 ohm resistor has to handle a fair bit of power of power.? Typ, several smaller resistors are used in parallel to make up the 10-20 ohms. 3rd issue is... the gain of the amp dropped by a bunch, and now MORE drive required.

?

On the drake, heath, henry, Kenwood amps, they tried the NFB trick by using grid caps. (The RF choke is to allow DC grid current to flow).? ( grid caps will block DC grid current).? Those grid caps are 1/2 of a V divider.? The 2nd 1/2 of the V divider is made up of the internal grid to cathode C of the tube itself. The values used for each 1/2 of the V divider are out to lunch, and will not work anyway.? And if you do get it to work, it's for one band only.?

What they managed to do with their fubar grid caps is..... the amp is now more UN stable.? Stability requires the best RF grnding of the grids you can get...and that's when each of the 3 x grid pins has it's own copper strap. And never use the screws used to submount those? Johnson style sockets.? I know of several TL-922 owners, that had instability issues and installed the nichrome junk.? Once the grids were bonded directly to the chassis, amp is rock stable...and oem suppressors were re-installed...abd amp still rock stable.? 20-25 watts less drive required.?

?

What drake, et all ended up with is..... just a tiny bit, fractional db imd improvement...and only on one band at most.? ?And the fractional NFB results in lower gain....and typ 20-25 watts? MORE drive required.?

When ARRL lab tested the Kenwood 850 yrs ago,? ?for whatever reason, they tested it on 20m, at 3 x different power levels, 100w, 85w, and 50 w.?

With 100w pep, imd was lousy at -30db pep (IMD3).? At 85 watts pep, IMD dropped like a rock.? ?At 50 watts pep, IMD dropped through the floor.?

The idea here is..... bond the grids directly to the chassis, then LESS drive required.? With 20-25 watts? LESS drive from the xcvr required, the? IMD of the xcvr will improve a bunch.?
As per usual, the limiting factor for system imd is not the amp itself, but the xcvr driving it... which is typ a magnitude worse than the amp.??

For those that missed it, the full story is here under the heading 'A Common "Bad Grid Idea" Super Cathode Drive' :



TLDR text below. Visit the web page to see the pictures.

"Floating grids on capacitors to add "negative feedback" is one of the worse things every done in grounded-grid triode PA's . This bad idea appears in the Collins 30L1 811A amplifier, and Japanese manufacturers copied the bad idea into their power amplifiers. Heathkit was also a victim of this engineering gaff in the SB-220 and SB-221 amplifiers. Here is how it started and filtered through Ham gear:

When I was designing PA's in the late 70's and early 80's, an employee of Eimac (who was also an author of many articles and a popular Radio Handbook) put considerable pressure on me to float the grids of 3-500Z PA's through small mica capacitors. He called the circuit a "super-cathode driven" amplifier. He wrote letters and called frequently, asking why I would not float the grids through small mica capacitors.

This quite likable fellow creatively "borrowed" this idea from the Collins 30S1, which was actually a proper application for this type of system. This system works in the 30S1 because it is a cathode-driven class AB1 tetrode. The 30S1, unlike later "copy-cats" using the floating-grid circuit, has zero control grid current. The grid has very high impedance all through the RF cycle. The high grid-cathode impedance does not shunt the upper capacitor divider with the low drive-varying grid resistance of stages with control grid current. Essentially R1 (see the circuit below) is infinite in the Collins 30S1. The 30S1, unlike triode copy cats, has a directly grounded screen. The screen shields the RF input (cathode) from the RF output (anode).

The theory seems pretty simple on the surface. Floating control grids through small mica capacitors forms a capacitive voltage divider, with the small grid-to-ground bypass capacitors forming the grounded half of a capacitive voltage divider. The small internal cathode-to-grid internal tube capacitance forms the upper leg of this voltage divider. Driving power requirements are increased by this negative feedback (the grid partially follows the RF cathode voltage, reducing effective grid/cathode voltage and reducing effective driving power applied to the grid). In theory, the amplifier should be "cleaner" and, with reduced power gain, be a closer match to higher power exciters.

Super cathode drive theory is the cathode to grid capacitance forms a divider with the grid bypass capacitor. This somewhat works in a class AB1 tetrode or pentode, because the cathode to grid circuit never biases into conduction. The idea falls apart with grid current in any amplifier, as well as in any triode.

After some thought, experiments, and questioning other engineers, I found no one actually measured performance or calculated feedback over a wide range of operating frequencies and control grid currents. It was assumed since everyone did it and an Eimac staffer endorsed it, super-cathode was already confirmed technically sound.

Good Feedback Dividers

In a good capacitive divider, sampled feedback voltage would be constant in both amplitude and phase regardless of frequency, power levels, and tuning. To be a "good" capacitive divider, the reactance of capacitors C1 and C2 would have to totally dominate system impedances. This is where the wheels fall off "super cathode drive".

We find a huge spike in grid-to-ground impedance at 2MHz, and very uneven response above that range. By manipulating the value of L1 (the grid chokes) we can move the spike around, but we are ALWAYS left with some low frequency where the grid isn't grounded! The Heathkit SB220, for example, peaks below the 160-meter band.

This is a very serious violation of good engineering practices in any grounded-grid PA, and is actually at the root of VLF and HF stability problems in a few popular PA's. Collins, for example, had a series of field modifications to the 30L1 grid system. They kept moving the spike around, trying to stabilize the amplifier. The best idea for the 30L1 Collins would have been to abandon the silly notion this system adds stable controlled negative feedback, and change the amplifier back to a true grounded grid with neutralization. If Collins wanted negative feedback in the 30L1, the PROPER method would have been the addition of a resistor in series with the cathode feed point near the tubes! We never want to float the grids in a grounded-grid triode amplifier.

There are obviously several major flaws with the super-cathode drive concept. Grid current causes grid-to-cathode impedance to constantly vary with drive level. When grid current is absent, the grid-to-cathode impedance is nearly an open circuit. Grid-to-cathode capacitance dominates the upper half of the divider, and everything appears to work as planned. Unfortunately, a problem appears whenever the grid draws current. Even the tiniest amount of grid current causes grid-to-cathode impedance to decreases rapidly. With only a few dozen milliamperes of grid current, grid impedance drops to a few hundred ohms or less. As grid current is drawn, the decreasing grid impedance dominates the upper leg of the voltage division circuit!

There are also new potentially destabilizing resonances added in the grid path.

This system causes four major problems:

Grid drive is effectively reduced as operating frequency is increased. This is the opposite of what we need! We need more drive to offset system inefficiencies on higher frequencies.
Feedback starts to show significant phase-lag with increased drive, especially on lower bands.
Grid-to-chassis impedance at VHF and LF is increased, making the amplifier much less stable. An SB-220 heath amplifier for example required nearly twice the parasitic choke inductance when the "super cathode" circuit was used. Still, because of pressure from this person, the circuit was added!
Protection for the exciter and cathode system, in the event of a tube arc, is greatly reduced. (see the mods on this link for the 572B and 811H amplfiers)

When I tested several amplifiers with this alleged "super-cathode" system added, IMD performance became significantly worse under some operating conditions. Stability also significantly decreased. Several amplifiers I tested using 572B, 3-1000Z, and 3-500Z tubes all had higher intermodulation distortion and required larger parasitic chokes when this super-cathode system was added!

Unless you have a class AB1 tetrode or pentode, ground the control grids directly with short heavy leads or use low-inductance high-value capacitors with very short leads to ground the control grids! The "super cathode drive" system system does not belong in any grounded grid triode amplifier. Get rid of it!"

...end of W8JI commentary.

73

-Jim
NU0C


On Sat, 30 Mar 2024 08:33:50 -0700
"Jim VE7RF" <jim.thom@...> wrote:

On Sat, Mar 30, 2024 at 04:46 AM, Richard Knoppow wrote:

<There was just some discussion about this, I think in this list. Drake's
idea was that the indirect grounding provided some negative feed back
which reduced the distortion. That may be desirable even at the price of
some reduction of power.
toggle quoted message Show quoted text ( #quoted-236065405 )



On 3/29/2024 6:53 PM, Jim Shorney wrote:

It is a PITA on an L7 due to the tight spaces. I just did a second one.
Works nice.
73
-Jim
NU0C

--
<Richard Knoppow
<Los Angeles
<WB6KBL
<SKCC 19998

Drake, heath, Henry, Kenwood, etc, all did the same thing.? It was a stupid idea, right outa the gate...and doesn't work anyway.? ?IF you want say 10 db of IMD reduction, you require 11 db of NFB...so not quite a 1:1 ratio.
And? 3 db of NFB means the gain just dropped by 3db...so now you require 200 watts? to drive the amp.... for a lousy 2.8 db of IMD reduction.? With 6 db of NFB,? it will require 400 watts to drive the amp....with a 5.9 db imd reduction.

Problem is, there is only one way to properly introduce NFB to any GG amplifier....and that's by inserting a low value resistor between the coupling cap and the cathode. Typ values are 10-20 ohms.? Problem with that is, the input Z of the cathode has now increased by 10-20 ohms, so now all your PI tuned input values are way outa whack.? 2nd issue is, the 10-20 ohm resistor has to handle a fair bit of power of power.? Typ, several smaller resistors are used in parallel to make up the 10-20 ohms. 3rd issue is... the gain of the amp dropped by a bunch, and now MORE drive required.

On the drake, heath, henry, Kenwood amps, they tried the NFB trick by using grid caps. (The RF choke is to allow DC grid current to flow).? ( grid caps will block DC grid current).? Those grid caps are 1/2 of a V divider.? The 2nd 1/2 of the V divider is made up of the internal grid to cathode C of the tube itself. The values used for each 1/2 of the V divider are out to lunch, and will not work anyway.? And if you do get it to work, it's for one band only.

What they managed to do with their fubar grid caps is..... the amp is now more UN stable.? Stability requires the best RF grnding of the grids you can get...and that's when each of the 3 x grid pins has it's own copper strap. And never use the screws used to submount those? Johnson style sockets.? I know of several TL-922 owners, that had instability issues and installed the nichrome junk.? Once the grids were bonded directly to the chassis, amp is rock stable...and oem suppressors were re-installed...abd amp still rock stable.? 20-25 watts less drive required.

What drake, et all ended up with is..... just a tiny bit, fractional db imd improvement...and only on one band at most.? ?And the fractional NFB results in lower gain....and typ 20-25 watts? MORE drive required.

When ARRL lab tested the Kenwood 850 yrs ago,? ?for whatever reason, they tested it on 20m, at 3 x different power levels, 100w, 85w, and 50 w.

With 100w pep, imd was lousy at -30db pep (IMD3).? At 85 watts pep, IMD dropped like a rock.? ?At 50 watts pep, IMD dropped through the floor.

The idea here is..... bond the grids directly to the chassis, then LESS drive required.? With 20-25 watts? LESS drive from the xcvr required, the? IMD of the xcvr will improve a bunch.
As per usual, the limiting factor for system imd is not the amp itself, but the xcvr driving it... which is typ a magnitude worse than the amp.

--

73

-Jim
NU0C

On Sat, Mar 30, 2024 at 06:27 AM, Clyde K2UE wrote:

<Generally speaking there are two types of tube transfer characteristics:

?

<The optimum characteristic can be described as two straight lines: constant transconductance at high input levels, and a second straight line of transconductance going from zero to the large signal value at low input levels.? When biased <at the half max transconductance point ideal class AB operation is obtained: for small signals there is half transconductance for both positive and negative going inputs, and for large signal there is twice transconductance for positive <going signals and cutoff for negative, i.e. the same time-average.? Triodes often have this characteristic, as the decreasing plate voltage keeps the gm (transconductance) from rising at high drive levels as the load line decreases the <instantaneous plate voltage.

?

<A non-optimum characteristic would be one where the transconductance is constantly changing, with no large constant regions.? This typical of many tetrodes, for instance, where the screen allows the gm to keep rising at hight currents <and the device is more constant current.

?

<In the latter case feedback can be very helpful in improving distortion, and without it IM tends to be awful no matter the bias point.? But in the former case the biggest factor is the bias point, and there will be a distinct dip in IM as the bias <is adjusted.? The 3-500Z is the former type, and you can see from the data sheets of various manufacturers that IM of -40dB or better can be obtained if the bias point is optimum (the 1500V zero bias case and the 3500V -15V bias <case).? An old-school analysis of the the tube transfer curves leads to the same conclusion.? So for the 3-500Z family choosing the optimum cathode bias voltage can be more effective than feedback in optimizing IM.

?

______________________________________________________________

<Clyde Washburn, k2UE

## NFB really only works for grid driven tetrodes, since they typ have high gain to begin with.? We measured 22.7 db gain on buddy's grid driven tetrode (tube with handles).? It used a 1k ohm grid termination. Had instability on 40-15m, so used a 500 ohm grid termination...and driven by an adjustable pi network. Parasitic suppressor installed right at the grid. After that, rock stable, and gain reduced to 20 db.?

The issue with say the heath SB-220 is..... it uses a grid choke on each socket, and the DC resistance of each grid choke is 25 ohms.? ?With 300 ma of grid current flowing, that's 150 ma per tube.? .150 X 25 = 3.75 vdc

That extra 3.75 vdc of UN wanted, and yo-yoing bias, is in addition to the oem 5 vdc zener bias.? ?Now the bias is swinging wildly from 5.0 vdc to 8.75 vdc.? That alone just screwed up the load line really bad.? The B+ regulation on the SB-220 is nothing to write home about...and sags badly.? With the UN regulated B+ dipping lower on peaks, that too just screwed up the load line.?

On my 4 x L4B's, the grids are bonded to the chassis with cu straps.? I also installed 10 x 1N5408 diodes for bias.? ( it idled at 220 ma? with Drakes zero bias scheme).? ?I tapped the diodes with a spdt-center off mini toggle..to obtain 3 x positions of bias.? On my HB metal tube amps, I use 50? x 6A10 diodes, with a 20 position rotary switch.? To make the diode bias scheme rock solid regulation, I wired a large value lytic across the entire string of bias diodes.? Does not budge. Without the cap, bias V will increase by 10-12%? between? idle current....and max plate current.?

These days, on the drake amps, I use all 10 diodes when on the SSB / 2650 vdc position? (and also the lower 1900 vdc position).? This dropped the idle current from 220 ma....down to 100 ma...and imd is superb? with 1290 watts pep out on 75m.? ?On CW position with 1900 vdc, idle is now just 40 ma.....and IMD is still superb with 625 watts pep out.? Aprx 7.3 vdc bias used. Slightly more drive required to overcome the 7.2 vdc of bias.?

?

The limiting factor is the IMD of the xcvr driving the GG amp.? No xcvr is good for -40db for IMD3.? An elecraft? K3 is good for -23 dbc for IMD3.?

?

The GG amp has it's input and output 180 degs out of phase. (cathode driven negative = grid driven positive).? That alone will? typ result in as much as a 6 dbc improvement in IMD.? The cathode on the GG amp is common to the input and the output.? With the PI tuned input installed, that feeds the cathode, the IMD can be substantially improved.? On my hb amps, on one of em,? I can adjust the Q between some wide extremes.? It uses a pair of air variable broadcast caps + a 0-4 uh roller coil (made from 8 ga wire). Both caps are padded on 160m only.? Eye opener to say the least.? Optimum linearity results with a total network Q of 4 ( input Q = 2..and output Q also = 2). It's a good compromise between IMD, BW, and circulating current in the coil.? The optimized PI tuned input also results in better harmonic suppression on the output side of the amp.

?

You can only do so much, but? bonding the grids on the L4B et-all? is a simple mod.? ?Adding a bit of bias to reduce heat is a bonus.? ?I tossed the drake oem pair of series 50k bleeders and the 5 kw? resistor. That dumps? a HUGE load off the supply.? ?Way less heat on the RF deck.?

开云体育