Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: Prev 1 2 [3] 4 5 Next   Go Down
  Print  
Author Topic: Copper Strip For Grid Grounding in TL922A  (Read 21718 times)
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #30 on: July 30, 2015, 03:14:32 PM »

More questions than answers.
 
I was speaking to the Collins 30S1, SB220, or TL922A.  Nothing to do with the 811 or 572B fitted amps. 

The reason for the capacitor connected from Anode to CG in the 30S1 is to provide negative feedback to the CG.  That capacitor is unnecessary in the triode configuration because the inter electrode capacity is adequate whereas the tetrode with the screen grounded, requires an external feedback path.
 
The PG feedback from the capacitor in the tetrode and the PG feedback from the inter electrode capacity in the triode are the same animal doing the same thing for the same reason. 

The PG capacitor in the Tetrode and the inter electrode capacity in the triode provide the top leg of the bridge that generates the feedback, and the G to Ground capacitor the lower leg.

The Grid in the 30S1 is not floating.  It has DC negative bias and ALC which prevent grid current, and a capacitor to ground for RF.  The combination is AB1 with negative feedback.

The grid in the TL-922 or SB220 are no more floating than the 30S1.  Why does the 30S1 get an OK nod regarding stability and the others not so?  There is a reason but I’m not convinced it’s a stability issue. 

Does it have a hard ground?  No but any coupling from anode or cathode to grid will cause negative feedback.   

What’s the difference between the 30S1, TL922A, or Heathkit?  Well organized bias, ALC, and AB1 mode.  Why would any be unusually unstable?

This is in no way meant to abrasive.  These are interesting questions that peak the grey matter.  Some of you have been down this path so please humor me.  Just haven’t heard enough to logically connect the dots of reality.

Thanks for taking the time to respond

Regards Jim
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #31 on: July 30, 2015, 05:32:07 PM »

Tom

What does a series resistor do in the cathode circuit to improve feedback?  I've seen comments where a series resistance will improve negative feedback, but I don't see the mechanism that provides this supposed advantage?

Pete

The resistor is in series with the cathode, and is NOT bypassed. When the cathode current increases, the opposing signal bias across the resistor increases. When cathode current decreases, the bias decreases. This reduces drive by reducing grid and plate swing. As long as cathode to ground shunting capacitance is reasonable reactance compared to the resistance, this makes pretty good RF negative feedback.

It really helps in some grounded cathode tetrodes or pentodes.

It is a good way to reduce gain by adding some NFB, and it does it without messing up performance. I've yet to see it cause a problem or make things worse.
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #32 on: July 30, 2015, 08:28:49 PM »

Tom,

One of the folks suggested that I read your explanation and calculations regarding the frailties of Bill Orr's Super cathode negative feed back system.  Well written.  Did so and believe you might review the explanation of the bridge.  Also, the spice chart for the SB220 that is in the http://www.w8ji.com/vhf_stability.htm page is in error.  Seems to suggest a high impedance spike in or very close to the 160 Metre band with the capacitor and choke combo to ground.  Don't believe it's anywhere near the HF bands.  Probably just a mis-calculation.

Regards Jim   
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #33 on: July 31, 2015, 02:52:29 PM »

Tom,

One of the folks suggested that I read your explanation and calculations regarding the frailties of Bill Orr's Super cathode negative feed back system.  Well written.  Did so and believe you might review the explanation of the bridge.  Also, the spice chart for the SB220 that is in the http://www.w8ji.com/vhf_stability.htm page is in error.  Seems to suggest a high impedance spike in or very close to the 160 Metre band with the capacitor and choke combo to ground.  Don't believe it's anywhere near the HF bands.  Probably just a mis-calculation.

Regards Jim    


At the bottom of the VHF stability page noted in the above quote there is a bullet point --- “GOOD FEEDBACK DIVIDERS”.  The schematic displayed in the Good feedback dividers is simply incorrect.  It totally ignores the fact that the feedback is not only from the cathode but from the anode as well.  It doesn’t include the feedback path from the Anode.  Also, the spice chart calculations are patently incorrect regarding the SB220 or for that matter any of the Super Cathode implementations.  The chart is also skewed badly to support your assertions in the writing.  
  
After reviewing these anomalies and incorrect information, I retract the statement that I made earlier regarding the direct grounding of the grid in a GG amplifier.

I cannot see any reason to avoid the Super Cathode Drive configuration for safety, stability, or performance reasons at this time.  It has merit that is worth exploring.  

If there is a question regarding instability causing damage to the amplifier, start looking for other reasons i.e. under rated or improperly designed band switches and frail design of T/R switching systems, not phantom instability.

G-Day Jim



« Last Edit: August 12, 2015, 01:12:47 PM by K9AXN » Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #34 on: July 31, 2015, 06:01:47 PM »

Tom,

One of the folks suggested that I read your explanation and calculations regarding the frailties of Bill Orr's Super cathode negative feed back system.  Well written.  Did so and believe you might review the explanation of the bridge.  Also, the spice chart for the SB220 that is in the http://www.w8ji.com/vhf_stability.htm page is in error.  Seems to suggest a high impedance spike in or very close to the 160 Metre band with the capacitor and choke combo to ground.  Don't believe it's anywhere near the HF bands.  Probably just a mis-calculation.

Regards Jim   

Jim,


What bridge are you talking about?

What frequency do you think a parallel 660pF capacitor and 10,000 nH inductor in the example circuit (taken from one of Orr's papers) would be resonant on? Are you saying all of the standard resonance formulas and SPICE are incorrect?

What part of F=1/(2pi(LC)^.5) is incorrect?

Tom



 
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #35 on: July 31, 2015, 06:08:47 PM »

At the bottom of the VHF stability page noted in the above quote there is a bullet point --- “GOOD FEEDBACK DIVIDERS”.  The schematic displayed in the Good feedback dividers is simply incorrect.  It totally ignores the fact that the feedback is not from the cathode but from the anode.  It doesn’t include the feedback path from the Anode.  Also, the spice chart calculations are patently incorrect regarding the SB220 or for that matter any of the Super Cathode implementations.  The chart is also skewed badly to support assertions in the writing.  
  
After reviewing these anomalies and incorrect information, I retract the statement that I made earlier regarding the direct grounding of the grid in a GG amplifier.

I cannot see any reason to avoid the Super Cathode Drive configuration for safety, stability, or performance reasons.  It has merit that is worth exploring.  

If there is a question regarding instability causing damage to the amplifier, start looking for other reasons i.e. under rated or improperly designed band switches and frail design of T/R switching systems, not phantom instability.

G-Day Jim





You obviously have no idea how the Orr circuit is configured. You must be looking at something else somewhere. You might want to look up some of Orr's circuits and have a look at them.

It is a grounded grid system. The cathode is driven. The cathode has a divider, formed by the grid-cathode capacitance and the grid to ground capacitance.

The grid is in the middle of that divider. The output in not part of that path.

It is not that complicated of a circuit. My explanation reflects the actual Orr circuit.

I don't know what on earth you are thinking of. Maybe you can post a link? Do you think the feedback is from anode to grid? Where is that circuit at?



73 Tom
« Last Edit: July 31, 2015, 06:11:49 PM by W8JI » Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #36 on: July 31, 2015, 07:33:01 PM »

    
[/quote]from Tom

Jim,


What bridge are you talking about?

What frequency do you think a parallel 660pF capacitor and 10,000 nH inductor in the example circuit (taken from one of Orr's papers) would be resonant on? Are you saying all of the standard resonance formulas and SPICE are incorrect?

What part of F=1/(2pi(LC)^.5) is incorrect?

Tom
------------------------------------------------------------------------------------------------------------------------------------------

Tom,
 
You referred to the SB220 when you ran the formula.  Neither the SB220 nor any of the other amplifiers use a 10000nh choke.  The SB220 uses a 1mh choke and if you calculate the resonant point of any of the amps that use Mr Orr's design or the Collins configuration you will immediately see that they are resonant somewhere around 500Kc NOT 2Mc.  

------------------------------------------------------------------------------------------------------------------------------------------
[/quote]  From Tom

You obviously have no idea how the Orr circuit is configured. You must be looking at something else somewhere. You might want to look up some of Orr's circuits and have a look at them.

It is a grounded grid system. The cathode is driven. The cathode has a divider, formed by the grid-cathode capacitance and the grid to ground capacitance.

The grid is in the middle of that divider. The output in not part of that path.

It is not that complicated of a circuit. My explanation reflects the actual Orr circuit.

I don't know what on earth you are thinking of. Maybe you can post a link? Do you think the feedback is from anode to grid? Where is that circuit at?


73 Tom

[/quote]

------------------------------------------------------------------------------------------------------------------------------------------

Tom,

I explained yesterday 03.14.32 how the feedback is implemented --- you must have missed it.  Look at the Collins users manual 3.7 on page 3.5 or look at Bridge C103/C104 on the schematic.  Also look at yesterdays post. 

I believe you stated the Bill Orr copied the Collins 30S1 configuration.  

A great day to you --- Kindest regards Jim  
« Last Edit: July 31, 2015, 07:43:39 PM by K9AXN » Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #37 on: July 31, 2015, 09:24:47 PM »

You didn't explain anything, Jim.

I don't think you understand the circuit at all.
Logged
G3RZP
Member

Posts: 1096




Ignore
« Reply #38 on: July 31, 2015, 11:21:13 PM »

I think I can see the reasoning about letting the grid float. Leaving aside for the moment whether that reasoning is correct, once grid current starts, there is a completely different phase shift involved, because the grid cathode admittance is no longer completely capacitive. Even before grid current starts, above about 15 to 20 MHz with most transmitting tubes, there will be a shunt conductive part to the tube input admittance because of transit time - inherent because of physical size.

So you have a capacitive divider with the bottom capacitor - grid to ground - shunted by a conductance that is frequency and drive level dependent, and by a capacitance that is drive level dependent because of Miller effect and the fact the tube isn't completely linear. In the case of tubes like 811s and 572Bs, you also have the effect of grid lead inductance to add to the frequency dependencies.

Once you add a screen grid and run AB1 with no grid current, you have a totally different situation. Especially with tubes that have a high gm and thus a small grid - cathode spacing, minimising transit time.

So my take on floating the grid with small capacitors in GG triode amplifier follows Lewis Carroll - "Bold, heroic was the notion"
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #39 on: August 01, 2015, 08:49:05 AM »

Tom,

Your calculations regarding the position of the high impedance point in the SB220 being close or within the 160 metre band are now a moot point.  They are incorrect and the concern is pointless.  All or these amps are well below 800Kc, not 2Mc.  Please confirm, is this true or not so?
 
Regarding your theory where a bridge with the grid at the center point of two capacitors, one connected to ground and the other, gc capacity.  The negative feedback effect from the CG capacity is reduced not enhanced.
 
Think about this.  The cg capacitor is in parallel with the cg resistance.  The low resistance would swamp any effect the capacity would exert on the grid bridge that has a 5/660 ratio (Heathkit).

If the grid is grounded the current leaving the grid will be rectified pulsating ac.  All energy sent to the dumper.

If it is the Orr configuration, the current leaving the grid will be DC, not pulsating anything.  This leaves a full +- sine influence on the grid proportional to cathode. 

Visualize this.  The grid is the anode, the choke is as in your favorite amp final circuit, the capacitor is the coupling cap to the tuned circuit, at ground, a very low impedance.  This combination results in negative feedback with both polarities because the cathode current provides the negative charge to the shunt cap and the choke provides the positive half of the cycle just like the output circuit of your amp.

The feedback from the anode with the triode is just frosting but your proposed bridge actually reduces feedback.

The Orr idea Sounds like a very good configuration to me.  Might take a bit more drive but the linearity has to be better. 

We're out of here for a few days, won't answer until midweek.

A great day to you!

Kindest regards Jim 
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #40 on: August 01, 2015, 09:15:05 AM »

Hi Pete,
I believe the effect/ transition from no grid current to grid current is mitigated by the Orr configuration. Before grid current flows, the GC and PG capacity play and when current begins, the capacity of the shunt becomes effective, leaving the major DC component to be carried off by the choke.  

There will be less AC current compared to the magnitude of the grounded configuration.

In the grounded configuration everything is pulsating AC after current flows.  Here the transition is profound.  Not so with the Orr config.

A super day to you!

Kindest regards Jim
« Last Edit: August 12, 2015, 01:04:08 PM by K9AXN » Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #41 on: August 01, 2015, 08:29:42 PM »

I think I can see the reasoning about letting the grid float. Leaving aside for the moment whether that reasoning is correct, once grid current starts, there is a completely different phase shift involved, because the grid cathode admittance is no longer completely capacitive. Even before grid current starts, above about 15 to 20 MHz with most transmitting tubes, there will be a shunt conductive part to the tube input admittance because of transit time - inherent because of physical size.

I'm not sure how much the IMD worsening comes from Miller effect, but I suspect not much.

Measurements at Ameritron and at Heathkit showed the IMD was worse on lower bands.

I always reasoned this was because Orr used just the filament to grid capacitance as the upper cap in the divider, and the reactance was far greater on lower bands. The grid-K conductance swamped that capacitnce worse with lower frequency.

This also had the weird effect of reducing the gain most on ten meters, where gain was already rolling off.

The circuit worked well in the 30S1 tetrode because it was AB1 with grounded screen, but it made anything else worse.

Quote

So you have a capacitive divider with the bottom capacitor - grid to ground - shunted by a conductance that is frequency and drive level dependent, and by a capacitance that is drive level dependent because of Miller effect and the fact the tube isn't completely linear. In the case of tubes like 811s and 572Bs, you also have the effect of grid lead inductance to add to the frequency dependencies.

Once you add a screen grid and run AB1 with no grid current, you have a totally different situation. Especially with tubes that have a high gm and thus a small grid - cathode spacing, minimising transit time.

So my take on floating the grid with small capacitors in GG triode amplifier follows Lewis Carroll - "Bold, heroic was the notion"


Yes, the road to IMD hell is paved with good intentions. A simple resistor would have reduced gain without adding distortion and stability issues. Plus at some lower frequency, depending on components used, the grid isn't grounded at all.
Logged
G3RZP
Member

Posts: 1096




Ignore
« Reply #42 on: August 01, 2015, 11:53:50 PM »

If lifting the grid of a gg triode has so many advantages in getting low IMD, how come those building BIG linear amplifiers - 30kw and up - never used it? Companies like Marconi, Standard Telephones and Racal?
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #43 on: August 04, 2015, 02:34:04 PM »

If lifting the grid of a gg triode has so many advantages in getting low IMD, how come those building BIG linear amplifiers - 30kw and up - never used it? Companies like Marconi, Standard Telephones and Racal?

Pete,

If I was a commercial buyer in the market for an HF amplifier, there are things that I certainly wouldn’t do.

Use an amplifier who’s design to date was unproven or questionable --  (Super Cathode drive).
Use 4 811’s in a parallel configuration or even use them at all. 
IMD performance?  Lowest, or non-existent on the priority list.
Don’t know what the lust for a few db improvement in IMD is???

Super cathode drive is still experimental and should remain in the amateur not commercial community.  Too many unanswered questions.
 
Any comments that I made or make in this thread are essentially to the use of the 3-500 as used in the SB-220, not the 811 or 572B designed amplifiers.

I believe there are a large number of faults blamed on the Super cathode drive design that are actually associated with the transition of the antenna relay while the tube is engaged.  Also believe band switch arcing is primarily the fault of the marginal design and configuration of the switch.

Just walked in --- four days and no bent parts in the wing.

Kindest regards Jim 
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #44 on: August 05, 2015, 05:23:16 PM »

You obviously have no idea how the Orr circuit is configured. You must be looking at something else somewhere. You might want to look up some of Orr's circuits and have a look at them.

It is a grounded grid system. The cathode is driven. The cathode has a divider, formed by the grid-cathode capacitance and the grid to ground capacitance.

The grid is in the middle of that divider. The output in not part of that path.

It is not that complicated of a circuit. My explanation reflects the actual Orr circuit.

I don't know what on earth you are thinking of. Maybe you can post a link? Do you think the feedback is from anode to grid? Where is that circuit at?

73 Tom

Good afternoon Tom,

I have to disagree. The bridge in the SB220 is composed of the (center leg 600pf paralleled with 1mh choke connected between the ground and grid), (the left leg 8.3pf interelectrode capacity between cathode and grid), (the right leg 4.7pf interelectrode capacity between anode and grid).

The left leg is negative feedback sensed from the input and the right leg negative feedback sensed from the output.

The feedback from the anode passes through a capacity of 4.7pf and cathode 8.7pf however the ac voltage applied to the anode side is perhaps 15 to 20 times that of the cathode side resulting in 7 to 10 times the negative feedback from the anode side of the bridge as compared to the cathode side. QST July, 1967 p. 38 bottom.

The high impedance spike due to the parallel resonance of the grounded leg, 600pf and 1mh choke, occurs at approximately 250kc, nowhere near 2mc.

Regarding the Miller effect. The miller effect on the cathode and anode sides are insignificant before grid current flows.  When current flows the negative Miller effect increases but it remains insignificant because it's terminated through a 600pf capacitor to ground. In the grounded grid scenario the Miller effect has a negative not positive effect on capacity because the end points are of the same polarity.

Thoughts?

Kindest regards Jim
Logged
Pages: Prev 1 2 [3] 4 5 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!