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

   Home   Help Search  
Pages: Prev 1 ... 6 7 8 9 10 [11] 12 Next   Go Down
  Print  
Author Topic: • Favorite Old Wives' Tales •  (Read 30465 times)
AG6K
Member

Posts: 1




Ignore
« Reply #150 on: November 19, 2011, 05:24:23 AM »

  Eimac states that each 3% increase in filament potential above that which produces full PEP decreases the emissive life of the tungsten dicarbide layer by 50%. 

The very first sentence Rich posted excludes his formula, and telling people to set voltage to 4.8 volts or any other fixed voltage. The key words Rich ignores are:
 
Quote
above that which produces full PEP

This is what happens when someone chooses what part of a long text he wants to use, and discards the rest.

  Didn't I mention that Econco Broadcast Services told me that fil-V should be 2% more than the PEP decrease point ?


Quote
If we include ALL of the text it is plain none of it allows blind application of some arbitrarily assigned voltage chosen by AG6K, or anyone else.

  what's blind about algebra?

Quote
As a matter of fact Eimac actually never intended that theoretical statement to be directly applied to anything.

  Does Eimac mention reducing filament-V to lengthen emissive-life in their larger power grid tubes Tom?

 
Quote
The actual full original filament management text is here:

http://www.bext.com/filament.htm


The newest revision is here:

http://www.w8ji.com/Vacuum%20Tubes/CPI%20AB-18%20New%20v2.pdf

The full Eimac CPI text on filament management disagrees with Rich.

  The full text is on p.163 of "Care and Feeding of Power Grid Tubes" and anybody who reads it for themselves can see that you are blowing smoke - which in German is the word rauch Mr. Rauch,
••• Rich, ag6k
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #151 on: November 19, 2011, 05:56:19 AM »

See how little fact is behind everything AG6K offers, and how it is his way or nothing?

1.) We never got past myth number one, where Rich thinks radiation comes equally from all parts of an antenna.

2.) We never got past the filament voltage formula he invented, that disagrees with plain text from CPI Eimac, and he never said he measured what Eimac says must be measured when filament voltage is reduced

3.) We never got past the complex impedance of the anode and Q of the system

4.) We never got past the balun on the input of the tuner

This is just a repeat of his Handbook and QST articles. A dozen or more people disagree with him, and give references. He references himself, and says the people who disagree have some agenda against the truth.

This pattern, which is very easy to see, repeats over and over.


73 Tom
Logged
AG6K
Member

Posts: 1




Ignore
« Reply #152 on: November 19, 2011, 05:59:40 AM »

Eimac said "theoretically a 3% increase in filament voltage will result in a 20º Kelvin increase in filament temperature, a 20% increase in peak emission, and a 50% decrease in life due to carbon loss." 

That's a small quote from a much larger text in which context is most important.

Expressing this in a formula results in: Relative Life = V2/V1)23.4.

But over what limits is the formula accurate?

  You seem to have missed the point that it applies only in the range where filament voltage is 2% higher than the PEP decrease point.  For my freshly Econco recarburized 8170 this was >6.8v.

Quote
It is not always accurate to look at a very small data range and extrapolate into a wide-range formula. But that's what you're doing.

  The range is defined to be small.

Quote
More important, Eimac states categorically that reduction in filament voltage should only be done if 1) the point of IMD degradation can be accurately measured, 2) the tube(s) have been "broken in" for a couple of hundred hours, and 3) the filament voltage can be carefully regulated and accurately measured.

  the absolute accuracy of the voltage measurement is not as important as being able to see a 2% increase in it once the PEP decrease point is determined. 

Quote
What this says to me is that, in an amateur-radio setup, I should only attempt life-extension by filament-voltage reduction if I have the required test equipment and know how to use it, plus an amplifier with a regulated or at least controlled filament supply.

  If you are worried by going too low on filament-V and, assuming that the tube's vacuum contains a worrisome number of air molecules, thereby posssibly contaminating Th atoms, having the fil-V rise a few % on receibe should seemingly put your mind at ease since increasing fil-V is the remedy for Th contamination.

Quote
And if I decide to use filament-voltage-reduction, I should evaluate the IMD performance at least every time I change bands, because operating conditions may be different on different bands.

  Econco told me that PEP measurement will do for determing the requisite point.  Care and Feeding ... says the point where:  "a noticeable reduction in  anode-current, power output, or an increase in distortion." -- so there's a choice.
•••  Rich, ag6k

Quote
It seems very clear to me that the Eimac verbiage was intended for applications where tubes are operated under the same conditons for long periods of time, such as broadcasting.

One of the biggest mistakes being made is the selective quotation of a small part of a big text, thereby losing the context and qualifications of the original.

73 de Jim, N2EY
Logged
N2EY
Member

Posts: 3877




Ignore
« Reply #153 on: November 19, 2011, 07:54:31 AM »

You seem to have missed the point that it applies only in the range where filament voltage is 2% higher than the PEP decrease point.

OK - so that means PEP should be measured every time the amp is retuned or operating conditions changed, and the decrease point determined, and the filament voltage set 2% above that. Right?

Quote
It is not always accurate to look at a very small data range and extrapolate into a wide-range formula. But that's what you're doing.

The range is defined to be small.

Not in the formula you gave.

Quote
More important, Eimac states categorically that reduction in filament voltage should only be done if 1) the point of IMD degradation can be accurately measured, 2) the tube(s) have been "broken in" for a couple of hundred hours, and 3) the filament voltage can be carefully regulated and accurately measured.

the absolute accuracy of the voltage measurement is not as important as being able to see a 2% increase in it once the PEP decrease point is determined.

Most ham stations are in homes where the line voltage can fluctuate somewhat. So a measurement system of adequate accuracy and precision is needed if an amp is to be operated that close to the edge.
  
Quote
What this says to me is that, in an amateur-radio setup, I should only attempt life-extension by filament-voltage reduction if I have the required test equipment and know how to use it, plus an amplifier with a regulated or at least controlled filament supply.

 If you are worried by going too low on filament-V and, assuming that the tube's vacuum contains a worrisome number of air molecules, thereby posssibly contaminating Th atoms, having the fil-V rise a few % on receibe should seemingly put your mind at ease since increasing fil-V is the remedy for Th contamination.

Nope.

What I'm saying is that the whole voltage-reduction thing is only worth doing if ALL the recommendations are followed. ALL. Not just some.

Quote
And if I decide to use filament-voltage-reduction, I should evaluate the IMD performance at least every time I change bands, because operating conditions may be different on different bands.

Econco told me that PEP measurement will do for determing the requisite point.

Then PEP should be measured at every retuning or change in operating conditions.

Quote
It seems very clear to me that the Eimac verbiage was intended for applications where tubes are operated under the same conditons for long periods of time, such as broadcasting.

One of the biggest mistakes being made is the selective quotation of a small part of a big text, thereby losing the context and qualifications of the original.

Which is the entire point.


73 de Jim, N2EY
Logged
WB6BYU
Member

Posts: 13169




Ignore
« Reply #154 on: November 19, 2011, 08:22:27 AM »

In looking for something else, I happened across W7EL's analysis of the a balun on the tuner input:

http://www.eznec.com/misc/ibalbrf.txt

Basically, the voltage across the balun, the current balance in the output, and the common mode
current in the input, are the same regardless of which side of the tuner the balun is placed.

I didn't download his further analysis, but it includes an experiment to validate these results.


But then, any good Conspiracy Theorist would, of course, discount W7EL's work because he writes
antenna modeling software based on the wrong model of radiation (regardless of how accurately
it models antennas in the real world.)  Besides, he is obviously one of Tom's groupies.
Logged
AG6K
Member

Posts: 1




Ignore
« Reply #155 on: November 19, 2011, 11:40:56 AM »

You seem to have missed the point that it applies only in the range where filament voltage is 2% higher than the PEP decrease point.

Quote
OK - so that means PEP should be measured every time the amp is retuned or operating conditions changed, and the decrease point determined, and the filament voltage set 2% above that. Right? [uote]

  The PEP does not need to be measured, a relative RF indication - or a plate/anode-I - and a filament VM is all that is needed. 

Quote
It is not always accurate to look at a very small data range and extrapolate into a wide-range formula. But that's what you're doing.

  the only significant item is +2%.

The range is defined to be small.
[/quote]

Quote
Not in the formula you gave.

  The limit is not in the formula, the limit is setting the +2%.  1% is not enough, 3% or more is $$ down the crapper. 

Quote
More important, Eimac states categorically that reduction in filament voltage should only be done if 1) the point of IMD degradation can be accurately measured,

  " ... or a noticeable reduction in anode current, or power output" ..." [ref:  Care and Feeding, p. 163]

Quote
2) the tube(s) have been "broken in" for a couple of hundred hours, and 3) the filament voltage can be carefully regulated and accurately measured.

 The typical few % increase in filament-V on receive can damage the tube?

the absolute accuracy of the voltage measurement is not as important as being able to see a 2% increase in it once the PEP decrease point is determined.

Most ham stations are in homes where the line voltage can fluctuate somewhat. So a measurement system of adequate accuracy and precision is needed if an amp is to be operated that close to the edge.
  
What this says to me is that, in an amateur-radio setup, I should only attempt life-extension by filament-voltage reduction if I have the required test equipment and know how to use it, plus an amplifier with a regulated or at least controlled filament supply.

  Hardly.  I once worked on a Henry 1KD-5 that had 5.9v on the filament of its 3-500Z @ 240vac in.  Would you be okay with this Jim because you knew that one of "our recognized amplifier experts" designed it?   5.9v instead of 4.9v on the filament is going to reduce emissive life to about 13%. 

 If you are worried by going too low on filament-V and, assuming that the tube's vacuum contains a worrisome number of air molecules, thereby possibly contaminating Th atoms, having the fil-V rise a few % on receive should seemingly put your mind at ease since increasing fil-V is the remedy for Th contamination.

Quote
Nope.

What I'm saying is that the whole voltage-reduction thing is only worth doing if ALL the recommendations are followed. ALL. Not just some.

Quote
And if I decide to use filament-voltage-reduction, I should evaluate the IMD performance at least every time I change bands, because operating conditions may be different on different bands.

Econco told me that PEP measurement will do for determining the requisite point.

Quote
Then PEP should be measured at every retuning or change in operating conditions.

  relative output indication, a filament VM.  and a means of adjustment is all that's needed.

Quote
It seems very clear to me that the Eimac verbiage was intended for applications where tubes are operated under the same conditions for long periods of time, such as broadcasting.

  I've heard lots of Amateur Radio operators do it.  .

Quote
One of the biggest mistakes being made is the selective quotation of a small part of a big text, thereby losing the context and qualifications of the original.

Quote
Which is the entire point.
73 de Jim, N2EY

 Morphing a gopher mound into a foothill.
••• Rich, ag6k


[/quote]
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #156 on: November 19, 2011, 01:39:12 PM »


If a person is unbendable and insists the rest of the world is wrong about EM radiation from antennas, and doesn't understand Rp and Q, and invents filament voltage formulas by discarding everything in the surrounding text he does not like (like warnings to not use reduced voltage on unregulated filament supplies), I'm not so sure anything he says can be trusted.
Logged
G3TXQ
Member

Posts: 1510




Ignore
« Reply #157 on: November 19, 2011, 02:27:31 PM »

In looking for something else, I happened across W7EL's analysis of the a balun on the tuner input:

http://www.eznec.com/misc/ibalbrf.txt

Basically, the voltage across the balun, the current balance in the output, and the common mode
current in the input, are the same regardless of which side of the tuner the balun is placed.
Dale,

I've been re-visiting Roy's analysis during the past few days and I'm coming to believe it is incomplete. Unfortunately his ftp site doesn't seem to work for me so I can't check his other work.

As presented the analysis appears sound; but crucially, when he introduces the tuner he assumes a 1:N transformer which has its lower input terminal commoned to the lower output terminal. That forces the voltage across Z2 to be the same as the voltage across Zw, and all his current balance formulas then work fine. However, if you assume alternative tuner configurations where the lower input terminal is not commoned to the lower output terminal, the voltages across Z2 and Zw can differ, thus invalidating Roy's balance formulas.

From the different worked examples I've tried so far, the situation never gets close to the simplistic "Zw needs to be 5*50 Ohms" that Rich was expounding earlier in this thread; however it appears not accurate to say that the requirements on Zw are identical whether the balun is at the tuner input or the tuner output - they may be the same, they be more severe, or they may be less severe, depending on several factors including the tuner topology.

It's getting late here and my head is spinning Smiley It would be good if someone else could confirm those thoughts or point out my error - I'd be very happy to be proved wrong!

73,
Steve G3TXQ
Logged
W7SMJ
Member

Posts: 120




Ignore
« Reply #158 on: November 19, 2011, 03:31:14 PM »

  Eimac said "theoretically a 3% increase in filament voltage will result in a 20º Kelvin increase in filament temperature, a 20% increase in peak emission, and a 50% decrease in life due to carbon loss."  Expressing this in a formula results in: Relative Life = V2/V1)23.4.  Verifying the formula:  (1.00v/1.03v)23.4 = 0.500.  Are you okay with the verification Scott?
•••  Rich, ag6k

Yes, I’m fine with that.  In fact I will say I am in complete agreement with the formula (X/(X)*1.03)23.4 = 0.5.  What I’m saying is I don’t believe you can read what Eimac wrote in total context and then say Relative Life = (V2/V1)23.4

Let’s look at the 3-500Z spec sheet published in 1968, the year after The Care and Feeding of Power Grid Tubes was published which you are basing your claims on.  It says “The rated filament voltage for the 3-500Z is 5.0 volts.  Filament voltage, as measured at the socket, must be maintained within the range of 4.75 to 5.25 volts to obtain maximum tube life“.  So I’m guaranteed maximum tube life by running my 3-500Z at filament voltage of 5.25V even though your formula says I should expect a reduction in life of 90%.  (4.75/5.25)23.4 = 0.096.  If your formula is valid, how can this be???

73,
Scott
Logged
AG6K
Member

Posts: 1




Ignore
« Reply #159 on: November 19, 2011, 03:34:16 PM »


If a person is unbendable and insists the rest of the world is wrong about EM radiation from antennas, and doesn't understand Rp and Q, and invents filament voltage formulas by discarding everything in the surrounding text he does not like (like warnings to not use reduced voltage on unregulated filament supplies), I'm not so sure anything he says can be trusted.

  The formula you are kvetching about Tom R. is based on paragraph 4. p.163 of Eimac's  in Care and Feeding ... .
•••  Rich, ag6k
Logged
AG6K
Member

Posts: 1




Ignore
« Reply #160 on: November 19, 2011, 03:58:29 PM »

In looking for something else, I happened across W7EL's analysis of the a balun on the tuner input:

http://www.eznec.com/misc/ibalbrf.txt

Basically, the voltage across the balun, the current balance in the output, and the common mode
current in the input, are the same regardless of which side of the tuner the balun is placed.
Dale,

I've been re-visiting Roy's analysis during the past few days and I'm coming to believe it is incomplete. Unfortunately his ftp site doesn't seem to work for me so I can't check his other work.

As presented the analysis appears sound; but crucially, when he introduces the tuner he assumes a 1:N transformer which has its lower input terminal commoned to the lower output terminal. That forces the voltage across Z2 to be the same as the voltage across Zw, and all his current balance formulas then work fine. However, if you assume alternative tuner configurations where the lower input terminal is not commoned to the lower output terminal, the voltages across Z2 and Zw can differ, thus invalidating Roy's balance formulas.

From the different worked examples I've tried so far, the situation never gets close to the simplistic "Zw needs to be 5*50 Ohms" that Rich was expounding earlier in this thread; however it appears not accurate to say that the requirements on Zw are identical whether the balun is at the tuner input or the tuner output - they may be the same, they be more severe, or they may be less severe, depending on several factors including the tuner topology.

It's getting late here and my head is spinning Smiley It would be good if someone else could confirm those thoughts or point out my error - I'd be very happy to be proved wrong!

73,
Steve G3TXQ

 RE:  Rich's antenna tuner, the 1500-ohm antenna, and the location of the balun.  With  1100W from my pair of 3-400Zs, there is 234v across the balun on the input.  If I move the balun from the input of the tuner to the output there will be 1284v across the balun.  Would it be a good idea to move the balun to where there would be more voltage across it?
•••  Rich, ag6k
Logged
AG6K
Member

Posts: 1




Ignore
« Reply #161 on: November 19, 2011, 04:21:38 PM »


If a person is unbendable and insists the rest of the world is wrong about EM radiation from antennas, and doesn't understand Rp and Q, and invents filament voltage formulas by discarding everything in the surrounding text he does not like (like warnings to not use reduced voltage on unregulated filament supplies), I'm not so sure anything he says can be trusted.

  chortle.  This from the man who tried to negotiate a settlement with me during the Grate Parasitics Debate.
•••  Rich, ag6k
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #162 on: November 19, 2011, 04:31:22 PM »

I've been re-visiting Roy's analysis during the past few days and I'm coming to believe it is incomplete. Unfortunately his ftp site doesn't seem to work for me so I can't check his other work.

It is incomplete.

Roy considers only an L, pi, or T network.

My webpage considers the others.

Quote
As presented the analysis appears sound; but crucially, when he introduces the tuner he assumes a 1:N transformer which has its lower input terminal common to the lower output terminal. That forces the voltage across Z2 to be the same as the voltage across Zw, and all his current balance formulas then work fine. However, if you assume alternative tuner configurations where the lower input terminal is not commoned to the lower output terminal, the voltages across Z2 and Zw can differ, thus invalidating Roy's balance formulas.

His formulas are only good for an L, T or Pi.

Quote
From the different worked examples I've tried so far, the situation never gets close to the simplistic "Zw needs to be 5*50 Ohms" that Rich was expounding earlier in this thread; however it appears not accurate to say that the requirements on Zw are identical whether the balun is at the tuner input or the tuner output - they may be the same, they be more severe, or they may be less severe, depending on several factors including the tuner topology.

Roy is fully correct with an L, T or Pi.

Rich is incomplete with everything.

Quote
It's getting late here and my head is spinning Smiley It would be good if someone else could confirm those thoughts or point out my error - I'd be very happy to be proved wrong!

With a floating balanced network, the system becomes unpredictable. Moving the balun can make it worse, better, or not change anything at all. It is random luck depending on common mode impedance and the component values required to match the differential impedance.

With a grounded center balanced network, no balun is really necessary if the load is perfectly balanced. Do the math on two L networks, one with the input grounded, and you will see that.

If, with a double network (balanced and floating), you want a current source to handle slight unbalance in the load or load common mode currents, the only way to do things is have a current balun on the output.

It really is a very complicated system when you add the balanced network, and include all the pronounced effects.

The bottom line is, the only time you can move the balun and use a poor balun on the input, is when you really don't need any balun at all. That makes sense when we think about it. Even a poor balun system works fine when it is not needed.

You follow what I am saying about the balanced network? The series reactance in the return line has the same current as the reactance in the output line, and drops the same voltage with the same phase shift. This pretty well balances the network even if the lower leg is just tied to ground. If you run a Monte Carlo analysis with tolerances in components, and in the load balance  limits, you will see you might just as well not use the balun at all. :-)


So in fairness to Rich his system works but, in completeness, it only works when the balun is unnecessary anyway. :-)

When the balun is absolutely necessary, then the rule about moving the balun not helping balance applies.

73 Tom
« Last Edit: November 19, 2011, 04:32:57 PM by W8JI » Logged
AG6K
Member

Posts: 1




Ignore
« Reply #163 on: November 19, 2011, 04:35:37 PM »

  Eimac said "theoretically a 3% increase in filament voltage will result in a 20º Kelvin increase in filament temperature, a 20% increase in peak emission, and a 50% decrease in life due to carbon loss."  Expressing this in a formula results in: Relative Life = V2/V1)23.4.  Verifying the formula:  (1.00v/1.03v)23.4 = 0.500.  Are you okay with the verification Scott?
•••  Rich, ag6k

Yes, I’m fine with that.  In fact I will say I am in complete agreement with the formula (X/(X)*1.03)23.4 = 0.5.  What I’m saying is I don’t believe you can read what Eimac wrote in total context and then say Relative Life = (V2/V1)23.4

Let’s look at the 3-500Z spec sheet published in 1968, the year after The Care and Feeding of Power Grid Tubes was published which you are basing your claims on. 

 The latest edition (2003) of Eimac/CPI's Care and Feeding. sent to me gratis by "Mr. X",  says that every 3% increase in filament V above what is needed to obtain full pep out cuts the life of the ditungsten carbide layer by 50%.

Quote
It says “The rated filament voltage for the 3-500Z is 5.0 volts.  Filament voltage, as measured at the socket, must be maintained within the range of 4.75 to 5.25 volts to obtain maximum tube life“.  So I’m guaranteed maximum tube life by running my 3-500Z at filament voltage of 5.25V even though your formula says I should expect a reduction in life of 90%.  (4.75/5.25)23.4 = 0.096.  If your formula is valid, how can this be???

  by reading between the lines. 

Quote
73,
Scott

Logged
AG6K
Member

Posts: 1




Ignore
« Reply #164 on: November 19, 2011, 04:48:18 PM »

I've been re-visiting Roy's analysis during the past few days and I'm coming to believe it is incomplete. Unfortunately his ftp site doesn't seem to work for me so I can't check his other work.

It is incomplete.

Roy considers only an L, pi, or T network.

My webpage considers the others.

Quote
As presented the analysis appears sound; but crucially, when he introduces the tuner he assumes a 1:N transformer which has its lower input terminal common to the lower output terminal. That forces the voltage across Z2 to be the same as the voltage across Zw, and all his current balance formulas then work fine. However, if you assume alternative tuner configurations where the lower input terminal is not commoned to the lower output terminal, the voltages across Z2 and Zw can differ, thus invalidating Roy's balance formulas.

His formulas are only good for an L, T or Pi.

Quote
From the different worked examples I've tried so far, the situation never gets close to the simplistic "Zw needs to be 5*50 Ohms" that Rich was expounding earlier in this thread; however it appears not accurate to say that the requirements on Zw are identical whether the balun is at the tuner input or the tuner output - they may be the same, they be more severe, or they may be less severe, depending on several factors including the tuner topology.

Roy is fully correct with an L, T or Pi.

Rich is incomplete with everything.

Quote
It's getting late here and my head is spinning Smiley It would be good if someone else could confirm those thoughts or point out my error - I'd be very happy to be proved wrong!

With a floating balanced network, the system becomes unpredictable. Moving the balun can make it worse, better, or not change anything at all. It is random luck depending on common mode impedance and the component values required to match the differential impedance.

With a grounded center balanced network, no balun is really necessary if the load is perfectly balanced. Do the math on two L networks, one with the input grounded, and you will see that. If, with a double network (balanced and floating), you want a current source to handle slight unbalance in the load or load common mode currents, the only way to do things is have a current balun on the output.

It really is a very complicated system when you add the balanced network, and include all the pronounced effects.

The bottom line is, the only time you can move the balun and use a poor balun on the input, is when you really don't need any balun at all. That makes sense when we think about it. Even a poor balun system works fine when it is not needed.

  Given:  50-ohm coax output, an extended double-zepp antenna, and a balanced L network to match 50-ohms unbalanced to the balanced double-zepp.  Should the balancing device go on the input to the tuner or its output?
•••  Rich, ag6k

Quote
You follow what I am saying about the balanced network? The series reactance in the return line has the same current as the reactance in the output line, and drops the same voltage with the same phase shift. This pretty well balances the network even if the lower leg is just tied to ground. If you run a Monte Carlo analysis with tolerances in components, and in the load balance  limits, you will see you might just as well not use the balun at all. :-)


So in fairness to Rich his system works but, in completeness, it only works when the balun is unnecessary anyway. :-)

When the balun is absolutely necessary, then the rule about moving the balun not helping balance applies.

73 Tom
Logged
Pages: Prev 1 ... 6 7 8 9 10 [11] 12 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!