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: [1]   Go Down
  Print  
Author Topic: 3CX800A7 / Heavy load?  (Read 8147 times)
K3JVB
Member

Posts: 837


WWW

Ignore
« on: May 22, 2010, 02:10:27 PM »

High all..
I have a question regarding loading a single 3cx800a7 amp

The amp is a rather rare Lk-750. On 160-80-40 meters it loads to an easy 1200 watts, with about 35 ma or less on the grid. And it draws 750 to 800 ma plate current. All with 60-70 watts drive.

But on 20 meters and up. To make this kind of power (1200w) it requires up tp 900-950 ma plate. And 80 watts drive.

The manual on this amp is rather brief on max power tune ups. It just warns to keep the grid below 36 ma. It seems a little happier at less plate current, and about 1000 watts. This keeps the plate at 800-850 ma. and I failed to mention that plate voltage was 2300v fulls load. The amp is wired for 240v.

What do guys you feel the safe full power load would be?
Thanks
John/k3jvb

Logged
W1QJ
Member

Posts: 2980




Ignore
« Reply #1 on: May 22, 2010, 03:23:24 PM »

Most, if not all multiband amplifiers (some better than others) require higher drive levels on higher bands.  Efficiency usually drops as frequency increases,as a result the higher drive increases plate and grid current with a reduction in output.  Almost every multiband amplifier makes less power on 15 and 10 meters than on 80 or 40 meters.  The effect of stray capacitance and inductance in switching and leads connecting to tank coils robs efficiency.  This is less prvilent on 80 and 40 than on 15 and 10 and makes a HUGE difference on 6 meters.  This is why you rarely see a 160-6 meter tube amplfier.  Some think it is the effectiveness of the tube at higher frequencies but that is not true at all.  A 3-500 will make as much power on 6 meters as it would on 80 meters in a monoband setting.  In the perfect world one would insist on a monoband amplifer for each ham band but that is not really practical or cost effective.
Logged
K3JVB
Member

Posts: 837


WWW

Ignore
« Reply #2 on: May 22, 2010, 03:56:35 PM »

Hi Lou,
Thanks. I realized that. But I guess I was fishing to see if 900 ma currunt draw , or more was was hurting the tube ?  1200 watts on 15 meters...is almost 1 amp draw. The grid is still in ,line at 36ma or less. But that just seemed a little seep for one tube.

I have seen your work on mono band 3-500's for 6 meters....LOL

John
Logged
K6AER
Member

Posts: 5743




Ignore
« Reply #3 on: May 22, 2010, 05:06:00 PM »

Your gird current is high  for a single 3CX800A7. Normally 40 ma will yeald 1500 watts out on a pair of 3CX800A7's. I imagine when you are getting 1200 watts out this is in the SSB mode and not CW. 1200 watts out on a 3CX800A7 will fry the grid in no time.

The tube sounds like it is getting tired and the high frequency gain is starting to go.

What are the currents Ip and Ig an drive level at 800 watts out?
Logged
K3JVB
Member

Posts: 837


WWW

Ignore
« Reply #4 on: May 22, 2010, 05:13:00 PM »

Grid-30 ma
Plate current-720ma
800 watts key down,cw mode @ 52 watts drive
Logged
K6AER
Member

Posts: 5743




Ignore
« Reply #5 on: May 22, 2010, 05:21:10 PM »

Your second posting came in just before my reply.

on fifteen meters at one amp times 2300 volts on your plate, your input wattage is 2300 watts. Minus the 1200 out and you are dissapating 1100 watts of heat on a tube rated for 800 watts continous dissapation. You are cooking the tube to death not to mention frying the gird.

You can run higher power out on SSB for the voice duty cycle is about 20-25%. In morse CW the duty cycle is about 45% including pauses. Key down CW is 100% duty cycle.

A fresh 3CX800A7 should require 20 mA grid current for 800 watts out.
Logged
K6AER
Member

Posts: 5743




Ignore
« Reply #6 on: May 22, 2010, 05:25:10 PM »

Our replays are passing in the night.

From you last reply my guess is the tube is getting soft. It still has a lot of life left in it but getting the last 20% out of the tube will just hasten it's demise. Go easy on the old girl for a new tube is about $650-750 in cost. Chinese tubes are cheaper.
Logged
K3JVB
Member

Posts: 837


WWW

Ignore
« Reply #7 on: May 22, 2010, 05:37:24 PM »

Point well taken. These readings are 4-5 second blips, to peak into a dummy load. I run it at about 400-500 watts cw.  I tune it for max power and reduce the drive to about 900 watts for SSB. These tubes are a little pricey to pound them with long tune-ups

I was wondering, because the manual is very vague on tuning. The amp does have grid and drive protection. I have only tripped it once at 40ma.  Amp supply does claim a 1200 pep out-put

Thanks for the info.
John/k3jvb
Logged
K6AER
Member

Posts: 5743




Ignore
« Reply #8 on: May 22, 2010, 10:31:55 PM »

I can not speak about the amplifiers grid protection circuit, maybe Tom, W8JI knows. 1200 watts out is pretty heavy for a single 3CX800A7. On my single 3CX800A7 amplifiers 1050 watts is about max in SSB.

Many hams have no idea what their amplifier real output power is due to Faith in cheap watt meters. Take your input power, volts times amps and figure about 55-60% efficiency. As you go up in frequency your efficiency will be reduced do to loses in the matching network.
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #9 on: May 23, 2010, 06:02:27 AM »

High all..
I have a question regarding loading a single 3cx800a7 amp

The amp is a rather rare Lk-750. On 160-80-40 meters it loads to an easy 1200 watts, with about 35 ma or less on the grid. And it draws 750 to 800 ma plate current. All with 60-70 watts drive.

But on 20 meters and up. To make this kind of power (1200w) it requires up tp 900-950 ma plate. And 80 watts drive.

The manual on this amp is rather brief on max power tune ups. It just warns to keep the grid below 36 ma. It seems a little happier at less plate current, and about 1000 watts. This keeps the plate at 800-850 ma. and I failed to mention that plate voltage was 2300v fulls load. The amp is wired for 240v.

What do guys you feel the safe full power load would be?
Thanks
John/k3jvb



John,

Amp Supply and Dentron were never especially good about engineering. You'll often find the tank circuits, for no reason at all, have some very strange operating Q's. The tank in itself is unlikely to be the culprit however, because it fortunately can be a mile off and not affect efficiency much. Errors in the tank mostly result in poor tuning range or needless tank heat, but the actual power loss or efficiency change is only a few percent for some pretty gross tank errors.

A problem like you describe sounds much more like tuned input problems. Commonly a large falling off in efficiency and high grid current at optimum output tuning is caused by rounding off of the cathode current transitions where the tube goes into and out of conduction, or excessive even-harmonic content in the cathode current waveform. This not only pushes the average anode current up for a given output, it pushes grid current up in a similar proportion.

To keep the waveform at the cathode crisp, and to keep grid current and anode current low for a given output, the cathode has to see a very low impedance at the second harmonic of the driving frequency. This means the tuned input has to be a C-L-C pi-network or a simple parallel tuned network with adequate loaded Q. That network has to located electrically close to the cathode of the tube, so looking back into the input circuit from the cathode the impedance is very low at second or all even harmonics. (If the impedance is high at the third harmonic and odd harmonics, efficiency can increase because the transition from conducting to off sharpens!!!).

My first guess would be Amp Supply either did not use a tuned input, or they placed it some distance from the cathode of the tube, or they used a wrong style of network. I have not actually seen that amp.

I never found this problem documented anywhere. I discovered it doing a pair of 3-500Z's for Heathkit. The prototype amp failed performance tests on ten, and barely passed on 15 meters. It took me a few days to discover changing the length of the coax between the tuned input and the tubes changed the ten meter problem. By just over doubling that lead length from the tubes to the tuned input, I could move the efficiency, high grid current, and two-tone IMD problem down to 20 meters!

After that I checked the Ameritron AL1200 and AL82, and discovered they had a small case of that problem. The AL1500 using the same configuration did not. The cure in the AL82 and AL1200, just as in the never released Warrior II Heath amp (they dropped the amateur line), was to reduce the impedance of the coax going to the tubes from the tuned input. This has the effect of moving the pi network capacitance toward the tube, and reducing the impedance seen by the tube cathode at the second harmonic of the drive frequency. This is why, if you look in an AL82 or AL1200, you will see two 50-ohm coaxial cables paralleled from the tuned input to the tube.

While no one can guarantee this is the problem in your LK, there could be other things, the description you have parallels the problems in amplifiers I have seen that were traceable to a high cathode impedance at the even harmonics of the drive frequency. This has NOTHING to do with exciter harmonics, the harmonics in question are generated at the cathode of the PA tube. With larger conduction angles, further from class B and toward class A, the problem goes away.

In the AL572 amplifier, to cure an efficiency problem on one band, I actually added a series tuned circuit that shorts the cathodes to ground at the second harmonic of the problem band!!! On other bands, to pass FCC, there has to be intentional series inductance with padding caps in the tank.

73 Tom
Logged
K3JVB
Member

Posts: 837


WWW

Ignore
« Reply #10 on: May 23, 2010, 07:51:57 AM »

Thanks all.

I think the safe bet is to keep the power at 900-1000 watts.
It tunes that very easy. I was looking at the Ameritron Al-800 specs to get an idea of what their max load is for this tube. They have slightly different grid limits. But claim 1200 watts.
If they don't hear me at 900...they won't at 1200 w either. Thanks again.
John
Logged
KG6YV
Member

Posts: 593




Ignore
« Reply #11 on: May 27, 2010, 07:29:15 AM »


Efficiency always drops with frequency.
Although that tube has 800W of plate dissipation, the difference between 800W and 1200W output isn't noticeable on the receive end.  At 800W you are already past the point of meaningful return which is probably 400-600W.  I would personally run it at around 800W output and save the tube for an extermely long life.

My $.02

Greg
KG6YV
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #12 on: June 03, 2010, 03:22:56 PM »


Efficiency always drops with frequency.
Although that tube has 800W of plate dissipation, the difference between 800W and 1200W output isn't noticeable on the receive end.  At 800W you are already past the point of meaningful return which is probably 400-600W.  I would personally run it at around 800W output and save the tube for an extermely long life.

My $.02

Greg
KG6YV


There is nothing that says efficiency has to drop as frequency is increased except design compromises necessary to meet a cost target.

Logged
W1QJ
Member

Posts: 2980




Ignore
« Reply #13 on: June 05, 2010, 05:19:10 AM »

TOM.... Since I have owned several AL-82 and AL-1200 amps, and repaired many as well, I recall seeing the single coax and the double coax leads to the tube on the input.  I often wondered why this was changed from one to 2 coax runs.  Now I know why.  This is VERY interesting to say the least!!!  When you went from the 1 to 2 coax leads, did that change the values of CLC in the input network?  Also, Your comment about the input must be or should be CLC, I know a few ham amps that the design of the inputs are such that by the time you get to 10 meters C2 is nonexistent.  The Ten Tec 425, SB-200 and a couple of others.  Can the input C of the tube count as C2?  Also you made an interesting comment about input matching circuits and plate and grid currents.  This is another FASCINATING issue.  As one that experiments with different input matching circuits, I have noticed the plate current and grid current changes you mention.  I have also noticed that on some occasions a "flat" input swr from a particular match does not always give the best drive from exciter to tube.  Sometimes a higher input swr will yield more drive.  It is amazing how this one issue of many with amplifiers can be so vast as to what can and should be done.  You mentioned a "simple" parallel tuned input as well.  That would be an inductor and cap to ground on the input line to he tube.  I never experimented with that, but that idea intrigues me.  Any pros or cons to that approach?  I think you can dedicate a small book to the input tuning issue with amps.  Lou
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #14 on: June 05, 2010, 01:51:33 PM »

TOM.... Since I have owned several AL-82 and AL-1200 amps, and repaired many as well, I recall seeing the single coax and the double coax leads to the tube on the input.  I often wondered why this was changed from one to 2 coax runs.  Now I know why.  This is VERY interesting to say the least!!!  When you went from the 1 to 2 coax leads, did that change the values of CLC in the input network? 

Yes, some bands required different values but I can't recall which bands or the changes. It has been years.

My only recent changes were to the 1200 to use the 1200Z7 tube, improve cooling, and reduce noise.


Quote
Also, Your comment about the input must be or should be CLC, I know a few ham amps that the design of the inputs are such that by the time you get to 10 meters C2 is nonexistent.  The Ten Tec 425, SB-200 and a couple of others.  Can the input C of the tube count as C2?
 


To a limited extent it can. Still, looking back from the cathode to ground at the cathode, we want the lowest possible impedance at the second harmonic.

I had to add a series L/C circuit to accomplish that on a problem band in the AL572.


Quote
Also you made an interesting comment about input matching circuits and plate and grid currents.  This is another FASCINATING issue.  As one that experiments with different input matching circuits, I have noticed the plate current and grid current changes you mention.  I have also noticed that on some occasions a "flat" input swr from a particular match does not always give the best drive from exciter to tube.  Sometimes a higher input swr will yield more drive.  It is amazing how this one issue of many with amplifiers can be so vast as to what can and should be done.


There is more going on than meets the eye, and like you I always spend time and resources getting to the bottom of any odd behavior.

Quote
You mentioned a "simple" parallel tuned input as well.  That would be an inductor and cap to ground on the input line to the tube.  I never experimented with that, but that idea intrigues me.  Any pros or cons to that approach?
 

There isn't much resistance transformation that takes place, but it can be an effective "ground" for cathode harmonics if it is close to the tube. I rarely use that system because few tubes are 50 ohms at the cathode.

73 Tom
Logged
Pages: [1]   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!