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: Matched Pair of 3-500Z --- how should they glow?  (Read 2953 times)
N2AXZ
Member

Posts: 90




Ignore
« on: April 08, 2006, 05:14:40 PM »

Well, I finally got the leaky 3-500Z that I had gotten from RF Parts replaced --- it turns out that there was a small hairline crack in the envelope right up near the anode pin that was almost impossible to see.....

Anyway, I have a new matched pair of 3-500Z's in the homebrew, and I finally got the plate supply all hooked up and threw the switches for the first time. Letting the tubes idle for several minutes allowed the plates to get a nice shade of red, but I notice that the plate of one tube glows significantly brighter than the other.  Is this to be expected?

Operating parameters for the tubes (grounded grid) are:

Vplate = 3560 V
Iplate = 100 mA (total for both tubes)

Both cathodes are biased at 15 volts positive with respect to the grids. Both tubes are forced air cooled by pressurizing the chassis beneath the tube sockets.

Any observations or thoughts are appreciated.
Logged
W5HTW
Member

Posts: 729


WWW

Ignore
« Reply #1 on: April 08, 2006, 05:42:58 PM »

Not good news!  If they are properly biased, why would the plates get red??  100 ma plate current idling?  That would not make the plates of a pair of 6L6s red.  Those red plates would be an indication of way too much plate current.    If the tubes are biased properly, there should be no red color, except on the filaments!  I know of no tube that has glowing red plates when it is biased to cutoff or idling.  

Something doesn't fit.  

Presumably there's a cathode bias resistor in the circuit.  I'd check it for proper value.

I would also remove one of the tubes and see if the plate current for the remaining one drops to about 50 ma.  Then I'd swap them.  And I'd try the good tube in each socket independently.  

I guess that bias reading you are getting of 100 ma is on the tube's plate current meter, yes?  

Anyway, take a look at the bias resistor.  

Ed
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #2 on: April 08, 2006, 08:02:01 PM »

I will try looking at the plate current with one tube at a time ---- that's a good idea.

I am not using a cathode bias resistor ---- I am using one of those "Triode Boards" that uses a large PNP transistor with an adustable zener in the feedback loop to provide cathode bias.  I have adjusted the cathode bias to 15 volts.
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #3 on: April 08, 2006, 08:03:13 PM »

Almost forgot ---- the plate current is measured with both the plate meter in the amplifier and an external ammeter.  Both of these meters agree with each other.
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #4 on: April 08, 2006, 08:50:06 PM »

Okay --- I tried each tube by itself.  The tube which had the brighter anode draws 50 mA by itself, and the tube which had the dimmer anode draws 39 mA by itself.  Even under these conditions, the anodes do glow.

I am wondering about your comparison of a 6L6 to a 3-500Z, and whether or not this is a valid comparison.  The maximum plate voltage rating for a 6L6 is 500 volts, whereas for the 3-500Z it is 4000 volts.  I am operating the tubes at a plate voltage of 3560 volts, so with a 50 mA plate current the total plate dissipation is 178 watts, which is almost six times the plate dissipation rating of a 6L6.  Seems like a 6L6 is in an entirely different league.  

I am not trying to sound critical here --- if I am wrong in my assumptions, please correct me.  This is the first amplifier I have built, so I am trying to find my "mental bias point" for how the amplifier is supposed to behave over its entire operational envelope.
Logged
KA5N
Member

Posts: 4380




Ignore
« Reply #5 on: April 09, 2006, 05:23:34 AM »

Consider the 6L6 comparison an off-hand comment.  He meant that the 3-500Z has a plate dissipation rating of 500 watts (1000 watts for a pair) and at 100ma and 3560 volts the plates were dissipating only 356 watts total.  This is well under the level which would cause red color.
You said this is a homebrew amplifier.  Was it working before you changed tubes?  Sounds like you have some major problem with the circuitry such as a short in the cathode/filament circuit, so that you are drawing much more current than you measure.  You need to check the amp throughly before you ruin a new set of tubes.
Allen  
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #6 on: April 09, 2006, 07:19:55 AM »

This is a new construction project, so the amplifier has never been operational.

I will check the circuit once again thoroughly, as you correctly suggest.  However, the currents that I measure with both tubes in place and then each tube individually jive very closely with what the data sheet for the tubes suggests.  I am measuring the current with a digital multimeter in series with the cathode return circuit --- between the center tap on the filament transformer and HV-.

I think that the only other way current can be sneaking around this measurement is if there is excessive grid current.  My grid current meter is not indicating any current flow.  Perhaps I can check this by looking at the voltage across the safety resistor which is placed between chassis ground and HV-.  This is a 17 ohm resistor.  Do you have any thoughts on what the value of this safety resistor should be?  I have seen some applications where the resistor is quite small as I am using, but I have seen others where it is much larger, perhaps 1Kohms or so.  Is it possible that I need to make the value of this safety resistor larger?
Logged
KA5N
Member

Posts: 4380




Ignore
« Reply #7 on: April 09, 2006, 11:15:54 AM »

Since I don't have a diagram to look at, I can't be sure what is going on.  Excessive cathode current can't be going through the grid, it can't dissipate that kind of power.  
You don't show any grid current.  Well you have no RF drive.  You show only 100 ma plate current but the plates are running red.  You must be drawing more current than indicated.
I suspect that your bias circuit and metering system is miswired or something has gone bad.  If your filament transformer has a path to ground, the noted symtoms would occur.  The grid safety resistor value isn't critical as it's function is too open if there is a short in order to save the tube.
Maybe someone else has a better idea.
GL Allen
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #8 on: April 09, 2006, 12:24:45 PM »

Okay ---- I think I may have found a clue to what is going on.  I started to think that perhaps the tubes are oscillating, so I placed a scope across the grid protection resistor.  This is a 17 ohm resistor connected between chassis ground and HV-. It is important to remember that I am not applying RF drive at any point in my testing and measurements.  

When I close the PTT circuit and allow plate current to flow, I measure 100 mA of plate current, as before.  However, looking at the scope, I can see AC across the grid protection resistor.  The frequency of oscillation is about 870 kHz, and the amplitude is about 150 mV pk-pk.

The parasitic suppressors are made from 3 150 ohm, 2 watt  carbon resistors in parallel, with three turns of flat silver plated copper strap encircling them.  I have one of these on the anode of each tube.

So, it looks as if the amplifier may be oscillating.  How should I go about troubleshooting this?  Should the structure of the parasitic suppressors be different?

Thanks again!

David
Logged
KA5N
Member

Posts: 4380




Ignore
« Reply #9 on: April 09, 2006, 02:42:27 PM »

The grids should be bypassed to ground with disc capacitors with short leads.  This should take care of the AC signal you are seeing.  It is probably just stray pickup not oscillation.  If the circuit was oscillating it would be either high HF or VHF not 800-900 kHz.  Also if the circuit was oscillating, the plate current would be much higher than 100 ma.
Allen
Logged
WB2WIK
Member

Posts: 21837




Ignore
« Reply #10 on: April 10, 2006, 12:53:08 PM »

I'd say also if it was oscillating, the peak RF voltage measured wouldn't be only millivolts.

A 3-500Z at 3500V and 50 mA is dissipating 175W and absolutely can glow red (anode color) after a couple of minutes.  It won't glow red immediately upon application of HV, but after a couple of minutes, yes...and how red depends a bit on how you're removing the heat from the tubes.

Sounds like you neglected a cutoff bias circuit, though.  Why let the tubes dissipate power in "standby?"  They'll sit there and get hot, and possibly generate noise in your receiver.  Silly.  Better to apply cutoff bias, which is usually done by inserting a large resistor in series with the cathode to cut the tube(s) off completely when they're not needed to amplify anything.

What kind of cooling are you using?  Are you using heat dissipator anode caps?  

BTW, the filament color is usually bright yellow, not red, for a 3-500Z...

WB2WIK/6
Logged
N2AXZ
Member

Posts: 90




Ignore
« Reply #11 on: April 10, 2006, 06:48:50 PM »

Thank you! Thank you!  Thank you!  I have been scratching my head and pulling my hair out over trying to digest all the conflicting information here and elsewhere on the web.  I agree with you that an oscillating amplifier would not have an amplitude of millivolts, and as I spent more and more time thinking about it, the more convinced of this I became.  I have made numerous, repeated measurements of plate current with one tube and both tubes, and everything falls right into place with respect  to the Eimac datasheet for the 3-500Z.  In spite of all the warnings that something must be wrong, all of the information I was gathering by probing the circuit myself convinced me that everything must be working correctly.

You are correct in that the anodes do not glow red immediately --- it takes about 2 or 3 minutes before the glow is really visible.  The anode of one tube glows somewhat more brightly than the other --- this is likely due to the slight difference in the idling currents between them.  The tubes were purchased as a matched pair from RF Parts, but measuring their idling currents individually reveals that they do differ slightly.  And yes, the filaments do glow bright yellow.

There is indeed a cutoff switch in my amplifier --- there is a 10K resistor in series with the cathode return circuit that is bypassed by a relay.  The relay is actuated through the PTT circuit.  So, no need to be concerned --- when the amplifier is sitting doing nothing, the idling plate current is indeed zero.

The tubes are forced-air cooled with a 54 CFM blower.  The blower is mounted underneath the RF deck and pressurizes the interior space of the chassis.  I am using a pair of glass Eimac chimneys, and finned aluminum anode caps.

I owe you a beer, my friend ---- thanks SO MUCH for your input!

73,

David, N2AXZ
Logged
WB2WIK
Member

Posts: 21837




Ignore
« Reply #12 on: April 11, 2006, 09:04:27 AM »

Every 3-500Z amplifier I've ever used makes the tubes glow red after 2-3 minutes of idling at 50mA per tube.  That's normal.

Hell, it's 175W of heat -- has to do something.

The tubes should get BRIGHT red under full-power key-down, and that's normal, too.  I can turn one from dull gray to cherry red in less than 10 seconds at 400mA Ip.

WB2WIK/6
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!