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 6 7 Next   Go Down
  Print  
Author Topic: Tubes going soft - Again  (Read 18821 times)
KH6AQ
Member

Posts: 8008




Ignore
« Reply #15 on: April 10, 2010, 06:28:52 AM »

You could compare filament current but what really drives cathode performance is cathode temperature.

Install a 'bad' tube and a known 'good' tube from another manufacturer. Take a look at the color of the filaments. Is the 'bad' tube filament much cooler or hotter?
Logged
K7ZRZ
Member

Posts: 279




Ignore
« Reply #16 on: April 10, 2010, 09:31:43 AM »

WX7G.... excellent suggestion. I can get the original Dentron tubes back from my buddy (who isn't using them currently), and compare. My observation of the two different sets of tubes has been that they glow in a very similar intensity.

I didn't note anyone raising plate dissipation as an issue,

In your first reply, you made the comment that if the anode is not glowing orange, it is not being overheated.  That (and other comments I have often read) cause me to wonder and try to determine. Also, glass envelope overheating could only be caused by excessive anode dissapation, it seems.

Tom... thank you much for your (as usual) very insiteful response. I have completed an order for another quad set of the tubes - again from RF Parts.  I got a response to an inqriry to Ameritron as to their Tube offering.. to which they indicated that they were using Taylors also, and that's what would come from them. I decided to make another purchase from RF Parts and have a new set on hand, with which to experiment and try to make a good combination of tubes. If I can find a bad tube or tubes, then I can go back to them for a warrantee again and will be in a better negotiating position.

Also, Tom, can you recommend a specific circuit or schematic that I could look at and try to copy for the Grid metering adaptation.  My first thought was that I could use the meter switch position now used for plate voltage. It seems that plate voltage is going to remain constant and not of any specific importance for normal day to day amplifier operation. A modification of that nature could always be returned to original later if selling the amp to someone to whom that was important.  Could you please also tell me what specific "teflon coated wire" I should be trying to obtain, and from where? I'm quite unfamiliar with it.

Tom... a recent suggestion you have made to me regarding the issue of filament winding center tap - for which you suggested a resistor network to "imitate" a CT.  Would that have any impact on filament voltage if implemented and how much?

All in all, I feel better about this whole situation. Hopefully I CAN find a problem with tubes - which is not related to operational errors and conditions. And with another complete set, I feel I may have improved my odds of being able to assemble a good healthy set into the amplifier.

Thanks again all.

Brian
« Last Edit: April 10, 2010, 09:35:44 AM by Brian J. Ingoldsby » Logged

Brian K7ZRZ
VK1OD
Member

Posts: 1697




Ignore
« Reply #17 on: April 10, 2010, 12:14:54 PM »


I didn't note anyone raising plate dissipation as an issue,

In your first reply, you made the comment that if the anode is not glowing orange, it is not being overheated.

Those last words were from WX7G, and IMHO, they dismiss anode dissipation as an issue rather than raising it.

To overheat the anode takes a more sustained effort than overheating the grid.

Owen
Logged
K7ZRZ
Member

Posts: 279




Ignore
« Reply #18 on: April 10, 2010, 02:03:41 PM »

Owen,

I'm sorry for the confusion. I better start checking my sources more closely before making quotes. My bad.

Brian
Logged

Brian K7ZRZ
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #19 on: April 10, 2010, 06:21:52 PM »

Brian,

The reason I suggested you measure filament current is because no one I know can read temperature with their eyeballs. If you know the filament current and voltage you KNOW the heat dissipation in the filament. If you see filament current is disproportionate in one tube, it is probably bad. Also make sure new tubes draw expected current at rated voltage, or real close. Defects in the filament will show up that way.

As for grid metering, it is necessary to float the negative rail of the supply through a grid shunt resistor. The zener for bias has to float from the chassis and return to the plate current shunt. The meter has to switch between the plate current shunt and the grid shunt. You should just copy what the Ameritron AL80 did, or the SB200 Heathkit.

The Teflon wire is used as a resistor. Just make the gauge large enough to not overheat and add enough in series to bring voltage down. #16 or 18 would probably hold up fine. It will get warm but that won't hurt the insulation.


The filament CT resistors won't help you.




Logged
K8KAS
Member

Posts: 578




Ignore
« Reply #20 on: April 11, 2010, 08:34:40 AM »

That FLICKING to 1800 watts output really has me, lets see at 60% eff. that comes out to 2700 plus watts input power. Come on guys if I was RF Parts I would throw a hammer at you.

 What is the combined grid current? I am sorry
but I lost it on you when you mentioned that FLICK to 1800 watts output.

I have owned 3 or 4 SB200, one I have now for a standby amp is using the same tubes it came with 30 years ago and I still get 525 watts out on 40 meters.

The Clipperton is not a great amp for sure but they will huff along if treated well, say 1000 watts
output on a peak reading meter for a along time.

I tune with 10 or 20 watts to get the tune and load peaked and the slowly bring up the drive to my desired output and at the same time check my total grid/plate current not to exceed number and this is for 10 or 15 seconds. 73 Denny K8KAS
Logged
VK1OD
Member

Posts: 1697




Ignore
« Reply #21 on: April 11, 2010, 02:15:20 PM »

I just browsed a schematic for the Clippertron L.

The schematic I see shows that it does not use a centre tapped filament transformer, so that filament current is not uniform. Not only is the filament voltage apparently high (presumably in the receive state) according to Brian, but cathode current contributes to additional heating in an asymmetric way. I expect that on SSB, the effects of this will be insignificant... but the long term effect on high duty cycle modes will be to raise part of the filament to higher than rated current and temperature, more so that if the anode current flowed in a symmetric way.

Does it use some kind of input matching network? The schematic I see shows a small box in series with the cathode feed, but doesn't describe it, and leaves me wondering if the input is direct to the cathodes without filtering or impedance transformation. It would be interesting to know what the input VSWR is when driven to rated output (whatever that is).

I can't make sense of the specs. It says HV is CW idle +1800, SSB idle +2700. There is no mode switching, are they trying to say the power supply sags from over 2700V to 1800V on CW key up (surely not), key down?

So I am left guessing what the HV was when Brian measured 1800W out at just less than 1A of anode current. If those measurements were correct, they suggest HV was close to 3000V loaded. Something is wrong here!

Owen
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #22 on: April 11, 2010, 02:59:19 PM »

The schematic I see shows that it does not use a centre tapped filament transformer, so that filament current is not uniform. Not only is the filament voltage apparently high (presumably in the receive state) according to Brian, but cathode current contributes to additional heating in an asymmetric way. I expect that on SSB, the effects of this will be insignificant... but the long term effect on high duty cycle modes will be to raise part of the filament to higher than rated current and temperature, more so that if the anode current flowed in a symmetric way.

That's one of many bad things Dentron did. But the filament current is about 4 amperes per tube, and plate current only 1 ampere in total, so the filaments are not in too bad of a shape with the funny wiring. Mostly it causes hum on the signal.

Quote
Does it use some kind of input matching network? The schematic I see shows a small box in series with the cathode feed, but doesn't describe it, and leaves me wondering if the input is direct to the cathodes without filtering or impedance transformation. It would be interesting to know what the input VSWR is when driven to rated output (whatever that is).

The Clipperton was "designed", and I use that word loosely, as a 1 kW DC input amplifier. 600 watts carrier out on CW, and about 1000 watts on SSB PEP.

It originally had no tuned input, the little box was a ten/eleven  meter trap.

Quote
I can't make sense of the specs. It says HV is CW idle +1800, SSB idle +2700. There is no mode switching, are they trying to say the power supply sags from over 2700V to 1800V on CW key up (surely not), key down?

It changes the HV winding for CW and SSB.

Quote
So I am left guessing what the HV was when Brian measured 1800W out at just less than 1A of anode current. If those measurements were correct, they suggest HV was close to 3000V loaded. Something is wrong here!

I ignored the "occasional flicks" and just assumed he had a funny meter or it was bouncing to untrue readings.

As for tube life, it is a thoriated tungsten filament. Emission current is NOT a player in tube life, so loss of emission would not come from hammering the filament emission like it would with a MOx cathode. Even the grids are difficult to hurt. Damage from excessive power would mostly be damage to the anode from excessive anode temperature.

I still would never own an amp without a grid current meter.

IMO this most likely is a combination of a little extra voltage with a poor material quality or poorly built filament. I've seen quite a few short emission life tubes from the same country that copper plates steel and calls it copper wire.
Logged
VK1OD
Member

Posts: 1697




Ignore
« Reply #23 on: April 11, 2010, 03:14:41 PM »

Quote
I can't make sense of the specs. It says HV is CW idle +1800, SSB idle +2700. There is no mode switching, are they trying to say the power supply sags from over 2700V to 1800V on CW key up (surely not), key down?

It changes the HV winding for CW and SSB.


I missed that, I don't know how!

Owen
Logged
K7ZRZ
Member

Posts: 279




Ignore
« Reply #24 on: April 11, 2010, 04:03:01 PM »

That FLICKING to 1800 watts output really has me, lets see at 60% eff. that comes out to 2700 plus watts input power. Come on guys if I was RF Parts I would throw a hammer at you.  73 Denny K8KAS

I ducked and you missed. Come on guy...it wasn't intentional, it just happened. I raised the drive level and up she went. Maybe it was the needle's momentum that carried it up there. My point was that the power raised to a point far beyond what it will now. Something's different in these few months.    And secondly, RF Parts has nothing to get violent about. I just spent another 200 bucks with them for another new set.

Can I come out now??

My Clipperton DOES have a tuned input board. It was added along the way, and I don't know by whom the parts were issued. It looks just like others I've seen pictured.  The added wafer that controls it - first one behind the front panel, is brown and the originals are white.


Brian K7ZRZ
« Last Edit: April 11, 2010, 04:57:57 PM by Brian J. Ingoldsby » Logged

Brian K7ZRZ
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #25 on: April 12, 2010, 09:12:22 AM »

That FLICKING to 1800 watts output really has me, lets see at 60% eff. that comes out to 2700 plus watts input power. Come on guys if I was RF Parts I would throw a hammer at you.  73 Denny K8KAS

I ducked and you missed. Come on guy...it wasn't intentional, it just happened. I raised the drive level and up she went. Maybe it was the needle's momentum that carried it up there. My point was that the power raised to a point far beyond what it will now. Something's different in these few months.    And secondly, RF Parts has nothing to get violent about. I just spent another 200 bucks with them for another new set.

Can I come out now??

My Clipperton DOES have a tuned input board. It was added along the way, and I don't know by whom the parts were issued. It looks just like others I've seen pictured.  The added wafer that controls it - first one behind the front panel, is brown and the originals are white.


Brian K7ZRZ

Does your Clipperton have the CW/SSB switch that changes transformer taps Brian?

Secondly, let's all get over the power thing. Unless someone has a 150-200 watt exciter, they would never be able to drive the 572's hard enough to get 1500 watts out. The AL572, despite running absolute maximum voltage and a whole lot better layout and tank, will only really run 1300 watts out on the best bands with 100 watts drive.

Even if you did hammer the tubes with more drive that should never cause premature emission failure. Thoriated tungsten cathodes are not depleted by emission or harmed by removal of the electron cloud around the cathode like MOx cathode tubes. That's why they can have full load on them while bring brought up from zero volts on the filament, instead of requiring a warm-up time before plate current is drawn.

I suspect you either have a "funny meter", or you have a little SWR causing reflected power to add to forward power and inflate the reading. There is no way you can get a REAL 1500 watts out of four tubes with only a typical 100 watt exciter. It won't happen.

Now if you were slamming the grids by under-loading the amp with full drive, you could heat the grids up. Normal 572 grids can actually heat to the point of faint color without failure. Although grid emission would be terrible while hot, and it isn't good to run them that way, it is unlikely that would poison the filament. You would most likely have tank component arcing long before cooking the grids.  I would still add a grid current meter, however.

I just checked a fairly reliable source on the service history of 572's from China, and they are averaging only a little of 1 year for emission life when operated at 6.3 volts. The real Russian tubes are all code dated from 1999, and while they have good emission life about half fail high voltage tests. A large percentage of those that pass HV screening fail in the field from arcing.

I hate to tell you this Brian, but I think you are fighting a typical Chinese quality control or manufacturing problem. I don't thing you are doing a thing to cause the problems, although your 6.6 volts might be accelerating the emission problems. 572's might not be worth the trouble until manufacturing straightens out.

I suspect as much from the start because it is nearly impossible for you to do anything that would make the emission go away so fast, other than running 20%-30% too much filament voltage.

Did you isolate the bad tube or tubes? You might try running 8.5 volts or so for just a few seconds, and then immediately step down to 7 volts and bake them over night at 7 volts.  Then see what they are like.

My bet is they have contaminated materials, or the wrong mix of materials, in the filament.
Logged
K7ZRZ
Member

Posts: 279




Ignore
« Reply #26 on: April 12, 2010, 03:20:42 PM »

Hi, Tom,

Yes, I certainly do have the switch which switches voltage from about 1900 V. to about 2700 V.

As I mentioned, I have ordered 4 more tubes from RF Parts, and am hoping there to be less than the full 4 that are in it now to be bad. I theorize that with enough tubes on hand, I have a better chance of actually possessing and being able to install 4 good ones... and be done with it.

When the tubes get here, I set about trying to isolate a bad tube or tubes and then deal with RF Parts for another warranty.

To elaborate a bit on the previous thing I had mentioned about raising the power level quickly when the tubes were first installed.... I tuned the amp at full rated power first advanced the load control some, then cranked up the drive on the transceiver (only capable of supplying 100 watts drive, period) - and when the plate current meter got up to just under an amp, the fuse on the back of the amp blew. (It was the proper value fuse - and the supply voltage was 120VAC.)  Again, I didn't expect to have that happen, and I certainly didn't maintain that level of drive nor have I pushed the amp past any reasonable level of operation.  But boy, those tubes really drew the current. I have the amp Load control considerably advanced, and leave it there for most tunning operations. Also, doing the extra "load" increase is part of my regular tuning process, and I always give it enough advance to pull the output power down by at least 100 watts from max.

The raising of filament voltage for a "shock" (I'd guess) and cooking them at 7 overnight is an interesting suggestion. But not one that I would have the ability to execute here in my motorhome at the park.  I think I'll be doing tube substitution and evaluation before I'd have the ability to do something like that. I'll keep it in mind.

Thanks for the guidance on all this. I not afraid to take my medicine when I pull a boo boo. But one thing that comes to mind is... what's happening to all the amplifiers being currently manufactured and sold (Ameritron AL-572) that are getting the same tube in them that I'm getting??  Or am I getting something less likely to hold up - from RF Parts - than Ameritron is?  Hmmm.

Brian K7ZRZ

Logged

Brian K7ZRZ
K7ZSA
Member

Posts: 12




Ignore
« Reply #27 on: April 12, 2010, 08:13:15 PM »

I have been following this thread closely as I have the same scenario here.  I have the Clipperton L now equipped with a new quad.  I have seen the infamous "flick" of Ip and I agree with Tom in that I suspect a reflection due to SWR.  That is why I pre-tune my antenna / tuner with an MFJ analyzer that I switch in.  This low level pre-tuning is non interferring and makes life easier for the transceiver and amp.  Get over the flick thing.  My money is on one or more bad tubes and I don't envy Brian in the task of rotating them around and repeatedly testing.  Perhaps if the Chinese did a little more testing we could just get on with Hamming.
As far as RF Parts goes, even bad publicity is publicity.
Jeff
KZSA
Logged

It all started with a spark!
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #28 on: April 12, 2010, 11:15:47 PM »

Thanks for the guidance on all this. I not afraid to take my medicine when I pull a boo boo. But one thing that comes to mind is... what's happening to all the amplifiers being currently manufactured and sold (Ameritron AL-572) that are getting the same tube in them that I'm getting??  Or am I getting something less likely to hold up - from RF Parts - than Ameritron is?  Hmmm.

Brian K7ZRZ



Everyone suffers the problems of poor or questionable 572B quality. It has been an issue from day one for the AL572 amp.

When the AL572 was released we used Svetlana 572's. Over half the brand new tubes were bad and would not handle 4000-5000 volts on peaks without arcing. Since the peak plate voltage goes nearly twice the dc anode resting voltage when properly tuned, and can be several times the dc anode resting voltage when NOT tuned properly, and since there were four tubes in parallel, field failures were terrible. 

Apparently the Chinese manufactured 572's now have similar filament emission life issues. It doesn't seem like it is anything you are doing.

Tom
Logged
KC8VWM
Member

Posts: 3190




Ignore
« Reply #29 on: April 13, 2010, 12:02:24 AM »

Does anyone know if anyone in the US has ever been in contact with the Chinese manufacturing facility responsible to suggest improvements and/or to address this ongoing "tube failure" issue?

It is my understanding the Chinese have expressed they are starting to take "quality control" initiatives more seriously and are working with US partners in resolving similar issues.

I am suspecting they don't know about certain test procedures or materials they can use that would serve to improve the current flawed design they are currently churning out.

I suspect this is a matter of talking to the right people.
Logged
Pages: Prev 1 [2] 3 4 5 6 7 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!