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 Next   Go Down
  Print  
Author Topic: 3-500 Filament Problem with Henry 2K Classic X  (Read 24624 times)
W8KR
Member

Posts: 28




Ignore
« Reply #15 on: February 24, 2014, 08:03:39 PM »

Thanks W8JI... So I take it you would suggest picking up a new 5 volt transformer and rewiring the circuit.  I'm inclined to do that.  Hopefully, someone has them.

In the meantime, at W1QJ's suggestion, I began looking at all connections, starting with the transformer, and I immediately found a finger-tight connection on one of the secondary posts, which sits just inches below the very high-powered blower, so it would be picking up some vibration.  That very well could be the problem.  It is also in a position where it is impossible to tighten, without taking out the transformer, tightening, and putting it back (if the wires are long enough).  It may never have been properly tightened.  This is an example of bad design; I could never have done something like that when I built my 4-1000.

Interesting about the sockets. I noticed last night that one of the filament socket connectors looked wider than the other one.  I tried to pull it together with a pliers but it didn't change much, if any.  I will look to see if that's the socket where the pin solder started to melt.  Do you think the tube will be OK?

Please let me know what you think.

Jack W8KR
Logged
G3RZP
Member

Posts: 1278




Ignore
« Reply #16 on: February 24, 2014, 11:56:31 PM »

Tom,

It was a military tx. Effectively, each tube had its own 5volt 15 amp winding with a common point and the transformer wound as 5-0-5, rather than 0-5, 0-5 with centre taps, which would have meant a bigger transformer. Similarly, a 2.5-0-2.5 would have meant much heavier wiring to the tube sockets.

Then the military decided that there would be a big improvement in performance from going from 350 watts with a pair of 4-125s to about 700 watts with a pair of 4-250s. So they fitted bigger transformers with no increase in the space available and didn't cool the tube pins effectively - that way, the 'design' was done very quickly, so pleasing both company and military management!

They didn't uprate the big ceramic tuning capacitors either, which was another source of reliability problems.....
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #17 on: February 25, 2014, 12:32:36 AM »

Thanks W8JI... So I take it you would suggest picking up a new 5 volt transformer and rewiring the circuit.  I'm inclined to do that.  Hopefully, someone has them.

In the meantime, at W1QJ's suggestion, I began looking at all connections, starting with the transformer, and I immediately found a finger-tight connection on one of the secondary posts, which sits just inches below the very high-powered blower, so it would be picking up some vibration.  That very well could be the problem.  It is also in a position where it is impossible to tighten, without taking out the transformer, tightening, and putting it back (if the wires are long enough).  It may never have been properly tightened.  This is an example of bad design; I could never have done something like that when I built my 4-1000.

Interesting about the sockets. I noticed last night that one of the filament socket connectors looked wider than the other one.  I tried to pull it together with a pliers but it didn't change much, if any.  I will look to see if that's the socket where the pin solder started to melt.  Do you think the tube will be OK?

Please let me know what you think.

Jack W8KR


The tube should be OK. Just be careful soldering.

Pin solder used to be a higher temperature harder alloy than regular 60/40 alloy.  I'm not sure what they are using now, probably clay with a little silver paint from some suppliers. Smiley

Buy an AL82 filament transformer. You should be able to find some used. They are planned to put 5.0 at the pins with a normal filament choke.
Logged
NJ1K
Member

Posts: 3




Ignore
« Reply #18 on: February 25, 2014, 07:01:19 AM »

I recently went through this same exercise with my Henry 3KA.  While I am not sure how close the design of the classic X is to the 3KA, here is what I found.

The 3KA does not use air system sockets, rather it has the square ceramic sockets, sometimes called Johnson sockets.  On the tube in question, one filament pin was pitted and I found matching pitting on the socket.  This first thing I did was removed the pitting by filing, sanding, scraping, etc. from both the pin and socket.  I re-flowed solder into the pin.  I removed all the socket springs and squeezed them together for better tension.  Then I swapped the tubes in the sockets.  This got me going again but I had a feeling that cooling air was possibly insufficient. 

Looking over the airflow design carefully, it looked like there was possible issues in component placement that could cause unequal flow between the two tubes.  The filament choke is placed right under one of the sockets, the entering air was offset between the tubes, and one chimney is shorter than the other choking airflow on that tube. 

I took air pressure readings under the chassis and above the chassis and compared that with the performance curve of the fan and with Eimac's datasheet.  I felt that the airflow was a little insufficient for full output and duty cycle, not to mention unequal between tubes. 

I wanted to replace the Johnson sockets with air system sockets but they wouldn’t fit without a complete redesign of the RF deck but that would require a complete custom fabricated chassis and I didn’t want to go to that length.

So, here is my current solution: I removed the blower.  I bought a new larger blower and built a box to mount it in.  This “remote air handler” sports a 2-inch high efficiency pleated filter, adjustment air vane and a 3-inch PVC outlet.  In the amp I piped the air inlet out of the back with 2-inch PVC pipe to a 2x3 reducer.  I wired a 220-volt outlet on the back of the amp for the fan to plug into.  The remote air handler sits outside on the deck and is piped into the amp with regular dryer hose and clamps. 

I also built an instrumentation panel that has a bird wattmeter, a magnehelic gage measuring deck air pressure and a thermocouple thermometer measuring outlet air temp.  Also included is an air pressure switch that alarms if the deck pressure goes below .18”wc.  I have yet to replace the short chimney but that’s on the to do list.

So far, this setup has worked FB.  I have a couple of refinements to do to it yet, but it’s on the air and working.

P.S. Tom W8JI, I ALWAYS trust your opinion.  I have a lot of respect for you..

Tom NJ1K

Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #19 on: February 25, 2014, 07:24:31 AM »

I've made hundreds of airflow tests.

The regular sockets are fine. They just need recessed enough for air to get around them.

You cannot use water column just at the plenum (chassis bottom), you have to use differential pressure, and even that does not provide the real answer. 

To fix the air system, you have to bias the tubes into the planned dissipation and use temperature measuring paint or crayons at the seals. The seals are far more critical than the glass envelope. If you get the base seals in line, the pins will follow.

Systems I have measured similar to the Henry require almost twice the blower capacity the Henry has. The base seal cooling is far more dependent on hole shape and socket recess than minor obstructions like chokes. This of course assumes air is not turbulent in the lower plenum.

I'm sort of "discussing" this with another fellow in another thread about "one 3-500Z or two".
Logged
NJ1K
Member

Posts: 3




Ignore
« Reply #20 on: February 25, 2014, 07:45:26 AM »

Yes Tom, you are correct about airflow, however, not having a real lab to do critical testing, I opted for the brute force approach.  I would not advise everyone or anyone to do what I have done, but the two advantages achieved were elemination of mechanical noise as well as about twice the mass airflow. 

True that I don't have the resources to measure actual seal temp, I think I have a workable solution now.  I ran the amp at full legal limit for 24 hours SSB during Field Day 2013 with nary a hickup.  Operating temp (outlet temp) stabliized at 135ºF with 95º+ ODT maintaining .22"wc under the chassis. 

I'll read into your other thread...
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #21 on: February 25, 2014, 08:58:26 AM »

Yes Tom, you are correct about airflow, however, not having a real lab to do critical testing, I opted for the brute force approach.  I would not advise everyone or anyone to do what I have done, but the two advantages achieved were elemination of mechanical noise as well as about twice the mass airflow. 

True that I don't have the resources to measure actual seal temp, I think I have a workable solution now.  I ran the amp at full legal limit for 24 hours SSB during Field Day 2013 with nary a hickup.  Operating temp (outlet temp) stabliized at 135ºF with 95º+ ODT maintaining .22"wc under the chassis. 

I'll read into your other thread...

Outlet temp does not indicate seal temp.

Get some temperature sensitive crayons or paint for ~20% below seal ratings, and at seal rating. They will tell you how the cooling really works.
Logged
W8KR
Member

Posts: 28




Ignore
« Reply #22 on: February 26, 2014, 02:19:51 PM »

G3RZP and W1QJ... OK, you are both right.  I found a loose connection on one of the filament transformer secondary posts and an even looser one on one of the 240V primary posts.  I haven't cleaned and tightened them yet because of the way this darned thing is built in layers Sad I will have to loosen the transformer and tip it back to get at the secondary posts.  Not impossible, just aggravating.

But here's my question: Why would these loose connections not affect both tubes?  The problem is only on one socket and that would seem to point to a bad socket, as W8JI and W1QJ suggested.

I am seriously contemplating a full re-build of the RF section to eliminate the series connected filaments, but that presents another rat's nest problem because they again built in layers, putting the input coils and switch right over the tube sockets.  If they had built this thing a couple of inches wider it would be a lot more serviceable.  For example, just to change out the suspect tube socket, I would have to remove the entire tuned input circuit, coils and all.  Grrr.

By the way, what is the opinion out there about the need for a tuned-input circuit in a GG amp?  My original 4-1000 amp didn't have one and it worked beautifully.

Jack W8KR
« Last Edit: February 26, 2014, 02:37:33 PM by W8KR » Logged
W1QJ
Member

Posts: 2966




Ignore
« Reply #23 on: February 26, 2014, 03:47:21 PM »

Jack, Don't even think about removing the tuned input circuit.  don't want to go into that because that is another full thread topic.  Do me a favor,  repair the loose connections and get the amp running and use it.  The design may not be the best, but it works.  Do what is necessary and don't open a can of worms.  Lou
Logged
W8KR
Member

Posts: 28




Ignore
« Reply #24 on: February 26, 2014, 04:40:33 PM »

Thanks Lou; I'm listening.  What about my question regarding the problem on one tube only? 

Jack W8KR
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #25 on: February 26, 2014, 06:02:22 PM »

G3RZP and W1QJ... OK, you are both right.  I found a loose connection on one of the filament transformer secondary posts and an even looser one on one of the 240V primary posts.  I haven't cleaned and tightened them yet because of the way this darned thing is built in layers Sad I will have to loosen the transformer and tip it back to get at the secondary posts.  Not impossible, just aggravating.

But here's my question: Why would these loose connections not affect both tubes?  The problem is only on one socket and that would seem to point to a bad socket, as W8JI and W1QJ suggested.

I suggested anything in the filament circuit, not just the socket. The socket only is pointed to by a pin unsoldering, but sometimes tubes are made that way. I don't know if a pin clip caused that, but I would check them for tension.

The jumping voltage can be any connection in the Henry filament secondary anywhere. This is because the filaments are in series, and depend on an equalizing tap to make voltages equal in the tubes. 

You get a loose connection at one end of the filament winding, and the tubes go way out of balance. One can go dark while the other stays lit.

You get a bad connection on the center, and the amp can eventually blow a filament out on turn-on surges.

I avoid high current filament connections through plugs and multiple mechanical connections to a remote transformer. 

I especially never series connect filaments on expensive directly heated tubes. It's silly for a number of reasons. One end of the filament in one tube is alternating from 7 volts plus bias to bias minus seven volts, while the other opposite end is doing the exact opposite.

Imagine what the electrons around the filament are doing, since operating grid-filament voltage is only ~100 volts peak with 100W drive and there is 14 volts of 60 Hz AC ripple.

73 Tom



 
Logged
W8KR
Member

Posts: 28




Ignore
« Reply #26 on: February 26, 2014, 06:15:04 PM »

Thanks Tom... Very interested in your response to questions I posed to u earlier today. I'm listening and weighing all options relative to my capabilities. I appreciate everything very much. 73,

Jack
Logged
W8KR
Member

Posts: 28




Ignore
« Reply #27 on: February 26, 2014, 08:31:39 PM »

W8JI and W1QJ... I put a micrometer on the 3-500 pins and they are 3/16.  I tried Tom's trick of putting the chuck-end of a 3/16 drill in the suspect socket's filament clips and found that one of the filament clips is definitely wider than the other, making "loose" contact on the drill.  Visually, you can see that clip spreads out more than the others, when there's nothing in it. (I had also noticed this visually several days ago.)

So, to prove my case, I put the tube, which had melted solder on a filament pin, into the socket and, sure enough, that's the pin that goes into the loose clip!

The combination of the loose primary and secondary wires on the transformer, plus the wider than normal filament socket clip, probably defines the problem.

Squeezing the clip with long-nose pliers doesn't seem to help.  Wondering if a little more brute force might work.  Otherwise, I will have to replace the socket, which requires removing the tuned input circuit Angry
Logged
N7BMW
Member

Posts: 215




Ignore
« Reply #28 on: February 26, 2014, 10:46:48 PM »

W8KR - You can remove the spring clip from the socket, bend the the two sides of the spring towards the middle and then put it back on.  I did this with an SB-220.  I admit I was worried that I would break the spring but it didn't happen. 

In theory you could take the smallest size bulldog paperclip and cut it down to use as a substitute.  That was my plan B but didn't have to do it.
Logged
W1QJ
Member

Posts: 2966




Ignore
« Reply #29 on: February 27, 2014, 03:32:26 AM »

would you be able to just replace the pin?  If you have room to work on just the pin you should be able to grind the rivet, remove the pin and install a good pin by using the right sized brass screw and associated hardware.  I have some good pins you could have one.
Logged
Pages: Prev 1 [2] 3 4 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!