Amplifier troubleshooting EDIT: Problem found.

[ARCHIVED THREAD] - Amplifier troubleshooting EDIT: Problem found.

Posted: 8/16/2013 6:44:22 PM EDT

While tuning up my amp today, encountered some issues. Everything tuned up fine. Then, all of a sudden, I lost the signal, and SWR went off the charts. I may have encountered a problem with my feedline, antenna, etc. Switched antennas, and was able to tune up just fine. However, I noticed one of the 4 tubes in my 811-H is no longer working. I popped the cover, switched positions, and it is the position that is dead, not the tube.

Green arrow is slot that does not glow.

Blue arrows point to 51 ohm resistors, all are good.

Red is a single 1N4007 diode. It conducts in both directions, and is toast. However, that doesn't seem to explain why one socket would be dead. I hear it's common for the 51 ohm resistors, and 4 other diodes to go. I am unable to locate the other diodes, but haven't broken down the amp any further.

No obvious signs of smoking. All tubes look in good shape....no melting or signs of overheating at this point.

Posted: 8/16/2013 6:55:26 PM EDT

[#1]

It's lost heater voltage if it doesn't glow. Heater inputs are pins 1 and 4 on those tubes at 6.3v. Check that socket for loose wire/solder on pins 1 and 4.
Off to find schematic..
ETA -- Looks pretty straight forward. The diode and resistors aren't in the heater circuit at all. Looks like the heater current comes direct from the transformer, through some chokes, then on to the tubes. Flop it over and get out the "Grandpa glasses".

Posted: 8/16/2013 7:25:52 PM EDT

[#2]

Yep ^^

Schematic shows 2 1N4007's, D2 and D16. D2 has a path to ground via T1 and DL1-2 (meter lamp), so it would show conductive to a degree when measuring in-circuit.

D16 cathode doesn't have a resistive path to ground so it can be measured in-circuit. Neither of these diodes isolate to a single tube though D16 can fail due to tube arcs.

Your tubes have a common filament supply (at least on the schematic I'm looking at) so with 3 of them lit it's not an issue with a discrete component in that circuit, but you should take a close look at the tube socket itself for a burnt filament connector pin.

Grids for each tube are separate circuits, R19-22 (the 51ohm resistors) connected in parallel with caps to ground. That little board in the middle of the tubes is the plate circuit for each. Try not to touch that while the unit is powered. :) Better yet, don't even power that puppy up with the cover off.

Posted: 8/17/2013 1:35:31 PM EDT

[#3]

Fixed!

Problem? One of the pin tensioners/clips was broken.

Broken clip was on red circled pin. I took the clip from the blue pin, and put it on the red one. Nothing is connected to the blue pin, so I figured connectivity there was less important.

Broken clip

four happy tubes

Posted: 8/17/2013 1:38:55 PM EDT

[#4]

Good Fix!

Posted: 8/17/2013 2:57:33 PM EDT

[#5]

D16 is a common problem part on these amps, too.

But you found the problem... good! An easy thing to miss.

Be very careful after opening, even if unplugged, as the high voltage capacitors can retain enough charge to kill you. There are bleeder resistors, but it is good to make sure. A very simple tool is made from an alligator clip, piece of wire, and a probe of some sort. The probe is fastened to a plastic stick of some sort, 1/4 nylon rod, handle of an old toothbrush, anything. You clip the alligator clip to the chassis, then touch the probe, using the plastic stick for manipulating, to the various points on the little board up by the top of the tubes. Make sure everything is safe before you begin pulling tubes.

Or ZAPPPPPPPPP!!!!! And your wife collects on life insurance. Hopefully.

Posted: 8/17/2013 3:00:29 PM EDT

[#6]

Quote History

Quoted:
D16 is a common problem part on these amps, too.

But you found the problem... good! An easy thing to miss.

Be very careful after opening, even if unplugged, as the high voltage capacitors can retain enough charge to kill you. There are bleeder resistors, but it is good to make sure. A very simple tool is made from an alligator clip, piece of wire, and a probe of some sort. The probe is fastened to a plastic stick of some sort, 1/4 nylon rod, handle of an old toothbrush, anything. You clip the alligator clip to the chassis, then touch the probe, using the plastic stick for manipulating, to the various points on the little board up by the top of the tubes. Make sure everything is safe before you begin pulling tubes.

Or ZAPPPPPPPPP!!!!! And your wife collects on life insurance. Hopefully.

View Quote

Always. I switch the meter to HV reading, and it will show if a charge is present. After it hits zero, I use an insulated needle nose to short out the top of the tubes to the chassis.

Posted: 8/17/2013 3:05:23 PM EDT

[#7]

I noticed that there is a direct ground mod out there, that has not been done to this unit. May look into doing so.

However, I found the problem that started this whole ordeal: My Alpha Delta switch!

Center position is the radio. Far right is my G5RV position, far left is my A99. As you can tell in the photo, when set to the far right, it also contacts the far left position....which is grounded out when not actively selected. Apparently....over time.... and heating.... it's developed a memory of sorts. 99% of the time, when in use, it's in the far right position. When I switch to the far left position, it in no way is near touching on the right. So... i took it apart, straightened out that center rocker, and we should be good to go!

Posted: 8/17/2013 3:19:37 PM EDT

[#8]

Quote History

Quoted:

I noticed that there is a direct ground mod out there, that has not been done to this unit. May look into doing so.

However, I found the problem that started this whole ordeal: My Alpha Delta switch!

http://i42.tinypic.com/11gjc5t.jpg

Center position is the radio. Far right is my G5RV position, far left is my A99. As you can tell in the photo, when set to the far right, it also contacts the far left position....which is grounded out when not actively selected. Apparently....over time.... and heating.... it's developed a memory of sorts. 99% of the time, when in use, it's in the far right position. When I switch to the far left position, it in no way is near touching on the right. So... i took it apart, straightened out that center rocker, and we should be good to go!

View Quote

That has got to be one of the most screwed up switch designs I've ever seen. Might be a good idea to glue in some plastic standoffs/braces that keep the ends of that curved common from twisting/tilting.

Posted: 8/18/2013 3:45:09 PM EDT

[#9]

Well.... Problem NOT solved.

I think I may have had several issues that compounded here. Still having problems when I go QRO. Around 250 watts out, SWR starts to intermittantly spike. A99 has no problems, so that tells me it's nothing in the shack. Even if the switch is taken out of the equation, still have the problem. Just checked the coax and windowed line. Found a small nick in the windowed line, but taped it up. No other issues found, no shorts.

Posted: 8/18/2013 5:50:23 PM EDT

[#10]

Are you running through a tuner? What happens if you go direct to the rig (minus amp)? What antennas are you running?

Posted: 8/18/2013 8:22:39 PM EDT

[#11]

If you are running a poor quality balun, or one that is not rated for the power you are using, or forcing a tune with a tuner for a band the antenna is not suited for, then the balun can overheat, the core saturate, and SWR go screwy.

Posted: 8/19/2013 4:35:40 AM EDT

[#12]

Technically, there is no balun in this system.

I'm running a 160M G5RV. Windowed ladder line runs to a PVC box that technically converts from balanced to unbalanced, but there are no chokes or cores. From there, 75 ohm LMR-400 into the shack. I pulled the coax off the pvc tube, and it's bright and clean as new. No signs of water penetration.

Posted: 8/19/2013 4:38:07 AM EDT

[#13]

Quote History

Quoted:
Are you running through a tuner? What happens if you go direct to the rig (minus amp)? What antennas are you running?

View Quote

160M G5RV and a 10 Meter A99. I'm going from the radio, to the amp (AL-811H), to the power meter (Daiwa CN-801HP), to the tuner (Dentron Super Super tuner... rated to 3kw) to the antenna switch (Alpha Delta).

No problems at all when the switch is set to the A99. If I connect the tuner to the A99, no problems. But any time I use the G5RV, it gets twitchy around 250W out.

Posted: 8/19/2013 5:56:10 AM EDT

[#14]

Isn't there an impedence matching "balun" in the PVC box?

Posted: 8/19/2013 6:36:46 AM EDT

[#15]

Quote History

Quoted:
Isn't there an impedence matching "balun" in the PVC box?

View Quote

Negative. It was my understanding that G5RVs did not use such devices. It is merely a physical matching. I had to replace it once prior, when water got in.

http://www.ar15.com/forums/t_10_22/655595__ARCHIVED_THREAD____Somehow______I_managed_to_arc_a_3kw_rated_amp_with__700_watts__ETA__problem_found_.html

Posted: 8/19/2013 8:12:12 AM EDT

[#16]

I have had issues with bad coax patch cords in the past. I replaced the patch cord and my issues went away.

Posted: 8/19/2013 8:16:50 AM EDT

[#17]

Quote History

Quoted:
I have had issues with bad coax patch cords in the past. I replaced the patch cord and my issues went away.

View Quote

I had considered this as well. Few things to note:

If it were a bad patch cable, the A99 would have issues as well
patch cables are Times LMR-240 ultraflex, commercially assembled, tested, and utilize silver plated Amphenol connectors
cables are less than a year old

Entirely possible it's a feedline issue. I do have a bend that might be a bit sharp for LMR-400. Only reason I'm using LMR-400 in 75 ohm is because it was cheap. As in ~10 cents per foot, brand new, still on the spool. Earlier designs of the G5RV actually used 75 ohm feedline. Since it was so cheap, I picked up the whole spool. So replacing it won't be that difficult. If I go that route, i will likely install PL connectors at the tight bend, and install a 90 degree adapter of some sort. That, or maybe switch to a short section of ultraflex.

Down to the shack, up and across the yard to a tree where the ugly balun and connection to ladder line exists. That top elbow might be a bit sharp.

Posted: 8/21/2013 5:44:13 AM EDT

[#18]

here's a better shot of the angle. If I do replace the coax, looking at a better way to deal with that corner. What's the recommended solution? 90 degree PL connector with some sort of strain relief? switch to ultraflex or LMR-240 for that corner? Leave it as is...not that big of a deal?

Posted: 8/21/2013 8:56:53 AM EDT

[#19]

If it were me, I would make a fairly large loop to negotiate the corner. I would not use 90deg PL "elbows". They just invite more problems. Reading the archived thread on the water intrusion, I wonder if the connections inside the PVC tube are wonky.

Posted: 8/21/2013 9:02:53 AM EDT

[#20]

Quote History

Quoted:
If it were me, I would make a fairly large loop to negotiate the corner. I would not use 90deg PL "elbows". They just invite more problems. Reading the archived thread on the water intrusion, I wonder if the connections inside the PVC tube are wonky.

View Quote

that's the next item on my list. However, when I peeled back the waterproof tape, the PL and SO connectors were bright and shiny as new. No oxidation or water penetration. not saying there's nothing bad inside, but since I inverted the tube (windowed ladder on the bottom), there's much less chance of water wreaking havoc.

Still... I don't see a problem with the windowed line, doubt there's an issue with the coax, and have ruled everything out in the shack. So it's probably time to crack open the PVC tube.

Posted: 8/21/2013 9:37:28 AM EDT

[#21]

Quote History

Quoted:

that's the next item on my list. However, when I peeled back the waterproof tape, the PL and SO connectors were bright and shiny as new. No oxidation or water penetration. not saying there's nothing bad inside, but since I inverted the tube (windowed ladder on the bottom), there's much less chance of water wreaking havoc.

Still... I don't see a problem with the windowed line, doubt there's an issue with the coax, and have ruled everything out in the shack. So it's probably time to crack open the PVC tube.

View Quote View All Quotes

View All Quotes

Quote History

Quoted:

Quoted:
If it were me, I would make a fairly large loop to negotiate the corner. I would not use 90deg PL "elbows". They just invite more problems. Reading the archived thread on the water intrusion, I wonder if the connections inside the PVC tube are wonky.

that's the next item on my list. However, when I peeled back the waterproof tape, the PL and SO connectors were bright and shiny as new. No oxidation or water penetration. not saying there's nothing bad inside, but since I inverted the tube (windowed ladder on the bottom), there's much less chance of water wreaking havoc.

Still... I don't see a problem with the windowed line, doubt there's an issue with the coax, and have ruled everything out in the shack. So it's probably time to crack open the PVC tube.

Rgr. By wonky I meant maybe the connections are burnt or loose. You didn't use wire nuts like the original setup did you? I hate wire nuts and Scotch-Locks.

Posted: 8/21/2013 10:03:49 AM EDT

[#22]

Quote History

Quoted:

Rgr. By wonky I meant maybe the connections are burnt or loose. You didn't use wire nuts like the original setup did you? I hate wire nuts and Scotch-Locks.

View Quote

Soldered, then used wire nuts as insulators. Original was set up the same way. soldered, wire nuts, then crimp rings for stress reliefs. Entirely possible the crimp ring cut into the ladder line, and shorted somehow. Or, connection got burnt. Or something else... who knows. But i think that probably is the next item. Shame it can't be opened without being destroyed, but oh well. I could always try to seal with RTV or caulk, instead of PVC glue...or maybe threaded ends. hm.....

Posted: 8/21/2013 10:29:17 AM EDT

[#23]

Threaded gets my vote since it seems you need to open it up now and then.

Posted: 8/22/2013 10:29:42 PM EDT

[#24]

This is a long shot, but what's the bend radius on that LMR-400? Times Microwave says it can be installed with as little as 1" radius. I think 1" is a bit too tight, but if your bend is tighter than that, the dielectric could have kinked. I try to keep bends greater than the repeated radius which in this case would be 4". It may be that if you're somewhere in between, swaying and flexing of the coax in the wind could weaken the dielectric.

Just a thought.

A time domain reflectometer would be very handy in this situation. If you have an oscilloscope, you can make a TDR pretty easily and find mid-cable faults.

Posted: 8/23/2013 4:41:12 AM EDT

[#25]

Quote History

Quoted:
This is a long shot, but what's the bend radius on that LMR-400? Times Microwave says it can be installed with as little as 1" radius. I think 1" is a bit too tight, but if your bend is tighter than that, the dielectric could have kinked. I try to keep bends greater than the repeated radius which in this case would be 4". It may be that if you're somewhere in between, swaying and flexing of the coax in the wind could weaken the dielectric.

Just a thought.

A time domain reflectometer would be very handy in this situation. If you have an oscilloscope, you can make a TDR pretty easily and find mid-cable faults.

View Quote

Yeah.... contemplating installing a strain relief of some kind, and a different elbow arrangement of sorts. I'll pop the cap off and take a look......as soon as I get a chance

May be a week or so.

Posted: 8/27/2013 3:56:06 PM EDT

[#26]

Posted: 8/27/2013 6:43:37 PM EDT

[#27]

Well the burn in phase is complete. Now you're clear to start transmitting!

Posted: 8/27/2013 6:53:45 PM EDT

[#28]

I tried a quick fix, but still having issues. Contemplating a hot swappable device. Some sort of quick disconnect for the ladder line side. So far, I've cooked two. Would be nice to swap on the go, versus an hour or more downtime gluing, soldering, and whatnot.

Of course, I'll be ordering some quality components, and building one that's a bit more bulletproof. But still...backups are nice.

On that topic, what's the best place to buy components, without getting raped? Probably order a few PL-239s....few SO-239s.... and a few other odds/ends.

Posted: 8/27/2013 7:27:48 PM EDT

[#29]

WOW!

QRO can be a PITA.

You might try RF Parts for connectors and teflon coated wire. Considering the high current at that point, I would be hesitant to use any type of "quick disconnect" device.
How about using a small waterproof box like Wally World carries with a SO-239 connector on one end and some 1/4x20 stainless bolts at the other for the ladder line. Join the SO and bolts with some heavy copper straps. This could be servicable by just opening the lid.

Posted: 8/27/2013 7:35:05 PM EDT

[#30]

I'm often pleased with connector pricing at Tessco. I avoided them for years until I realized that if you create an online account. Then you see much better pricing. However, they do have a $50 minimum.

Also, I've been pretty pleased with W5SWL for adapters and stuff. But for high power or stuff that involves calibration, I stick with Amphenol components.

What's the toasty thing? Is that your coax-to-ladderline transition?

ETA: I wouldn't use "UHF" connectors on anything that is A) outdoors or B) high power. They do not maintain a constant 50 ohm impedance through the connector and they are not waterproof. N connectors are far superior in both respects.

Posted: 8/28/2013 4:28:18 AM EDT

[#31]

Quote History

Quoted:
What's the toasty thing? Is that your coax-to-ladderline transition?

ETA: I wouldn't use "UHF" connectors on anything that is A) outdoors or B) high power. They do not maintain a constant 50 ohm impedance through the connector and they are not waterproof. N connectors are far superior in both respects.

View Quote

Correct...coax to ladder line. I used "UHF" connectors.. cause... well... that's what it came with, that's what the amp came with, that's what the tuner came with, and that's what the radio came with. Don't get me wrong....I'm not in any way saying it's the best for the job; probably just the cheapest.

I did use fusible tape, so there was no water penetration. I do think the waterproof box is a good idea. But, rather than a wal mart waterproof box, why not just use an electrical junction box? Outdoor rated, of course. Something like this: http://www.calcentron.com/Pages/allied_moulded/allied_moulded_pvc_junction_boxes.html

The feed point, and ugly balun, are in a tree. So I can easily drill the box into the tree. They already come with flange mounts. Larger box = more heat handling. Also thinking about using standard windowed line instead of regular wire inside the box. This way, there's no possible way for the two lines to touch.

Posted: 8/28/2013 5:00:34 AM EDT

[#32]

Quote History

Quoted:
I'm often pleased with connector pricing at Tessco. I avoided them for years until I realized that if you create an online account. Then you see much better pricing. However, they do have a $50 minimum.

View Quote

I created an account, but it's still showing only the list price for some reason. Logged out, and back in. I would definitly prefer amphenol connectors for the QRO stuff...I was using some silver plated connectors from Ham City, but naturally, Amphenol wins.

Posted: 8/28/2013 12:41:19 PM EDT

[#33]

Here's the plan:

SO-239 on one end, two holes for windowed line on the other end. Insulated underground electric fence wire connecting the two. The same stuff that'll hold ~16kV (i have half a spool left.

)

If something cooks, open the access panel, replace.

Posted: 8/28/2013 2:22:27 PM EDT

[#34]

That'll work. What is the fence wire conductor made of?

Posted: 8/28/2013 3:22:33 PM EDT

[#35]

Why not lose both the PVC "balun" and the ugly "balun" and replace them with something like the Balun Designs #1115? Or the model #1171.

You'll get a ferrite-cored common mode choke designed to handle 5KW. And it will also provide a transition from coax to balanced line. Drill a couple of weep holes at the bottom of the enclosure and you're all set.

Or get rid of the coax altogether and run balanced line back to your tuner. Turn your G5RV into a ladder line-fed multiband doublet.

Posted: 8/28/2013 3:25:52 PM EDT

[#36]

Quote History

Quoted:
That'll work. What is the fence wire conductor made of?

View Quote

Good Point. Pretty sure it's galvanized of some sort, which probably won't work.

I guess with this box, I'm at least guaranteed separation, so I don't have to worry about the wires getting close enough to arc.

Posted: 8/28/2013 4:44:35 PM EDT

[#37]

Use some brass or copper strap in place of wire, no chance of touching. But like mentioned above, shouldn't there be a "real" balun there?

Posted: 8/28/2013 5:11:11 PM EDT

[#38]

Quote History

Quoted:

http://i42.tinypic.com/ek1gyu.jpg

View Quote

looks GTG!

ar-jedi

Posted: 8/28/2013 5:23:00 PM EDT

[#39]

Quote History

Quoted:
Use some brass or copper strap in place of wire, no chance of touching. But like mentioned above, shouldn't there be a "real" balun there?

View Quote

Maybe? Never tried a real one there. Might be worth a shot.

Posted: 8/28/2013 5:31:03 PM EDT

[#40]

Ok... new box in place, so far, so good. Antenna switch is the only remaining problem. When tuning with the switch, I can get up to ~400 watts no problem. Let off key, rekey, SWR off the scale. Start at low power, and work up, no problem. But as soon as I let the key up, and rekey, SWR through the roof.

Take switch out of the equation, no problem. Tune up to full power, rock out with cock out.

So all in all, looks like I had several problems. Going to look at a better switch. Any recommendations? As for the balun, probably a good idea. Will likely go with the model posted earlier, or may make my own (I already have the box and connections, just need the choke).

Posted: 8/28/2013 5:36:44 PM EDT

[#41]

Quote History

Quoted:
Here's the plan:

http://i41.tinypic.com/2yn342t.jpg

SO-239 on one end, two holes for windowed line on the other end. Insulated underground electric fence wire connecting the two. The same stuff that'll hold ~16kV (i have half a spool left.

)

If something cooks, open the access panel, replace.

View Quote

Add This and These and This

Posted: 8/29/2013 5:43:03 AM EDT

[#42]

So....been pondering the current issues.

I think you guys have me sold on the balun. I'll probably get the one posted in this thread. I have some spare ladder line that's loosely coiled (4ft loops every 2 ft). I think I am going to move the feedpoint to the house, rather than running coax across the yard to a tree. This will avoid any stress on the coax from hanging in the air, and having that 90 degree turn. Likewise, it will also take up the slack in the windowed line.

Kill a few birds with one stone, so to speak.

As for the switch....that's probably going to the low priority list. Might be able to fix it, might not be worth effing around with.

Posted: 8/29/2013 6:15:43 AM EDT

[#43]

I've had good results with Daiwa switches.

Posted: 8/29/2013 6:22:02 AM EDT

[#44]

Quote History

Quoted:
I've had good results with Daiwa switches.

View Quote

I've had good luck with my Daiwa power meter, so I'll have to check out their switch offerings. I liked the arc plug design of the AD, but quality is shit.

Posted: 8/29/2013 4:22:17 PM EDT

[#45]

Here's the latest.

Took out the ugly balun, took extra coax run across yard out. Now it's ladder line, straight to the box, then coax down to the shack.

All separation needed for the windowed line is observed. It's tied off to an old insulator post up top with para cord. Its' then turned away from the house till it hits the box.

Red shows the path the windowed line takes. It zig zags to some para cord tied to a tree, then slopes up to the center point on the top of the tower (sticking about 2 feet off the tower, to stay as far away from the metal as possible).

SWR has improved. No RFI in the shack thus far. Will probably still get the balun.. but given it's the same shape and size, will be an easy swap out.