The AM Forum
March 28, 2024, 11:47:23 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: Where should I start: Troubleshooting hum in my CW signal on a Thunderbolt  (Read 11615 times)
0 Members and 1 Guest are viewing this topic.
W1TTL
Member

Offline Offline

Posts: 61


« on: January 14, 2022, 09:01:12 AM »

Hi everyone,

I am trying to figure out where to start with tracking down what appears to be AC hum on my CW signal when using my EF Johnson Thunderbolt. Listen to the attached file from WB3AVF when he was copying me.  You can hear what I'm talking about.

I am driving the Thunderbolt with a Ranger.  To make sure it's not the Ranger with the hum, I bypassed the Thunderbolt and the Ranger's CW transmission sounds fine according to local reports.  

The Thunderbolt's chassis is grounded both to the house's electrical ground and a grounding rod outside the shack.  The Ranger is, too.

I installed new electrolytic capacitors in the Thunderbolt.  Where is a good place to start with tracking down what sounds like AC on my signal?  I do have a 'scope with a HV differential probe (it goes up to 5KV) so I am isolated from the voltages in there (as much as possible, anyway).  I've never had to look for AC ripple in a transmitter or amplifier, so I am looking for some Elmer advice.  Do I use a properly rated capacitor with my probe to see if AC is present when transmitting?  Or do I just switch my scope's input to AC to look for traces of AC?  Do I start probing at the voltage regulator tubes?  

If anyone is curious, I have the "late" model Thunderbolt and the schematic is here: https://bama.edebris.com/download/johnson/tboltlate/tboltlate.pdf

Thanks for any advice and 73,
Tony W1TTL

* New Recording 46.mp3 (363.01 KB - downloaded 126 times.)
Logged
W1NB
Member

Offline Offline

Posts: 72


« Reply #1 on: January 14, 2022, 09:41:54 PM »

I would start by sampling some of the RF and looking at a continuous carrier on your scope with the sweep rate adjusted to view lower audio frequencies. If the problem is hum it will be seen on the carrier. If you see it, determine if it’s 60 Hz or 120 Hz. If it’s 120 Hz, check the bias, screen and then plate voltages under load. If it’s 60 Hz, look at the possibility of a heater/cathode short on one of the tubes.
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #2 on: January 14, 2022, 11:25:26 PM »

Hi everyone,

I am trying to figure out where to start with tracking down what appears to be AC hum on my CW signal when using my EF Johnson Thunderbolt. Listen to the attached file from WB3AVF when he was copying me.  You can hear what I'm talking about.

I am driving the Thunderbolt with a Ranger.  To make sure it's not the Ranger with the hum, I bypassed the Thunderbolt and the Ranger's CW transmission sounds fine according to local reports.  

The Thunderbolt's chassis is grounded both to the house's electrical ground and a grounding rod outside the shack.  The Ranger is, too.

I installed new electrolytic capacitors in the Thunderbolt.  Where is a good place to start with tracking down what sounds like AC on my signal?  I do have a 'scope with a HV differential probe (it goes up to 5KV) so I am isolated from the voltages in there (as much as possible, anyway).  I've never had to look for AC ripple in a transmitter or amplifier, so I am looking for some Elmer advice.  Do I use a properly rated capacitor with my probe to see if AC is present when transmitting?  Or do I just switch my scope's input to AC to look for traces of AC?  Do I start probing at the voltage regulator tubes?  

If anyone is curious, I have the "late" model Thunderbolt and the schematic is here: https://bama.edebris.com/download/johnson/tboltlate/tboltlate.pdf

Thanks for any advice and 73,
Tony W1TTL

With a scope, check your screen supply to determine if any 120 or 60 Hz ripple exists on that supply.

Phil - AC0OB
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
KB1VWC
Contributing
Member
*
Offline Offline

Posts: 40


« Reply #3 on: January 15, 2022, 07:30:59 AM »

   It's induced 60HZ from the filaments being unbalanced.  They ground one side of the filaments, rather than having a center tap. You might be able to drive it hard enough and load it heavy enough to get rid of the 60hz. The better way is to float the filaments and either ground them through a set of resistors to ground, or add a center tapped transformer in.  My Chippewa had the same issue and I installed a 12vct transformer across the existing filaments and grounded the CT.  Obviously float the other side of the filament. It had about 10% 60hz modulation at normal power levels before the mod. I'm assuming since most ham receivers didn't have a lot of bass response, no one cared.   Good luck!


Steve
KB1VWC

Logged
W1TTL
Member

Offline Offline

Posts: 61


« Reply #4 on: January 16, 2022, 04:26:11 PM »

Gentlemen,

Thank you for the replies.  OK, looking at the screens on the 4-400 finals, I am seeing 120Hz AC at the 5mV division.  (My differential probe is set to 200x.) This both when I am transmitting and not transmitting.  Is that enough AC ripple to cause the buzzing I'm hearing on my CW signal?  The screen voltage is correct (~ 510V).  I traced the ripple back to the screen's voltage supply, where it comes from the heater of a 5U4GA rectifier (and then goes through a few voltage regulators).   Is it possible that this tube is causing the 120Hz ripple or am I barking up the wrong tree?

Thank you for any continuing advice.

73,
Tony W1TTL
Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #5 on: January 16, 2022, 06:03:51 PM »

5mv on a 510 volt supply is excellent filtering and won't be the issue.  How does the bias supply look?  The control grid is very susceptible to ripple voltage.

What band was that on?  The Johnson VFO is VERY sensitive to RF getting into it when it is operating on the same band as the output frequency (which it does on 40 meters).  Do you have any crystals you can use to test it?

Rodger WQ9E
Logged

Rodger WQ9E
W1TTL
Member

Offline Offline

Posts: 61


« Reply #6 on: January 17, 2022, 11:24:42 AM »

Hi Rodger,

Thanks for the reply.

I'm sorry I wasn't clear earlier... the ripple is actually 1VAC.  My probe is set to 200x so, at the 5mV division that should be 1V.  Is 1V of ripple still OK?  I installed all new electrolytic capacitors but I don't know how much ripple is "OK."

I just checked the bias supply for the grids on the 4-400's and the 120Hz 1VAC ripple is there, too, whether or not I am transmitting.  The bias voltage is -166VDC and, by the book, -165VDC is the nominal operating voltage.

I am using 80m.  I have a 40m crystal somewhere around here, so I can try that unless you think I should be going in a different direction.

Thanks and 73,
Tony W1TTL
Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #7 on: January 17, 2022, 12:51:27 PM »

Tony,

That percentage ripple on the screen is fine.  But what is the ripple on the bias supply?  It needs to be a low percentage.

If you have a crystal handy, give it a try to rule out RF into the VFO.

Thanks for clarifying the 200x measurement! 

Rodger WQ9E
Logged

Rodger WQ9E
W1TTL
Member

Offline Offline

Posts: 61


« Reply #8 on: January 17, 2022, 02:15:46 PM »

Hey Rodger,

OK, good.  Yes -- the ripple on the bias supply is the same: 120MHz @ 1VAC.

I popped in a crystal on 7.100 and tuned everything up.  I can actually see the ripple by varying the sweep on the 'scope when sending dits and not turning on the plate.  I have my scope connected to the grids of the 4-400's.  See attached photo.

Earlier when I was testing on 80m, I did take the tbolt out of the equation to make sure the Ranger wasn't putting hum on the signal.  The Ranger wasn't. 

I turned on the plate and transmitted while listening to myself on a remote SDR in the next state.  I can hear the buzz on the CW carrier. 

Thoughts?

73,
Tony W1TTL


* IMG_1707.jpg (348.98 KB, 4032x3024 - viewed 204 times.)
Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #9 on: January 17, 2022, 03:22:00 PM »

Quick question on the 1 volt ripple for bias, is that RMS, peak, or peak to peak?  I can't remember and couldn't find in a quick search of references what the acceptable ripple percentage is for the bias supply in a class C final but it requires lower ripple than the screen and plate supplies.  I would expect very little ripple to show up on that bias supply since the current draw is so low.

But one other quick test:  test the Thunderbolt for hum on CW running at fairly low drive and input output level compared to the amount of hum running at rated power.  If one of your HV rectifiers is weak, the ripple you measure unloaded will be far less than the ripple when it is producing rated output.

Rodger WQ9E
Logged

Rodger WQ9E
W1TTL
Member

Offline Offline

Posts: 61


« Reply #10 on: January 17, 2022, 06:59:03 PM »

Hey Rodger,

The 1V ripple on the bias (as well as the screen) is peak to peak or one full cycle.  Please see the attached photo.

Thanks for checking into what the appropriate ripple should be.  I have been doing research, too, and have been coming up dry.

I lowered the drive on the Ranger so I could lower the output of the Thunderbolt.  I am still getting a buzz on my CW signal while receiving on a remote SDR.

Thanks,
Tony W1TTL


* IMG_1709.jpg (361.77 KB, 4032x3024 - viewed 172 times.)
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #11 on: January 17, 2022, 08:28:28 PM »

You’ve found trivial amounts of ripple on the bias and screen supplies. What about the plate supply? We make AM transmitters by varying that after all ;-). I don’t recall you saying it was ripple free, but I’m looking at this on my phones tiny screen and could’ve missed it.

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
KA2PTE
KA2PTE
Member

Offline Offline

Posts: 118



WWW
« Reply #12 on: January 18, 2022, 12:19:23 AM »

Is the humm still there in the TUNE mode or just in CW mode?
Logged

==================
EF Johnson RF Tube Linears
==================
https://groups.io/g/johnson-tube-linears
W1TTL
Member

Offline Offline

Posts: 61


« Reply #13 on: January 18, 2022, 01:13:22 PM »

Ed,

I was so caught up in the screen and grid voltages, I completely forgot to consider the plates!  D'oh!

See the attached photo of my 'scope on the plate supply AFTER the filtering capacitors.  I have my probe set to x2000, and the AC waveform appears to be 3 divisions from peak to peak (the 'scope is set to .5 V / div).  So, does that mean this is 3kV of AC??  The frequency is 120Hz and this waveform is present both when transmitting and not transmitting.  I see the same waveform at the 4-400s plate caps, too.

Note: I did get the five Illinois Capacitor 80uF 450V electrolytic capacitors for the plate supply off eBay...  :/

Thoughts?

Thanks and 73,
Tony W1TTL

You’ve found trivial amounts of ripple on the bias and screen supplies. What about the plate supply? We make AM transmitters by varying that after all ;-). I don’t recall you saying it was ripple free, but I’m looking at this on my phones tiny screen and could’ve missed it.

Ed


* IMG_1715.jpg (325.24 KB, 4032x3024 - viewed 164 times.)
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #14 on: January 18, 2022, 03:30:33 PM »

See the attached photo of my 'scope on the plate supply AFTER the filtering capacitors.  I have my probe set to x2000, and the AC waveform appears to be 3 divisions from peak to peak (the 'scope is set to .5 V / div).  So, does that mean this is 3kV of AC??  

If I read that right, you have 3kV PP ripple on some DC voltage, or no filtering at all. I can't tell if it is riding on top of some DC voltage from the scope picture. But yeah, that would modulate the CW waveform...

In fact, the data sheets for many transmitting tubes have a section for self-rectified RF oscillator use -- running it on AC, putting out modulated CW. I don't know if that was actually a communications use back in the day, or if it was intended for industrial uses like RF heating/materials processing.


RE Illinois capacitors, I've used them over the years, never had issues. Sounds dumb, but are they hooked up? That is, no accidental bit of insulation between things or broken lead wires? They may have failed, as well...450v *5 is 2250v, 3kV is well past that.

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
W1TTL
Member

Offline Offline

Posts: 61


« Reply #15 on: January 18, 2022, 03:45:37 PM »

Ed,

OK.  I'll pull the filtering board (see photo) and check all the connections.  You can see where I was probing the plate supply in the picture and looking closely, that resistor at the probe looks a little toasted.  I'll check all the resistors on the board, too.  

I'm attaching another photo with the AC waveform from the plate of one of the 4-400's when transmitting a CW tone.  (the volts/division is .2, so that's around 1.5V peak to peak in the photo... x2000, of course)!  I'm also attaching a photo of the 'scope when the plates are on but I'm not transmitting CW so you can see the difference.

I'll report back after I dig around the filter board.

Thanks to everyone for their help here.

73,
Tony W1TTL


* 4-400-Plate-CW.jpg (320.03 KB, 4032x3024 - viewed 175 times.)

* Tbolt-HV-Filter.jpg (385.09 KB, 4032x3024 - viewed 191 times.)

* 4-400-Plate-At-Rest.jpg (326.32 KB, 4032x3024 - viewed 160 times.)
Logged
W1TTL
Member

Offline Offline

Posts: 61


« Reply #16 on: January 18, 2022, 05:38:11 PM »

I think I know what the problem might be now.  Sometimes it helps to take a step back.

The plate filter assembly has SIX electrolytic capacitors according to the schematic.  After looking at the photo I took of the assembly, then looked at the schematic, and then looked at the photo of the assembly in the manual, I realized I have only FIVE capacitors on there.  When I re-capped this board, there were only five capacitors to start.  I took pictures before I re-capped the board and reviewed them.  If I look close enough, I can see a mark on the phenolic board from where the missing capacitor was.   At least now I see the smoking gun.  I am missing a capacitor.

73,
Tony W1TTL
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #17 on: January 18, 2022, 07:15:35 PM »

Even six 450v caps isn’t enough if the PS is actually at 3kV idling. Six is 2700 volts. They’ll live a while but nothing like they should. Newer caps are smaller, maybe you can fit one or two more?

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #18 on: January 19, 2022, 07:16:12 AM »

If the power supply voltage is near 3KV, it sounds like the bleeder resistor is also open or disconnected.  The Tbolt uses six 5K resistors as a bleeder, the HV supply is choke input and should be ~2,350 volts no load dropping to 2,200 volts at rated input.

One of the complaints about the Thunderbolt for SSB/AM linear is the plate voltage is on the low side for  4-400A/PL-175A tubes resulting in reduced efficiency and forcing it into class AB2 for running at the old 2KW PEP legal input rating.

Rodger WQ9E
Logged

Rodger WQ9E
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #19 on: January 19, 2022, 07:46:20 AM »

Sounds like it's time for 600 or 750v DC rated caps.

They are rare, but they are available.

--Shane
WP2ASS / ex KD6VXI
Logged
W1TTL
Member

Offline Offline

Posts: 61


« Reply #20 on: January 19, 2022, 08:12:33 AM »

I was just going to say that there should only be 2.2kV - 2.35kV on the plates of the 4-440's but Rodger beat me to it! Smiley  

That 3kV sounded off to me so I went back to the manual and saw the operating voltages that Rodger mentioned.  Those six 450V capacitors should be fine as long as the correct operating voltage is going through them.

Rodger - thanks.  That is very helpful.  I will check all six 5K resistors.  I see those on the schematic: R101 through R106 and they sit on a phenolic board next to the finals.  Looks like they are 50W each.  

73 all,
Tony W1TTL
Logged
KA2PTE
KA2PTE
Member

Offline Offline

Posts: 118



WWW
« Reply #21 on: January 19, 2022, 01:44:51 PM »

Since the cap change was done and directly could cause the humm, could be one is in backwards (I do it all the time).
Also just because you buy a new cap, its not automatic anymore its to spec like it use to be. Nearly all of them are
mass produced in Chyna, and quantity for low price != quality.

Even six 450v caps isn’t enough if the PS is actually at 3kV idling. Six is 2700 volts. They’ll live a while but nothing like they should. Newer caps are smaller, maybe you can fit one or two more?

Ed
Logged

==================
EF Johnson RF Tube Linears
==================
https://groups.io/g/johnson-tube-linears
W1TTL
Member

Offline Offline

Posts: 61


« Reply #22 on: January 29, 2022, 04:55:13 PM »

Gentlemen,

Well, the Thunderbolt appears to be A-OK.  Since we have a nor'easter up here, I figured it would be a good time to spend some time troubleshooting.  I added the missing capacitor on the C101 capacitor board.  Since a couple of the 120K resistors were out of tolerance on the C101 board, I replaced them all.  After that, when transmitting, I could hear arcing.  I looked around while sending dits and I saw it... right were C19 connects to L7.  A broken wire!  Some cleaning and soldering later, the Thunderbolt is back on the air and no buzz in the CW note!

Thank you everyone for your help with trying to find the cause.

73,
Tony W1TTL
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #23 on: January 29, 2022, 08:00:50 PM »

Gentlemen,

  After that, when transmitting, I could hear arcing.  I looked around while sending dits and I saw it... right were C19 connects to L7.  A broken wire!  Some cleaning and soldering later, the Thunderbolt is back on the air and no buzz in the CW note!


Glad you found the problem, Tony -   a broken wire.

Hindsight is 20-20, of course, but the old technique of tapping the circuits around with a 4' long plastic rod would have worked here. (Plexiglas or Lexon are OK)   The old shotgun approach.  I've fixed many rigs this way, especially homebrew stuff that has intermittent parts, like tuning caps or any kind of bad splice. Splices include relay and switch contacts, etc. Any connection can be called a splice.

First tap and pull on the rig parts with all the power off to expose obvious bad connections, loose screws, etc.

If no luck, then do a hot test.  For example, what to do about a BIG rig with sneaky high voltage that arcs during operation?   I usually put a large mirror behind the rig and darken the room. This gives full viewing. A small headband light on dim works well. Then key the rig up.  Tapping the plastic rod on every part in the rig usually shows the intermittent culprit. It will also expose parts that were destined to go intermittent years in the future.  

Just be super careful and be sure to use REAL insulated material for the HV tapping rod.

* Also, a hair dryer gun to warm parts and then freeze them with cold spray works like a champ to expose marginal connections.  (and bad solid state parts too)   ALWAYS spray cold freeze with the HV power OFF and caps discharged.  The actual spray stream may conduct current and fry you thru the metal can..  Same precaution goes for heating with a hair dryer gun.  Keep the HV off as much as possible as a standard practice.


T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W1TTL
Member

Offline Offline

Posts: 61


« Reply #24 on: January 29, 2022, 08:24:42 PM »

Thanks, T!  It's always something simple, isn't it?

And thank you for the sage advice.  I will add that to my book of tricks!

And yes... when troubleshooting TTL logic I've always used the freeze spray approach... but never thought to use it for connections in vacuum tube technology!

73,
Tony W1TTL
Logged
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.068 seconds with 19 queries.