What is causing this audio highs attenuation/ distortion problem?

[K1JJ]

Was on 75M for a coupla hours last night. The rig ran FB and then had an arcing type crap-out.   It was the old problem of the grid input tuning. I was using thin miniductor for the coil and the plastic form melted and shorted the turns. The last time the input bandswitch melted and shorted the 4X1 grid to ground - fireworks!

I underestimated the current in the coil. The current is Q *  circuit current. If the grids were drawing 80ma, then a Q of 20 = 1.6A circulating thru the coil. It takes about 80 watts to drive the rig. That very thin miniductor is no match for 1.6A of RF.

 Today I'll wind my own coil with #16 wire and be done with it.   We all tend to underestimate the input tuning requirements and focus on the plate tank, I guess..

Other than that, Fabio II did pretty well last night. At least I haven't gotten frustrated, given up and torn it down to build another  4X1 PDM rig.  That is my normal MO .... :-)

T

[KA2DZT]

I wind my input coils on ceramic forms with double cotton wire.  I then varnish the winding,  seems to work FB.

Fred

[K1JJ]

Here's a pic of the new grid tank coil. I took out the bad mini-ductor and wound up a real maaan's coil.  This one uses #12 enamelled wire with Teflon spaghetti wound to separate the turns.  It's about 20 uH and covers 160M too.

I ran some full operating rig tests into the dummy load and it's rock solid.

However, there was actually a second problem:  The FT-102 exciter drive was moving up and down - unstable. I found a bad cap in the driver and fixed it.  Now Fabio II can drop a carrier for 10 minutes without a sweat.

It seems that whenever I get on the air, there is some new bug that gets weeded out. Eventually all the bugs will be found and it will be bullet-proof.  (I hope)    I think the grid tuning cap has some spacing problems when fully meshed, so its arcing may show up on 160M.... [sigh]



T


Covers 160 - 40M.   Underneath the sub-chassis uses all Teflon wire, of course. (of course)    That sub-chassis has had a hard life - it's housed about four rigs in the past. A coupla of Rico-Suaves and Fabio I and II.
 

[N2DTS]

Looks tight in there.
Are those crimped connections???

[KA2DZT]

They are crimp connectors.  Hopefully he soldered them.

Fred

[K1JJ]

Crimped connections?  Crimped connections?     That's when my problems would really begin.

Yes,, but I pour solder down each and every hole.   One thing about my work is it is solid and rarely fails from a bad connection. But because I change things around a lot, over time they may not look as pretty as when first built. I don't usually spend the time making thangs real pretty cuz they may be recycled at any time.  So you see the warts and all...


But I think Fabio II is a keeper.

T

The beginning is the prettiest:

[KA2DZT]

Tom

OK FB,  glad you solder the crimped connectors.  Very often I remove the plastic, crimp, solder then use shrink tubing.

Nice soldered connections really adds to the longevity of the project, which, in your case can be as much as a month or two.

Fred

[Opcom]

I also crimp them solder. I even do this when getting a car stereo installed, usually I "help" by soldering what they have crimped. Oh well it is good for someone else to crawl under and into things, and I know the shop so they only charge me about $50 more if I supervise, because it takes more time. Crimps are what they are and some last 40 years but how the industry accepts it as a 'solid' connection is beyond me. 8-)

and I see your grid coil won't be shorting out any time soon.. Learned the same lesson on an MB-40 in the grid of a 4-1000. The link wire is very inefficient above 20M and will melt the insulation, making a smell that is not easy to find.

[K1JJ]

Finally.   After many spectacular arcing crap-outs over the month, I finally found the problem.

Fabio II  ran fine out of the cabinet up to full power. But once in the cabinet, at medium power there was always fireworks in the RF deck and the HV fuses would blow. I looked many times to see any signs of carbon or arcing. I thought I found it many times, but it kept happening.

Tonight while testing into the dummy load with the cabinet on, there was a sustained flashing, five in succession - the worst arcing ever. I just knew there had to be some signs inside the cabinet. Using a magnifying glass, I found none.  An inspection of the HV RF deck showed nothing.... until I spied a copper connection on the vacuum variable with a single tiny black pit mark.  It had a melted plasma look to it, but almost invisible.  However, the cabinet did not have a matching mark. Where was the path?

Turns out that the copper had a sharp edge and the flame spread out onto the full cabinet aluminium side panel, dissipating the heat, leaving no secondary mark.  I machined the connection with a smooth edge and moved it farther away from the cabinet.  It was originally about 2" away, now 4".  RF can jump amazing distances. The 1/2" = 10KV is for DC.  Under modulation, the HV at the final plate can become X2 to X3 depending on the modulation level.  

Fixed.    With the cabinet on I fired Fabio II up at full power and really beat it hard.  No arcs.  Tried it on the air for a few Yallos. No arcing.   I expect some more problems, but this one was a toughie and may be a breakthrough to a more reliable rig.   Troubleshooting HV RF problems has always been an unnerving  and difficult job for me. Just gotta stick with it.

T

[K1JJ]

SPARK GAPS:

Today I set all four spark gaps.   [xfmr pri (2), sec (1)  Heising reactor (1) ]   Since the rig is in the cabinet when testing, I can't see the gaps arc. I put a small 1/2" piece of Scotch tape on the each gap. When it arcs, it is easy to later see which one did.  So, fire up the rig, sock some Yallos into the mic, pull the cabinet off and close the gaps of the ones with no melted tape and open the ones with melted tape.

I found that the Heising reactor gap was about twice the width as the gaps across the primary and CT. This makes sense since we are seeing 1/2 the voltage across the CT to ends of the primary.

In the end, after about eight trials, I can JUST get the gaps to arc when really hammering the audio at full power.

I know the transformer can take it cuz the gaps, when set wider, used to arc like crazy before I did spark suppression using vac relays and power resistors across the reactors.  The mod xfmr and Heising reactor are on Easy Street now.

I ran Fabio II today at full power with the cabinet on with no problems. The grid circuit and drive is very stable now. We are almost there.

Hope to be on with it this weekend for the AM rally.
  
T
 
 

[K1JJ]

Those of you who have heard Fabio II over the last few weeks probably heard the 3,500 Hz tone acoustic feedback problem I had.  If I turned the audio gain up to modulate heavily, this feedback would become very loud.

I first thought it was the mod iron talking back, but it was not.

Yesterday I covered the plate coupling cap with a towel and the feedback went away.  I replaced it with a bigger cap, but the problem remained. Hint: The towel also covered the plate tuning cap during this test.

Today I set the rig up to do a constant feedback tone and went in with a long 5' flexible plastic pipe against my ear, looking for the source.  It took about 60 seconds to find the tone was coming from the plate tuning cap C1, a nice 15KV 500 pF vacuum variable. If I touched the end cap with the hose, the noise disappeared.  The end cap is normally brazed? to the lead coming out of the glass and had loosened up. I added some reinforcing copper strap and the noise was gone.

I can now turn the audio gain up to full with just the hint of acoustical feedback.  OK for now.

The moral of the story is to suspect EVERY part, even if it is a FB vacuum variable. That was the last thing I wud have thought of. The plastic tube to the ear is a very precise and effective method to zero into the exact spot of acoustic feedback generation.  

** Be sure to test the listening tube for HV breakdown and electrical conduction before using it. For example, I tried to add some foam rubber to the end of a fiberglass rod for poking around. The foam rubber drew huge arcs off the RF areas. It acted like a carbon pencil when touched to an antenna tuner coil.   Pure fiberglass is OK as well as PVC pipes. I used a black, 3/4" PVC plastic flexible water pipe for the listening tests.

Be sure the pipe does not have any metal or carbon fiber ribbing. Testing it first is mandatory.

Tom, K1JJ

[W2VW]

Weird stuff must be contagious. I zorched through a ceramic standoff toward the end of last night's QSO. Only acted up on modulation peaks!

Symptom was reduced peak power and modulator clip lamp lighting some.

 

[w3jn]

The vacuuuuuum variable in my HB K1JJ tooner talks back also, especially on 160 where it's tooning the 75M dipole.

The tooner sits atop a tall rack and isn't that far from a drop ceiling, whose tiles are scarred with a multitude of scorch marks which are mute evidence of something busted in the antenna or feedline.

[KB2WIG]

" The moral of the story is to suspect EVERY part, "

So, you are saying to check everything out, before you throw in the towel?


klc

[K1JJ]

Weird stuff must be contagious. I zorched through a ceramic standoff toward the end of last night's QSO. Only acted up on modulation peaks!

Symptom was reduced peak power and modulator clip lamp lighting some.

 

Your audio peaks and BC sound were outstanding.  Rigs and race cars always blow when they are running at their best - most stress.   No wonder...


Johnny: Did you ever try touching the end cap of the vac variable with an insulated rod when the feedback is happening?  I wonder how common this end cap loosening problem is.


KLC:  You and 51 watt Fred get the coveted Corny Comedy Writing award on this BB....  


Update:  Fabio II made it thru a very lively three hour 75M and 160M session last night with no mishaps.  I can't believe it's not butter.


T

[W2VW]

Your audio peaks and BC sound were outstanding.


T

Thanks! Engineering credit goes to Stu AB2EZ. It's the modulator he used on the GS-35B rig.

All I did was make a change to the ratio to match my xmitter.

[K1JJ]

Thanks! Engineering credit goes to Stu AB2EZ. It's the modulator he used on the GS-35B rig.
All I did was make a change to the ratio to match my xmitter.

Yo -

Please describe your lash-up in mo detail.  I'm axing for mo details, chump.

T

[KA2DZT]

KLC:  You and 51 watt Fred get the coveted Corny Comedy Writing award on this BB....   


T

Thanks Tom,

I needed an award for something, anything.  I'm practicing for my stage comedy act when I come back in my next life.

Fred

[W2NBC]

Regarding the GS-35B modulator:

http://amfone.net/Amforum/index.php?topic=13307.50

indeed 

[w3jn]

Johnny: Did you ever try touching the end cap of the vac variable with an insulated rod when the feedback is happening?  I wonder how common this end cap loosening problem is.




T

No sir, never did.  Never got feedback but you could certainly hear midrangey audio from the tooner on voice peaks.

[K1JJ]

A big group on 75M tonight.

I think Fabio II is starting to get dialed in.   Jeff/ W2NBC made this recording from NJ using his R-390 receiver. About a 150 mile path. Running medium power tap.  Audio at -97% negative and 140% positive peaks.

The first station is W1LLY and the second station is Fabio II.

I like what I hear.

T

Listen:

[W3GMS]

Tom,

I heard you last night on 75M and the rig sounded excellent.  All of that work certainly has paid off.  When I first tuned in it took me a second or so to know it was you!  All your rigs over the years have sounded very good but this one was different.  So congrats and I have to ask, whats the next one going to be?

Joe, GMS

[K1JJ]

Hi Joe,

Thanks much for the comments.  That is the last test - hearing reports from  Bros on the air who have good ears. I'm really happy with the latest results.

The difference with this rig is I went to the ends of the Earth to optimize every detail the best I could. I also got some good email advice from Jeff/ W2NBC, Frank / GFZ, Chuck / K1KW, Bruce / W2XR , and some on-air help from others like Tron.  

This time I didn't give up when something blew up or the arcing continued for months. The high frequency sweep top end was stalled out at 4-5 KHz for weeks until the break-thrus came using the smaller bypass / coupling caps.  In addition, using the GFZ solid state audio driver was key for transparency - as well as testing and tweaking exhaustively the THD and IMD  -  using NFB, regulated supplies and any other trick I could find to make it clean.

In the past I never really had a negative and positive peak limiter like the CRL - PMC - 300A. This is a nice box. And I think using two tubes in the final, but backing it off to 1/2 power, running oversized modulators in AB1, etc., all added up.  

Jeff/ NBC gave good advice when he said not to give up, but continue to refine - refine - refine.  That is where it's at when it comes to ironing out the finer details.

Of course there is more work to do and it will probably never end. I'm now looking at a faster attack compressor to control overshoot of the highs.  I also had some breath pops recently that I think were corrected by using a double wind screen and staying away from the mike by 2".  

Building and getting a big plate modulated rig running pristine, in my opinion, is a very difficult venture.  As Bruce /XR said, this is archaic technology.  There are  many parts that can create problems, phase shifts, mismatches, etc.  The testing and optimization is possibly the most difficult part of the project.

Hope to catch ya on and hear your FB rig soon,  Joe!

T

[W3GMS]

Tom,
Just listening to it, I think you will have a hard time improving on the on the air sound.  The way you have it EQ'ed it sounds exactly like you do in person.  Its a very transparent sound.  I was listening on an 8 kHz bandwidth with the audio pickup off the detector driving a 100W amp into a studio monitor. 

I use an Inovonics 222 look ahead type of negative peak limiter.  I like it, but I am not sure I like the sound as much as the CRL which your using.  I have that CRL unit and its always worked very well but has some crazy  intermittency that I have not spent the time to find.  So I called Jim up at Inovonics and got the 222 but am not convinced that its as transparent as the CRL.  For average gain riding compression, I like the Symetrix 528E.  It won't do high speed limiting, but its good to increase the modulation density a bit.   

Joe, GMS

[N2DTS]

Sounds like you have the bugs worked out.

I think the 300a is a fantastic box. Thanks for pointing out you could get one for a good price.
Someone sent me a recording of me on 40 meters today, and I think I sound good, not very smart, but the audio sounds good.
Same deal, pair of 4x150a's running very light at 300 watts out ,380 watts in, 40 watts plate dis per tube, and modulated by a pair of 4x150a's at reduced voltage and power in AB1.

Tends to be a kind of large transmitter for the power, like broadcast rigs.

I suppose its fun to go the other way as well, say a 300 watt transmitter made up of a pair of water cooled 6aq5's modulated by a pair at 3000 volts.

In the past I never really had a negative and positive peak limiter like the CRL - PMC - 300A. This is a nice box. And I think using two tubes in the final, but backing it off to 1/2 power, running oversized modulators in AB1, etc., all added up.  


T