KW Flashbox. Here is the damage

[WA1GFZ]

Sounds like Tank Q is too low. Try reducing L which will increase C values of tune and load caps. This will increase Q.  Make sure you have a good ground on the tuner.  You could be generating a VHF oscillation that the tuner is seeing as  a high voltage off frequency. A clue is you can't get the tuner to reduce VSWR to 1:1.

[WQ9E]

Clark,

I could well be wrong but my suspicion is that your new rig is developing bursts of VHF energy under certain conditions.  I think there are some "old school" ways to determine if this is the problem but the only method I know involves a spectrum analyzer; hopefully someone else will chime in with another diagnostic method.  

You might try a different set of 4-400 tubes (or maybe even some 4-250 types if you have them since they would likely have a different parasitic profile because of the small anode) and see if there is any change.

Off topic but sorry I didn't see any 9 pin connectors at Peoria for your 6N2.  There was a 6N2 but no connecting cable.  

[ke7trp]

Joe suggested I put two 100puff door knobs on each side of the cap to coil. This would devide the voltage across the caps and probably stop the flash over.  Any suggestions.

C

[W2DU]

For anyone planning to HB a tuner, for example, duplicating the Johnson MB for its advantage of the inherent balun via link coupling, forget about using the more expensive dual-diff cap in the output. The Johnson engineers who designed the MB with the dual-diff cap didn't know what they were doing, because the inner cap serves no useful purpose, and its only advantage is to the mfgrs of those caps for the more money the get for them.

The Johnson manual I have says the dual-diff cap as used in their MB is an impedance divider, and as such helps perform the matching function. It is NOT an impedance divider, because the inner section of that cap is useless in the matching function, and actually performs against obtaining a normal impedance match.

Note that the inner section of the cap is connected directly across, in shunt with, the input to the feed line. The outer section forms the output cap of a balanced L network, thus the outer section is what performs the impedance match. The purpose is to match to the input impedance of the feed line, but as the outer section is tuned, the inner section shunting the line is continually changing the impedance seen by the outer section, so in effect, you're chasing your tail as you adjust the dual-diff cap.

You can prove the uselessness of that cap by disconnecting it. First, adjust the MB with the inner cap connected. Then disconnect it and observe what an insignificant change in turning the cap is required to re-establish the match.

In other words, don't waste yer moola on the dual-diff cap.

Walt

[ke7trp]

Interesting.  I think its time someone made an upgrade faq on bringing these tuners up to todays day and age.  Add 160 with a switch and a vac cap, Beef up the voltage capability and maybe the range.  THere are thousands of them out there.

Hopefully, I will figure this out. I am running the T3 at full legal limit right now on 7293 with no arc over.   Band seems dead.

C

[WA1GFZ]

If two rigs work and the third doesn't why are you concentrating on the tuner.

[ke7trp]

I am not, I spent the day testing this rig out and reported the results. 

C

[KA1ZGC]

I could well be wrong but my suspicion is that your new rig is developing bursts of VHF energy under certain conditions.  I think there are some "old school" ways to determine if this is the problem but the only method I know involves a spectrum analyzer; hopefully someone else will chime in with another diagnostic method.  

Put a neon bulb on the end of a stick. Hold the bulb close to the signal path and apply RF. It will glow a sort of amber color in the presence of HF. VHF will make it glow a more purple hue. Work your way through the stages and you'll find your parasitic, if there's one to be found.

[Steve - WB3HUZ]

he new Rig is only 21 DB down.

If this is correct, regardless of your tuner problem, you need to fix the second harmonic problem. The second harmonic should be at least 40 dB down.

[KA1ZGC]

Good point. That rig's not even legal to run with only 21dB suppression of the second harmonic.

But these harmonic measurements won't tell you if you're generating a VHF parasitic under modulation, even if you take them while it's happening. Parasitics don't necessarily manifest themselves at harmonic frequencies. They can show up anywhere, depending on what creates them.

Still, fix the harmonic, and you might clear up any parasitic by accident. A broken resistor on a parasitic suppressor can make a rig do all kinds of neat things the designers didn't intend.

It does sound like you've got a parasitic, but you've got to at least get the rig legal before worrying about that.

[W1UJR]

"Put a neon bulb on the end of a stick. Hold the bulb close to the signal path and apply RF. It will glow a sort of amber color in the presence of HF. VHF will make it glow a more purple hue. Work your way through the stages and you'll find your parasitic, if there's one to be found."


Now that's a very cool and most practical thing Tom, never heard of that before regarding color and frequnecy.
The neon bulb idea of course, but what are the physics which make the gas change color in the presence of VHF?

[KA1ZGC]

Now that's a very cool and most practical thing Tom, never heard of that before regarding color and frequency.
The neon bulb idea of course, but what are the physics which make the gas change color in the presence of VHF?

Haven't the foggiest. I could speculate about amber being a longer wavelength than purple, which it is, but I honestly don't know the true physics behind it. I only know it's an ancient old-buzzard trick from yester-century that hasn't let me down yet.

[WA1GFZ]

one of the oldest tricks in the book. OK some smart guy tell why neon glows purple when excited by VHF RF.

[KA1ZGC]

We were all waiting for you, Frank! 

[WQ9E]

ZGC, thanks for posting the neon bulb trick.  I vaguely recalled reading about it but I couldn't remember the details.

Steve,  If the second harmonic is only 21 db down that is a problem.  But I don't think Clark posted all of the details on his measurement setup and measuring the level of harmonics and spurs requires a very precise setup to get accurate results.  The biggest issue in the hamshack is that whatever you are using to sample the output must be frequency insensitive across the range of interest.   A typical ham setup is to use a small pickup wire for the spectrum analyzer and the gain is going to vary greatly by frequency so the results are meaningless unless the sampling setup is flat.   I was fortunate to cheaply acquire a sniffer setup years ago but for those who need to do these measurements there is a very practical setup in a number of the 80's or 90's era ARRL handbooks using 4 Heathkit cantennas to provide a suitable load/sampler.

[Steve - WB3HUZ]

I agree. If we assume his test set-up was consistent between the measurements made on the T-368 and his HB rig, the HB rig is 10's of dB worse than the T-368. If the T-3 hasn't been modified, we can safely assume the second harmonic output from it is probably 40 dB down or more. From this, I concluded his HB rig has a problem.

Yes, I made some assumptions and it would be good to go back and do the measurements again or use a spec an.

ZGC, thanks for posting the neon bulb trick.  I vaguely recalled reading about it but I couldn't remember the details.

Steve,  If the second harmonic is only 21 db down that is a problem.  But I don't think Clark posted all of the details on his measurement setup and measuring the level of harmonics and spurs requires a very precise setup to get accurate results.  The biggest issue in the hamshack is that whatever you are using to sample the output must be frequency insensitive across the range of interest.   A typical ham setup is to use a small pickup wire for the spectrum analyzer and the gain is going to vary greatly by frequency so the results are meaningless unless the sampling setup is flat.   I was fortunate to cheaply acquire a sniffer setup years ago but for those who need to do these measurements there is a very practical setup in a number of the 80's or 90's era ARRL handbooks using 4 Heathkit cantennas to provide a suitable load/sampler.

[Pete, WA2CWA]

one of the oldest tricks in the book. OK some smart guy tell why neon glows purple when excited by VHF RF.

Been using this technique since my Novice days. Nothing new; check the old Handbooks and the ARRL Hints and Kinks books.

Most "neon" lamps are a mixture of neon(mostly) and argon gases. Argon has a lower striking voltage then neon. So, when looking around for a VHF parasitic, the lamp generally will glow a shade of blue/purple. If the striking voltage is high enough, the lamp will start to glow red/yellow. I also suspect that the generated RF frequency also plays a part in the type of glow.

[W1UJR]

All right, excellent info from Tom and Pete.

I've used the neon bulb and incandescent bulb trick as well, in fact I use one on each side of my feed line to check balance.
But I was not aware of being able to distinguish the RF freq. by the color, Pete's explanation makes sense to me.

Did find this, 3/4 down the page is the series of photos with the glow of various gasses --->> http://en.wikipedia.org/wiki/Neon_sign

Now I've got a new trick to add to the bag, thanks!

[ke7trp]

Thanks for helping me guys.  

I got the spec AN and tested the rig again. I did not see much difference from the chart made using the icom.  

The transmitter runs into a Bird Dummy load. The Spec AN is on the table. The same coax is used for the T3. Nothing changes. The power output on both transmitters is the same during the test.

I got Robert on the phone and we experimented in loading. The trouble seems to be that when he had the RF deck, His line voltage was alot lower then mine.  I have about 300 extra volts on the B+ side.  When we changed the loading the second harmonic went down.  Its not as clean as a T3 but it wont ever be. The T3 uses an L and this is a simple Pi network.

The big trouble here is that with BC iron and no filter or EQ, I am pushing way to much audio bandwidth through the Rig. Once the Spec AN was inline, I realized that the transmitter was 35 to 40 KC wide! This goes back to the HI-FI antenna post. I tested the KW 20 KC up and its real touch on the tuning. 20 KC off set from center is a huge SWR.  So that cant be helping me here.

THe Plan today is to fix this DAP310 with NRSC filter.  Once thats in, I suspect I wont arc the tuner anymore.  I am not going to run it into the KW anymore.  I will have to just wait

Learning alot!  Thanks for the tips.

C

[WA1GFZ]

If you are modulating at 20 or 30 KHz this could be outside the bandpass of the matchbox. This will present a different load on the rig and maybe causing a high voltage to be generated in the network because it isn't going to the antenna. I bet you notice the swr change with modulation.

[W1UJR]

If you are modulating at 20 or 30 KHz this could be outside the bandpass of the matchbox. This will present a different load on the rig and maybe causing a high voltage to be generated in the network because it isn't going to the antenna. I bet you notice the swr change with modulation.

Won't the voltage be higher on voice peaks any way Frank, causing the arc-over?
When he is measuring power, I assume he's measuring an unmodulated carrier?
A few atomic hellos will push things up a bit.

[WA1GFZ]

Yes, but if it is inside the bandpass of the network it will be passed rather than be reflected. If reflected the loading cap in the rig could even flash over.
like pissing through a key hole, too wide a flow and you get splashed.

[ke7trp]

Thats what I just said 

I will let you know what happens.

C

If you are modulating at 20 or 30 KHz this could be outside the bandpass of the matchbox. This will present a different load on the rig and maybe causing a high voltage to be generated in the network because it isn't going to the antenna. I bet you notice the swr change with modulation.

[W2DU]

It was back in the late 1930s that I used neon bulbs as indicators, and yes, on low-freq HF the glow was red, but as the frequency increased the color turned toward purple. I wondered then what was happening, but never found out why.

Walt

[Todd, KA1KAQ]

Yes, but if it is inside the bandpass of the network it will be passed rather than be reflected. If reflected the loading cap in the rig could even flash over.

Which is what I was trying to get at earlier, just wasn't sure of the 'how' part. You guys are smarter than me so I knew you'd figure it out. Same thing is happening with my big rig up on 75 as soon as I hit the audio. Doesn't bother an equal distance from resonance below 3800. Only difference is I don't have the flashbox in line.