4-1000 grid driver and MB-40 question

[AB2EZ]

Opcom

Great!

What you did is similar to what I suggested in my last post... but even easier to implement

Grounding the frame has the same effect, on the tuning range, as putting a 400pF capacitor between what I called the "center bus", and ground. I believe the frame is electrically connected to what I called the "center bus".

[Note: the problem that your solution fixed is the following:

The fixed capacitance across the MB-40's low band tuned output circuit is the grid-to-ground input capacitance of the tube. I.e. even if the variable capacitors have very low capacitance when the plates are fully open... you still have this fixed capacitance across the inductor(s).

If you temporarily ignore this fixed capacitance, then whether you use both inductors and both variable capacitors, or only one inductor and 1 variable capacitor , you can still achieve resonance with roughly the same position of the tuning knob. I.e. with two inductors in series and two variable capacitors in series, you have 2L and 0.5C so... ignoring the grid-to-cathode capacitance of the tube... [2L x 0.5C} = LC. But, when you use both inductors (in series) and both variable capacitors (in series), the grid-to-ground capacitance of the tube more strongly constrains the highest frequency at which you can achieve resonance... because you have 2L instead of L... and therefore, at any frequency, the capacitance required to achieve resonance (including the fixed grid-to- ground capacitance) has to be half as big.]

By grounding the center bus, you are now only using the bottom inductor and the bottom capacitor to form the grid-to-ground resonant circuit]


What is different (between your solution and my last suggestion) is the way the neutralization comes into play.

If you don't have any parasitic oscillations (apparently you don't), you are all set.

Best regards
Stu

[WD5JKO]

Pat,

   I been following this thread for a while. Glad your making progress on the Tucker. Looks like Stu has been active helping you. I bet Colonial Tucker is Saluting you now.

In your absence, 40M 7160 in the 5th call area has gotten sparsely occupied by the occasional AM QSO. Meanwhile the crud moved in at 7158 where guys on LSB fairly local talk all day..fall asleep, then start talking where they left off an hour prior. Finding the frequency clear, and calling a CQ on 7160 usually kicks a sleeping hornets nest. A few weeks back they were sleeping again, frequency was clear for over 30 minutes. So I call a CQ on 7157, and Darrell Wa5VGO comes back on his KW1. That was nice in that it pushed them guys down to 7155. Since then they have moved back. This is reminiscent of up north on 7290 when guys slept and chatted on that frequency all day...then came Astabula Bill W8VYZ who keyed up while they were sleeping, and 7290 was clear. There is a long story on that, and a FCC ruling in the favor of old buzzard Bill.

   Maybe we can get you on 40m this weekend?

Jim
Wd5JKO

[Opcom]

I sure have received a lot of good help from Stu and others, all of the good suggestions made me think about the whole circuit in ways I had not before.

It's hard to say yet but the stability/neutralizing might be better now because I did not find the usual messy stuff on modulation peaks and I did not seem to notice funny plate-grid business when tuning the grid around while the TX was making a 400-500W carrier.

There is a certain range of drive wattage that produces a cleaner looking sine wave of RF from the unit, and this corresponds to its apparently 'best' operation. Above or below that drive range (10-18W) and there is a notching effect of the waveform that seems like a harmonic caused by a parasitic circuit. The TX due to its BC-Engineer-Designed layout has been 'plagued' with such annoyances and I do not blame COL Tucker, but accept this as a challenge for me to resolve. It needs more investigation and may be nothing more than an oddity of grid drive level vs RF feedbacl from the neutralizing circuit, however, as long as the setup is done with a scope this does not appear and did not seem to be a problem.

The above artifact appears after the LPF that is connected to the TX output, and I should move the scope tap to the input side of the LPF so as to see more and not just hide it. Question: isn't the neutralizing circuit really just a feedback loop?

The somewhat better behavior might be due to the three main things that have changed.
1. putting a 100pF cap in series from the tank to the grid
2. using an RF choke to supply bias right to the grid
2. insulating the capacitor shaft of the high frequency section from GND

I hope tomorrow to test the other bands and see how they work. I may try to be available about 11AM to Noon CST, if Jim or anyone wants to try 7160. I'm not retired, so it will be rare occurrence.
Barring any crap outs I will be on the air Saturday at the 'usual' time some time in the 11AM to 4PM range, suggestions welcome. We must re-take the slot for AM, as today I heard two gentlemen on 7160 LSB who were AM users bemoaning the demise of AM and saying that so and so with the large TX or big signal was gone, etc..

Lastly, I have to figure out a set of dial settings for different conditions. The TX has many adjustments for the class C stage and the best pi-network dial settings for 2250V are not the same as for 3000V. It's already been suggested that not knowing the right pi-network circuit values is a bad idea and a reason why those with a roller sometimes fail to get good performance.
I try to make the loading heavy to cut down on the chance of parasitics, and from what I understand this means using heavy loading, and if it is loaded heavily, the plate tuning won't show a huge dip as it does with light loading.
That is where it gets confusing. If it is running 500W out with 3KV on the plate and the plate tuning gives a dip from 300mA to 200mA, that is more of a dip than I see on, for example, the class AB1 NCL-2000 linear amp,  where heavily loaded it's like a 400mA current with a 50mA dip at resonance. Do class C stages give more dip when loaded correctly than linear amp stages?

Til tomorrow AM.. I ought rise before 8.

[WD5JKO]

Pat,

   It is good to hear that you've made significant progress on the Tucker. Maybe someday before you retire you will get to that onboard scope.  

   I'd suggest you log the settings, band to band. That way it will be much easier in the future to get the same 'Q' every time. Back when I ran a KW (mid 1980's), I ran 2500V @ 400ma with two big triodes in push pull. That dip to 400ma would 'peg' the 500ma meter with just a slight de-tuning of the plate tuning capacitor.

  I'm beginning to doubt myself now about how a class C plate modulated stage works. I am referring to this thread:

http://amfone.net/Amforum/index.php?topic=36906.0#


  Perhaps while you are at it, explore what happens to the modulation peaks when you vary the grid drive. Does the upward peaks compress as the grid drive excitation is reduced?

  As to 7160 local to us, I realize that nobody owns a frequency. What I wonder though is when there is a group that operates on a given frequency operating in a quick break-in type of operation, how long of a pause is required before the frequency is available for other amateurs to use it? Is it 15 minutes, 45 minutes, or perhaps 1 hour? This group getting on 7158 may not say a word for an hour or more, and then if I get on 7160, they start talking, and claim they have been on the frequency for hours. I have a vision of a bunch of retired guys talking themselves out, nodding over and drooling on the microphone, and then when I come on, it nudges them such that they continue on as if a silent hour had never passed.

  I'd like to keep 7160 for AM weekends since many of the stations that check in are using a crystal, and most Johnson gear PM when on 40M (VFO at same Freq as Xmitter). When the PM level is high, the audio is very muddy, and one sideband is nearly 30 db down from the other.

Jim
Wd5JKO

[DMOD]

I have a vision of a bunch of retired guys talking themselves out, nodding over and drooling on the microphone, and then when I come on, it nudges them such that they continue on as if a silent hour had never passed.

  I'd like to keep 7160 for AM weekends since many of the stations that check in are using a crystal, and most Johnson gear PM when on 40M (VFO at same Freq as Xmitter). When the PM level is high, the audio is very muddy, and one sideband is nearly 30 db down from the other.

Keep it up and if I can remember your time slot I'll join you. My Halli HT-40 does have a crystal for 7.160.

Phil

[Opcom]

I heard those guys on 7158.. one actually said where is so-and-so, and the other actually did say he probably fell asleep. Could they actually sit in chairs all day long except for making a sandwich or going to the pot, and all the rest of the time sit there??

However, if there's nothing for > a few minutes, how can they claim? I got on a couple times with an on-air test and asked if anyone was using the frequency. No one answered. 8-)

I did go through the bands, logged everything @ 500W carrier and 85% modulation, which is about what it'll do. It never has done 100%. Wow the drive, voltages, and currents are quite different progressing from 80 to 15M. It will get to 29MC, but no farther. That's OK I guess (but unexplained since it did not un-powered on the bench..)

It took a few hours to do this all, and I believe something cooked. I could not find it but smelled it, something pretty awful I never smelled before. It might be one of the fixed mica RF loading caps, because after that happened, the loading has to be set different, to include more of them. haha If so will have to go through the tests again. The caps are the brown ones as shown. The problem happened when only one of those was in the circuit as loading.

Has anyone had these caps go bad when used as loading caps in a KW size HF transmitter and keyed down for >30 minutes at a time with modulation, and 10 minute breaks or less. I thought those kind of caps were really durable.

But, it could be something else, just guessing. The insides of the RF deck is not accessible easily, that is, whatever was hot would cool off before I could get into it.

I have a serious grounding problem too. What is the best way to ground all the racks together? a piece of #6? a braid from RG-8? Then, how to GND this mess to earth? The wall is 6FT away, and the floor is concrete. At low power levels with a linear and plastic radio, a 20FT wire thrown on the floor has worked, but it does not work well for this. Should I run cable or braid to the girder in the metal wall of the building and then run a GND stake on the outside of the wall?

There's a parasitic tuned circuit in there at about 60MHz as well.

[Opcom]

I have a vision of a bunch of retired guys talking themselves out, nodding over and drooling on the microphone, and then when I come on, it nudges them such that they continue on as if a silent hour had never passed.

  I'd like to keep 7160 for AM weekends since many of the stations that check in are using a crystal, and most Johnson gear PM when on 40M (VFO at same Freq as Xmitter). When the PM level is high, the audio is very muddy, and one sideband is nearly 30 db down from the other.

Keep it up and if I can remember your time slot I'll join you. My Halli HT-40 does have a crystal for 7.160.

Phil

There is a mysterious substance in nasty old mikes that those in my circle call "moustache dandruff". It is crud in the screens and so forth that can be crustily blocking the mike. It may be the result of these old doddards activities as described above.

7160 AM on weekends is great! I'm in for next weekend now that the rig seems to work. If it blows up, I'll use the amplifier to get a 100W carrier.

I didn't know about the Johnson PM issue. Anyway XTAL stations are fine by me. The rule for the sat/sun 7160 RT is that everyone is welcome. I guess around Noon it gets started.

[WD5JKO]

   Looking at that RF deck, I'm wondering where the 4-1000 plate hooks into it.

Could the smell have been from RF heating whatever is on that paper plate? Is that left-overs from Colonel Tuckers last dinner? Did he wash it down from that antique beer bottle? 

Hope to see you Saturday on 7157...If the frequency is clear that is. 

Jim
Wd5JKO

[Opcom]

I should be on Saturday around noon or shortly after. In the morning is an antique radio swap meet.

Dennis W5FRS tells me that the old brown mica transmitting caps like those loading caps have serious problems at the higher HF frequencies and high power.

[WD5JKO]

Pat,

     perhaps some of those 100pf door knobs with npo dielectric would do better. use several in parallel.

      I have had something come up, and I cannot do 11 am to 3 pm. when I get back I will see what's left.

Jim
Wd5JKO

[Opcom]

OK well I'll probably be in the place listening anyway. At some point an air variable might be better as a loading cap for this monster. I wonder what voltage would be enough to avoid arcing when it is mis-tuned or unloaded by mistake. Would not want it popping like some storebought thing.

[VE3AJM]

So these fixed loading caps are not padding some kind of air variable/vacuum loading cap in the tank circuit already? Really?

Yeah, why not just install some kind of air variable loading cap in there as per many Radio/West Coast/ARRL Handbooks specs from over the years on variable loading caps for such a tx. Something that will fit into the space, etc. you have there in the RF deck. This thread and other multiple threads on this tx have been hard to follow over the years. They never seem to end coherently with resolutions of the issues raised. A new thread is simply started.

Al VE3AJM

[Opcom]

well it has issues that defy fixing and are sometimes tolerable and sometimes a show stopper. Most of the posts happen when something fries.

Right, the fixed caps are not padding. They are 'the' loading cap. It's shown in the schematic.

[Opcom]

I could have always put all of it in one thread, and continued threads that were months old, to continue after such gaps, but I think I was asked to start a new one if the old one is very old, so sorry if that is annoying. The part of the rework of this that has been discussed here has been going on for a few years, no doubt the longest rework of any piece of equipment on this site, and the total job is well over 10 years old, many head-scratchers and horrors having been found in the equipment. There have been many modifications from the old circuit 'as found' to the current evolution, so there are things that will not be consistent from one topic to another, and other things that were never resolved because they were inappropriate-design, could not be made to work, and the entire section was changed. Like many home made equipments, it may have all the bands on it but may have never worked right on more than one or two, or been used and debugged on more than the OM's favorites.

Between topics/posts for this, here is an example (from yesterday) of what is missing between the posts, so believe me you probably do not want to see the in-between stuff:
When it is modulated too hard, the PA plate current goes to 0, the mod current goes to 800mA, the RF goes to half voltage with square wave-like modulation, the screen current goes to 0 and the screen voltage to max, and the grid current disappears although the bias remains the same, and half the drive is reflected back to the exciter (SWR goes to 4:1). It repeatably does this over a time period of about 0.3 second (after adding 110% mod) and 'locks' in that mode. Once the modulation is throttled back it all comes back to normal in about 0.5 second, but there is hysteresis of this effect by the modulator drive level.  OK what is that?? Yes, what indeed is broken? No I do not expect any answers to that mystery and will have to figure it out myself over time -  it is one of the kinds of weirdo haunted things I do not ask about here, and explains why the topics do not go together because they are separated by complex or hellish problems that can barely be described much less discussed. So, see there are ghosts in this thing and that is why these topics to not have a conclusion yet, because as each wrong thing is found, it is only one step closer to having it finally work right.

On the good note, I had a nice QSO with a new person today and there was not a bit of trouble from the above. He has a 32V and 75A-2 and a strong signal for 100W. On more on 7160. Just having fun.