Neutralizing question

[ka1tdq]

I have a question about a neutralizing circuit for the rf deck that I'm building.  It's a pair of KT-88's being driven by a rice box.  I'm putting a 10db attenuator in line to knock down the power sufficiently and vary grid current by the amount of drive and/or detuning the grid tank.  Here's how I have the grid circuit setup:

Input tank with bottom end grounded.  The top end of the inductor has 5 turn link coupling.  100pf cap on top of tank then goes to grid.

Grid has a 2.5mh choke then grid dropping resistor to ground.  I'm measuring grid current on bottom.

Here's my question:  Can I run the neutralizing capacitor from the plate to the bottom end of the choke and then from that point the 330pf-ish bypass capacitor to ground (it would be in parallel to the grid dropping resistor)?  Will it work?

If not, what would be a good alternative so that I could keep the tank circuit how it is.  For construction ease I want to keep the variable cap grounded and not have it floating. 

Thank you.
Jon KA1TDQ

[K9ACT]

Judging by the reads vs responses, I would guess a schematic might be in order.

I do not like word problems when a picture would make it all clear.

js

[ka1tdq]

This pdf should make it much clearer.  I'm pretty sure it'll work but I want to run it by others to eliminate hours of experimentation.

[The Slab Bacon]

It probably WONT work as drawn. The neutralizing cap is tied into the bypassed side of the grid choke. this will prevent most (if not all) of the out of phase rf from getting back to the grid. Thus making it somewhat null and void.

                                                          The Slab Bacon

[KM1H]

Why use a 20-25W signal to drive a pair of 6L6GC's anyway? Your gain is less than zero.

Carl
KM1H

[ka1tdq]

I'm not sure how much power I'll need to get the necessary grid drive so I'm starting off with 1/2 watt of RF and I'll increase it as needed.  My goal is to use a rice box for excellent frequency stability and have the benefit of plate modulation.

Thanks for the input on the neutralization circuit.  What would be a good way to make it work (keeping the tank the way it is)?

[K9ACT]

Thanks for the input on the neutralization circuit.  What would be a good way to make it work (keeping the tank the way it is)?

You can't keep the tank the way it is.  The traditional method is with a balanced tank somewhere, either grid or plate.  You need a dual cap and a center tapped coil.  Feed the bias to the center tap and you can ground the cap.

You will never achieve 180 phase shift with the method proposed.

But the more important question is, why do you want to neutralize a tetrode in the first place?

js


 

[KM1H]

Lift the base of the coil/cap above ground with a .001 cap and connect the Cn to the junction.

However I doubt it it will require neutralization at 20M or below. CE managed a pair of 6550's 160-10M without neutralizing.

You can also eliminate the input altogether and drive a 50 Ohm resistor, that solves all the problems.

Carl
KM1H

[WD5JKO]

Jon,

    Look at the Valiant schematic posted elsewhere on this forum. The neutralizing circuit shown there would be easy to copy, and works. Notice the 10pf variable N/C has both sides hot. You could replace this with a suitable fixed HV cap (maybe use coax), and vary the other cap at the low side of the input coil.

    Also beware that the recent production high transconductance beam power tubes and power pentodes might not be the same as the vintage tubes by RCA, GE and others. The interelectrode capacity and lead inductance might be far different than the old stuff. The KT-88 (recent production) might have a plate to G1 capacitance (Cg-p) of about 1.3 pf.

    The only tube in the family that I know of that was successfully used as an RF amp up thru 10 meters was the 6550. Again the new 6550's may not behave like the old GE's did at RF.

    The best tube for RF in that family that I have found is the 7591. This tube has the GM high, and is small like a 6V6 yet it boasts a 19 watt CCS plate dissipation. Further it has a Cg-p of about 0.25pf, has a very short cathode wire, and has dual G2 pins to help bypassing. Using four 7591's will be in the same league as two KT-88's. Unfortunately the recent production 7591's fall short on power output when compared the the vintage 7591a's. Both JJ Tesla and EH make 7591's today.

   I use the 7591's in my Central electronic 20a as a class Ab2 RF linear amplifier. These tubes are extremely linear. I can get 65 watts PEP out from them on 80m without neutralization with a plate voltage of 560v, G2 at 400v, and about -22v G1 fixed bias.

It would be interesting to evaluate some of those high Gm audio tubes with class C plate modulated service. My fear however is that high interelectrode capacitance, and high lead inductance (cathode and G2) might lead to an application where instability might be hard to get rid of. Doubling the plate voltage at 100% + modulation might cause internal flash over too.

Please let us know how your Kt-88 amplifier turns out.

Good Luck,
73
Jim
WD5JKO

    

[ka1tdq]

Thanks everybody for the input!  Let's see if I can hit everybody...

I never did think about the increase in plate voltage during peak modulation, but since the B+ is about 425 I think I'll be okay. 

I only plan on using this on 80 and 40 meters so maybe neutralization won't be a big concern.

I've seen the balanced tank schematics but haven't put too many neurons to work on it, so it's new to me.  If I did need a neutralization circuit, I'd probably just lift the ground end of the tank and run a .001 to ground like someone said.

And no-neutralizy on the tetrodes?  Hooray!  Don't know why, but Hooray!

...and I will let everyone know how this turns out as I hope to have it on the air in a few weeks after I get the fundage to buy the tubes.  Everything is constructed except for this neutralization thing which I will forgo for right now and see how testing goes. 

[K9ACT]

This sort of project usually comes about because one has a couple of tubes on hand and wants to do something with them.

It's not clear why one would design a new rig from scratch using tubes that are not even designed for RF if you have to now go out and buy them.

js

[ka1tdq]

I know, KT-88's are for audio but I think they'll do for this as well.

My reason?  I plan to use this as my main AM rig and I wanted to use 'low' power tubes that will still be in production for years to come.  Plus, they have that 'Made in Russia' label on them that makes them really cool.

[WU2D]

Jon,

Gee I think that those KT-88's will sing without some attempt at neutralization, especially if you have a Hi-Q grid tank and less than perfect shielding, grid to plate. The circuit that you are trying for is called the single ended Bruene (grid) method. It was developed as a compromise circuit for single ended tetrodes for many of the same reasons that you are describing. It basically uses a ratio of capacitances to squeeze a little negative feedback into the bottom of the grid tank.

Take a look at this article http://www.cpii.com/docs/related/22/C&F5Web.pdf

73's Mike WU2D

[ka1tdq]

Thanks again for the tip!  I will use a balanced tank neutralizing circuit like someone had mentined before... I'll be able to ground the variable cap.  BUT, how do I calculate the resonant frequency of the tank... Do I use the value of just one side or factor in both?

[K9ACT]

Good question and one that has stumped me many times.

The answer is:  use the inductance of the full coil and the capacitance of one section in your calculations.  I will never understand why but that is what works.

Works out to about 40pF and 40uH for the ghetto.  Q depends on your source impedance if you can determine what it is,

A dual 100pF is a good choice with lots of windage.

The best calculator I have found for these calculations is Mini Ring Calculator. Not sure where I got it but google should turn it up.  It works for air and ferrite and has every combination you could want.

js

[ka1tdq]

Mini Ring Calculator... got it!  Thanks again for the tip!  Now to hunt eBay for a dual 100pf.

[WD5JKO]

I'd like to relate my experience with using a pair of EL-34's as RF amplifiers. Without neutralization I had a lot of problems.

Look at the article "Evolution of the QRO 20A":
http://pages.prodigy.net/jcandela/Central%20Electronics%2020a%20QRO%20%232/

Later I tried using a single E34L (more stout EL-34) and with neutralization it worked well up through 20 meters.

See 20aQRO#3:
http://pages.prodigy.net/jcandela/CE20AQRO/


   These Hi- Gm audio tubes CAN work at RF, but I think the challenges are greater than when using tubes made for RF.

  One tube I want to try at HF is the 7984. These are VHF tubes capable of about 40 watts out on 2 meters. They are pretty stout with ratings that impress. The neat thing however is that these are not scarce and they are not expensive. Anybody try the 7984 on HF AM?

73
Jim
WD5JKO

[K9ACT]

Mini Ring Calculator... got it!  Thanks again for the tip!  Now to hunt eBay for a dual 100pf.

ebay is probably a waste of time.  I bought one last night from RF Parts for $22.

Good luck.

js

[ka1tdq]

Yeah, you're probably right about using audio tubes for rf.  I will take your advice and use the 7984's instead of the KT-88's.  Since I already have the rig built (minus tubes) I'll need to remove the octal sockets and stick in a couple 12 pinners and do minor circuit changes.  But the plate will still be below the chassis, along with the high voltage, as it is right now and will work nice cosmetically.  Otherwise, I probably would've just gone with a couple 6146's or 2E26's.

...and, it may be original too.  Like you aluded to in your email, maybe noboby's done this before. 

Thank you again.
Jon