Viking 2 Modulator

[KZ5A]

I'm working on a Viking 2 restoration for AM use. 

My goal for this radio is to produce high quality, low distortion audio within reasonable bandwidth constraints like 150hz to 3.5kc.    Not looking for high bandwidth, just high quality.   My parameters for this are to generally stick with the original architecture and iron, all controls "work as original", re-built/re-designed with modern components as required, but no active semi-conductors.  So far, all the power supply filter caps have been "up-sized" by a factor of 8 to 10, all the rectifiers replaced with SS, and a Harbach soft-start module added to the HV.  The modulator screen will be regulated, derived from the low voltage supply, and switched on for AM TX by the PTT relay.  The original 5R4 HV rectifier sockets are being recycled to hold a pair of VR-xxx tubes for the modulator screens. 

I expect to be running this TX barefoot so want to produce as much PO as reasonably possible.

After reading and re-reading everything on the Internet and in back issues of ER, the K6AD audio modifications in ER #110 and 111 seem the most compelling.  Those mods preserve the 6AU6 audio tubes and will work with the Hammond T124B inter-stage transformer that I have on-hand.

However getting normal 2X 6146 power levels with near full modulation requires adding an additional audio stage and is still marginal, power wise.

Considering that a fundamental shortcoming of the original architecture is the inability of the 807's running in AB1 to fully modulate the 6146's, I wonder if it wouldn't be easier the convert the 807 modulator tubes to 6146's in AB1 than to add the additional audio driver circuitry to support running the 807's in AB2 with low distortion?

This seems like it would be an easy conversion, particularly since I'm rebuilding the screen and bias supplies anyway.  But I haven't heard of this being done, and I wonder if anyone on the board has any experience, thoughts, or ideas about this.

73 Jack KZ5A

 

   

[WA1GFZ]

I built a zener regulator for my 807s and run AB2. I put a 6aQ5 in the driver.

[The Slab Bacon]

Viking 1 & 2s, Rangers, Valiants, Apaches, and others, there are more mods out there than Grandpa's secret recipes for barbecue sauce! ! ! ! ! !

You got to look over as many as you can and pick the one that works for your needs and skill level to do. I have done bunches of them over the years and have kinda settled on the 12AX7 speech amp ones. Whether or not to use a fuzz inverter or driver is strictly dependent on the mod tubes used and the class of operation that you want to run them in. (Or how much audio you need to have on tap)

If you dont need to push the modders into the grid current range, a fuzz inverter is preferable. But a driver and driver transfoma helps keep it simple. (kiss rule)

[WD5JKO]

Jack,

   As Bacon pointed out, there are so many options to follow here. It depends so much on your ability, and time available devoted to these modifications. I cannot see the ER issues you mentioned, but it is typical for posted mods for these rigs to not be 100% complete. There is usually a gotcha somewhere that you have to figure out.

   One suggestion might be to look at the original design, and then weigh the pros and cons of it. What did Johnson do right, and where did they fall down? Then based upon your already stated goals, try to take a stab at it based upon your experiences, and motivations. This way the changes will all be yours, and you will learn a great deal. When I do this myself, I have a blast getting there.

Jim
WD5JKO


Oh, look at the VR tube data, and pay attention to the "minimum DC anode supply" specification:
http://www.mif.pg.gda.pl/homepages/frank/sheets/141/g/GL874.pdf

[WD8BIL]

Considering that a fundamental shortcoming of the original architecture is the inability of the 807's running in AB1 to fully modulate the 6146's,.......

The problem is NOT the 807s, in this case. It's the 6AU6's inability to drive the 807s. I use a variac on the plate supply of my hotrod Viking 1. I typically run 135 watt carrier with the 4CX250B final at 100%+ modulation using the 807s. The trick is to drive the 807s sufficiently which the 6AU6 cannot do.

Slab's 6AQ5 is a great choice or, like mine, I use a 12AX7A input amp followed by a 6C4 into the interstage transformer. The 6C4 has twice the drive of the 6AU6. Checkout the mod here:

Viking 1 and 2 Audio

Good luck, have fun and be safe!!

[KB2WIG]

FWIW, The RCA tub manuel has a ckt fer runnin the 807's in special triode connection.  10 watts of drive gives 120W output. This is enough to make the Vikings sing.

Hear is a link.

http://www.youtube.com/watch?v=g8huXkSaL7o


klc

[KZ5A]

Slab, Jim, and .....

I'm a little vague on what a "fuzz inverter" is

The original design certainly has it's interesting aspects, and for the most part seems to accomplish the designer's goals.   One of those was the "high intelligibility" audio passband which I'm sure seemed like a good idea at the time but resulted in the Viking 2's reputation for thin audio.  This limitation was compounded when the finals were upgraded to 6146's in the Viking 2 without a corresponding upgrade to the modulator section. 

I have looked at several "audio upgrade" articles for the V 2, most of which focus on the low level audio stages and don't deal with the basic problem of the original 807's not being able to fully and cleanly modulate a pair of 6146's.

Originally I worked up a design for new low level section based on the 12AX7/6C4 approach but with a 6EU7/6C4 circuit partially plagiarized from a 60's vintage Heath Hi-Fi amp and a better inter-stage transformer.  This would be a big improvement but still requires running the 807s in AB2 which gives rise to a bunch of complications I'd just as live avoid.   AB1 is much simpler to implement well due to the more consistent impedance presented to the driver.     

The June/July/August 1998 ER articles are by far the most comprehensive and well explained look at the Viking 2's shortcomings that I have run across.   The author's solutions were too "transistor-ish" for my purposes but the technical analysis of the root problems was excellent.  In addition to a solid discussion of the AB1/AB2 807 issues there is also a good look at the 6AU6 audio stages with what seems to me like a good low distortion re-design that preserves the 6AU6s.

The only "gotcha" I detected in that article was doing away with the 20K HV bleeder resistor which is probably OK for AM only, but looks like it would not maintain minimum current to keep the HV from soaring on CW during key up.   I could be wrong on that, but that's my understanding of choke input filters.

So baring hearing a "been there, tried that, didn't work" my direction on this project is to replace the 807 modulator tubes with a pair of 6146W's (B's) running in AB1, preserve the 6AU6's with new circuitry, use the Hammond 124B inter-stage transformer, and also replace the finals with 6146W's.

The lower screen voltage on the 6146 modulators (~200V) relative to the 807s (~300V) makes deriving it from the LV supply practical assuming a provision is made to switch it on and off along with the HV.

The power supply mods are focused on substantially improving the dynamic regulation, eliminating the voltage drop of the tube HV rectifiers and regulating the bias and mod screen voltages.

The RF section will probably end up unchanged except for replacing the usual old out of tolerance R's and C's.

If all this works as hoped, I'll end up with a really good sounding, well modulated, 125W AM TX with no new holes, no transistors, no empty sockets (the rectifier sockets are being recycled to hold VR tubes), and no changes to the normal operating procedure.   The only externally visable changes will be the  VR tubes in the ex-rectifier sockets.

This is a long term project, no big rush.   It has already caused me to review/relearn a bunch of stuff that I had forgotten decades ago and has been a very enjoyable endeavor.   I'm taking pictures as I go, mostly to help me remember which wire went where originally.  Maybe if it works I'll post the details.

I'm always looking for input and appreciate everyone's comments and suggestions.

73 Jack KZ5A






 

[WD5JKO]

Jack,

   You definitely have a plan, and I'd say run with it. If your like me, 90% of the fun is in getting there.

   Heck, some would argue that 6146 are poor audio tubes, but I still recall that Acrosound application note where a pair of 6146's made "enough sound to fill a stadium", and did so with high fidelity. They used tertiary windings for the screens, and cathode follower drive though. Also the resting class AB1 idle current might need to be set at the tubes Pd rating for low IMD.

   So taking this approach you are running away from the pack, and doing something pretty much original. Good luck with that. I'm no stranger in that arena, pushing a Retro75 to 30w PEP output, or running a single EL-34 on 20m SSB grid driver, AB1, running a pair of 7591's on 160-80-40m at 65w PEP RF output SSB, or 15w AM. It's fun breaking away from the pack. 

Jim
WD5JKO

[WD8BIL]

Well, ok, Jack.

I still totally disagree with your evaluation of the capabilities of the 807s considering I am modulating a 135W carrier with a pair in AB1 now. The Viking Bud is well known to many on this board and it's operation can be easily verified.

CAN I GET AN AMEN ON THAT???

But, as Jim points out, getting a plan and working it to conclusion is a lot of fun. I hope to hear ya strappin' into Northern Ohio soon!!!

[W1AEX]

CAN I GET AN AMEN ON THAT???

I will second, third, and AMEN that Buddly. The Viking Bud has gobs of sparkling clean full-range audio. I'm also puzzled by the comments regarding the capabilities of a pair of 807's. I have never found them to come up short in the power department for handling the task of modulating a 100 watt final. I also agree that the issue with the Viking I and II series is in the low level audio stages. Before my grand expansion into a rack to allow for a pair of 811A's, I followed the 12AX7 -> 6AQ5 -> heavier driver transformer route and the 807's handled the task quite readily. As the RCA data sheets indicate, with +600 volts on the plate, a pair will produce +60 watts in AB1, which is more than enough to cleanly modulate a pair of 6146's to 100% all day long.

Then again Jack, if you come up with something better, go for it! That's more than half the fun of this whole adventure. It'll be fun for you and we'll all enjoy following your progress!

73,

Rob W1AEX

[The Slab Bacon]

Well, ok, Jack.

I still totally disagree with your evaluation of the capabilities of the 807s considering I am modulating a 135W carrier with a pair in AB1 now. The Viking Bud is well known to many on this board and it's operation can be easily verified.

CAN I GET AN AMEN ON THAT???

AMEN!!

[The Slab Bacon]

Slab, Jim, and .....

I'm a little vague on what a "fuzz inverter" is

73 Jack KZ5A

Jack,
       A "Fuzz Inverter" is amers slang for a phase inverter. If you are running tubes that make enough audio running AB1, you only need to tickle the grids with some voltage. 807s and 6L6s are fine with this. 6550s really get it done just barely tickling the grids with some voltage, and a phase inverter is OK-FINE to drive these tubes in AB1.

However if you want to beat the snot out of a pair of 807s in AB2 (much more outpoot) or run anything in class B where you must push them into the range of drawing grid current, a transformer coupled beefy driver is the way to go.

Typically in class A or AB1 a step up driver transformer is used, where in the classes of operation that want to see grid current, a 1:1 or even a step down driver transformer is often used.   (dont beat me too hard, Don    ) 

[KZ5A]

Slab,

A "Fuzz Inverter" is amers slang for a phase inverter.

Ahhh.   Yes, I would consider a phase inverter to be a better arrangement than most any inter-stage transformer for driving an AB1 stage, however I already have what I think will be a good xfrmr on-hand and preferred not to add a tube socket.   Might end up doing that anyway later if the transformer gives me any grief.

WD8BIL

I don't have anything against 807s, I'd rather keep them, but..... my understanding of plate modulation is that the modulator should deliver roughly the same PO as the RF stage and I don't see how one would get there with 807s in AB1, but then I don't know what a "Viking Bud" is either???

Hopefully I will be back up on AM this weekend with my HT-37/SB-200 setup .  The Halli needed a new LV filter cap (can) and I thought I was going to have to use dynamite to get the old one loose from the chassis. 


73 Jack KZ5A

[WD8BIL]

Well Jack, the Viking Bud is my (my name is Bud) Viking 1. I've modified it with a 4CX250B in the final. I've solid stated the power supplies and put the plate transformer primary on a 140% variac. I get just under 1050VDC on the plate with 250ma current that's @ 260W input. The full carrier in CW hits 170 watts.

Now, in phone the modulator is turned on and the plate sags a bit. There's a big power resistor in to drop the plate voltage to the 807s to 750VDC. I drop the carrier to 135 watts and the 12AX7A - 6C4 - 807s FULLY modulate the carrier beyond 100%.

It's been working that way since 1973!!! Only changed the 807s once. I thiink it was 1998?
25 years on a set of 807s singing for their supper ain't bad in anyones book!!!!

 

my understanding of plate modulation is that the modulator should deliver roughly the same PO as the RF stage and

That's one school of thought and it works well. But I've found that a 1:2 power ratio works just fine. Ifn you want "overhead" then the skys the limit. 200 watts will mod 100 watts too but why??

Whatever you choose I hope to hear it soon. Remember... we're all in this together!!!

Viking Bud

[Steve - K4HX]

I don't think the 807s are running AB1.

I agree with Buddly. Given enough drive, the 807s in the Viking I and II will easily modulate 100 percent. At about 100 watts out, I was able to get 120-125 percent modulation with mine.

[WA1GFZ]

AMEN 807s are a good tube for modulators and look the part
AB2 gives you more clean plate voltage swing. but you need to motivate the grid 1 a lot harder.

[WD8BIL]

Frank said:

AB2 gives you more clean plate voltage swing. but you need to motivate the grid 1 a lot harder.

Thus the 6AQ5 or 6C4 or (insert any tube stronger than a 6AU6 here) upgrade in the low level audio section.

I don't think the 807s are running AB1.

Yes, Steve, they are if you set the idle current according to the manual ala R13's tap. You can, however, set the screen voltage to idle them in AB2.

The VK Bud has a variable BIAS supply so I can keep it in either class when changing the plate voltage with the scrote knob. I typically run it in AB2.

[W2ZE]

As others have stated already. The greatest defeciency with the Johnson design is the 6au6 lower stages, not the 807's. The 6au6 driver stage is at best barely up to the task. To keep the original socket, use a 6aq5 or 6c4. For real umph, drill and blast a new 9 pin socket and use a 12ax7. My father has used a 12au7 push pull with amazing results. The 807's only need to make 50% of the final power to modulate the class C amp to 100%. With the 6au6 left in there, that will assuredly never happen.

[WD5JKO]

I agree completely that we can make the 807's play ball in the Viking I or II transmitter. That said, if somebody wants to redesign to get 100 watts of audio in class AB1, a single pair of 807's won't do it within ratings, so another tube choice that will do it needs to be looked at. Tubes such as the 6550, KT-88, KT-90, KT-100, KT-120, and yes even the 6146 can make that kind of power without drawing grid current. In fact the 6146 ICAS at 750V plate can make 120 watts of audio, and do so class AB1.

So Jack, if you want to go that route, by all means do so. As far as all the QRM your getting, that is what passband tuning is all about.

Jim
WD5JKO

[WD8BIL]

It's not QRM, Jim. If you consider his original claim;

KZ5A said:

Considering that a fundamental shortcoming of the original architecture is the inability of the 807's running in AB1 to fully modulate the 6146's, .....

it simply isn't true. The 807s have no problem modulating 2 X 6146s to 100% in AB1. Many Vikings 1 and 2s have proven this time and time again. I'll meet ya tonite on 3885 with my Viking at stock power with the bias on the 807s set for AB1 and prove it.

Jack's notion of a 1:1 power ratio is where his original statement above shows an error. You simply do not need 100 watts of audio to 100% modulate a class C RF stage that is setup correctly.

It may be a different story when trying to modulate a class C triode but we ain't talking there. The Viking 2 plate modulates the final stage at 100 watts output. In this configuration the 807s have more than enough scrote for the job. That is undeniable.

The answer, for Jack, will be found in what Jack ultimately decides he WANTS to do.

[The Slab Bacon]

In theory, you need audio power equal to 50% of the DC inpoot to the final to modulate 100%. So for 180w inpoot (I think that is what a V-2 runs) you should, in theory need 90w of audio power to moduate it. However...................

In practice, it is nice to have the ability to produce audio power equal (or somewhere close to) the carrier outpoot power to have some extra "headroom" and be to make in excess of 100%. This really makes for good "Munky Swing". Also, this will also reduce distortion as you are not beating the snot out of the modulator.

[WD8BIL]

Also, this will also reduce distortion as you are not beating the snot out of the modulator.

Where's the fun in NOT making a tube sing for it's supper? 

The Viking 2 manual states phone output is 100 watts. 807s can handle that no sweat!

[W2PFY]

CAN I GET AN AMEN ON THAT???

You be the MAN Bud!!!

[WD5JKO]

In practice, it is nice to have the ability to produce audio power equal (or somewhere close to) the carrier outpoot power to have some extra "headroom" and be to make in excess of 100%. This really makes for good "Munky Swing". Also, this will also reduce distortion as you are not beating the snot out of the modulator.

   Well said. Not discussed thus far is that modulation transformer. Between its turns ratio, and power rating my bet is it has to go to get that audio headroom. Consider the Heath Apache with its ~ 2:1 step down ratio (imp is 11K/3K). That thing by design just cannot get it done. Anybody know the Viking II mod transformer turns ratio?

   Johnson by design had several built in limitations to the audio headroom:

1.) Hi-Z audio driver (6Au6) that cannot supply any grid current to the 807's, so class A, and AB1 are it.
2.) A 807 grid bias supply that is high impedance. Anything done to fix the 6AU6 problem is then offset by the bias supply that shoots more negative as the 807's draw grid current, AB2. That same bias also cascades to the RF stages as pulsating grid modulation.
3.) As soon as the 807's start to draw some serious plate current, the Hi-Z screen supply (tap on HV bleeder) caves in on audio peaks, further limiting the power.
4.) The modulation transformer is sized such that it can only handle so much audio with the 250ma DC current flowing through the secondary. The turns ratio might be deliberately limiting as well.

My last Viking experiment used the Stancor 125W multi-match transformer with outboard audio.

Jim
WD5JKO
 

[KZ5A]

Buddy,

The "Viking Bud" sounds really interesting. The 4CX series of tubes are hard to beat in terms of "bang for the buck".   There is a set 4CX300A's with sockets and chimneys in my parts collection that may find themselves in my SB-200 after I finish the Viking 2.   Last year's RTTY contests took a toll on the T-160L's and they are a little weak now.

Did you change the HV filter in the Bud to capacitor input or just use the variac to raise the output voltage?  Also wondering if the screen is modulated.

In any case it's good to know that the power iron in the Viking can deal with fully modulating 135Ws long term.

The Viking 2 manual states phone output is 100 watts. 807s can handle that no sweat!

I think this a function of the finals being upgraded in the V2 to 6146s from the 4D32 of the V1 without a corresponding increase in the modulator.


Steve,

"My understanding" is that the Viking 2 modulator in stock form operates in AB1 up to around 75-80% modulation.  Above that it goes into AB2 and generates tons of distortion due to the inability of the 6AU6 driver to provide the necessary power cleanly.  That is the issue dealt with by the various 6C4/6AQ5 mods.  The alternate approach I've run across is to just run abt 80W of carrier.

All,

One of my regrets in this project was that my Viking 2 was "toast" when I acquired it so I didn't have an opportunity to experience it in stock form before setting out to modify it.   That eliminates a handy frame of reference for evaluating potential mods.  There is a bunch of mod info on the internet and in old ER magazines but none of that is as useful as the comments and experiences related on this forum.

The difference between "plate modulation" and "plate and screen modulation" is no doubt the source of the confusion on my part about the requirements for modulator power.  The RCA TT-5 manual has a note recommending plate plus screen modulation for the 6146.         The only 4x 6146 rig that I am familiar with is the B&W 5100 which also does plate plus screen modulation.  Eimac also recommends it for tetrodes in general in some 1952 vintage application notes.  So I guess its just a tetrode thing and I'm stuck with it.  There is still the basic "conservation of energy" principal involved, meaning the output can not exceed the total of the RF power and the modulator power, i.e. "there are no free watts".

In the Eimac notes I did run across a discussion regarding the "degree" of screen modulation.  Thinking abt it, totally screen modulated finals (DX60) tend to have PEP~=CXR, fully plate modulated triodes generally have PEP=4x CXR, and apparently tetrodes fall somewhere on the line between depending on the degree of screen modulation.

Seems to boil down to either upgrade the driver to support AB2 or upgrade the mod tubes to supply the needed power in AB1.   Being the contrary old codger that I am, doing something differently than it has been done before has a lot of appeal.   It's all about having some hollow state fun anyway.  If it was just about getting on AM I could just stick with my HT-37 or, heaven forbid, put my K3 on AM.   

So I'm going to proceed with my 4 - 6146B's configuration and see how it comes out.   If it works there will be yet another Viking 2 mod floating around the internet.  If not I'll have to come up with a "plan B". 


73 Jack KZ5A