The AM Forum

My California Kilowatt.  IOW, let's make legal limit AM.

Well, almost done.  Did a test fire today  no fires or smoke except into the dummy load!

Power supply is Pete Dahl (new, not original) 4kv 3A CCS xformer, cap input with A FWB of K2AW diodes.

Has 242 volts on standby, 230 volts full power.    As such, it's 7.5 volts on the filament standby and 7 under fire.

I don't think I need the extra emissions of the half volt....

Pics speak for themselves.  I was able to keep the pep value all the way down to a 250 Watt carrier with my assymetry board, but if you don't have a sync detector  or sdr....

--Shane
KD6VXI

Power like that would definitely make you the Channel Master!

Jon

 :'(I just want QIX, JJ and HLR to be able to hear me.  I miss my mountain top.

--Shane
KD6VXI

And no, I don't plan on running it like this.  More along the lines of an 8930s in class A and an ANAN 10 for the exciter.

Quite impressive for a single 3CX-3000A7, Shane!

242V > 230V is about a 5% drop which is not too bad. 7.5 fil to 7V is about 7%.   Though, both of these sagging conditions have the potential to create minor IMD issues. Voltage variation is not good.  But by increasing your HV filter capacitance, you will be able to soften the HV drop to the tubes. HV will vary at an audio cyclic rate and time constant as created by the tube current draw and the  capacity of the filter caps. Just imagine if you had infinite capacitance - there would be no HV drop at all once the caps were initially charged. How much does the HV sag now from 0- 10 KW and from 2K to 10K?

I run about 200 uF in my HV supply. This consists of two caps weighing 60 pounds each... :-)

You might consider a 120V "power conditioner"?... is that what they are called?...  to hold the 7.5V fil voltage tight. Varying voltages are probably one of the biggest issues hams have to a clean signal.  (Besides flat topping / overdriving and loading improperly.)

You will need plenty of air, but 2K carrier and 10 pep (into a dummyload) is making that amp sing for its supper.

Good job!

T

I have a couple more oil caps I can add.  It's at 32 or 34 uF now.....  I can't remember which cap is in it.

I bought a CNC mill to do boards with.  Soon as I master that, I'll mill some capacitor boards and string some electrolytic.  Best I have now will cover 2.8 kv.

I'm contemplating running the fil xformer, control circuitry  and blower off a secondary power source. That would allow me to change to a regulator as your talking about. As it is, I have to change the tap to 240 volt.  As I told Pat (he's got the same fil xformer)  the fil xformer is really good, but you do have to fine tune it.  With taps for 208, 30, 40 and 250 I can dial it in.

--Shane
KD6VXI

Tom,

I'll have to look at the hv sag for different levels of output tonight.

I was able to run 6kw carrier for a minute and 20 seconds last night.  Was my smoke test.

Tonight I'm thinking of blasting the dummy load with Rush - - Spirit of Radio for a good smoke test. :-)


On another note.  I have a Johnson kw tuna.  I understand that it's rated for desk kw output, which is 600 carrier and 100 pct modulation, correct?

Don't need to replace things that don't need replacin....

--Shane
KD6VXI

" Time Has Come Today "  by the  Chambers Brothers would give it a workout.



KLC

I'll give that one a listen.

On my drive over here, the song 'I'm on a Mexican, whoa oh.... Ah radio! '...  Sounded like a good one too lol.

--Shane
KD6VXI 

My go to test audio of late has been the Stray Cat's "Gene & Eddie"

https://www.youtube.com/watch?v=afkfos_oQro

'Ring My Bell' is a favorite as well...  The multiband, etc I run really likes her voice...!

Spirit of Radio is more to for the lyrics....  But I do love me some Rush!

"Begin the day
With a friendly voice
A companion, unobtrusive
Plays the song that’s so elusive
And the magic music makes your morning mood

Off on your way
Hit the open road
There is magic at your fingers
For the Spirit ever lingers
Undemanding contact
In your happy solitude.

Invisible airwaves
Crackle with life
Bright antennae bristle
With the energy
Emotional feedback
On a timeless wavelength
Bearing a gift beyond price –
Almost free…
All this machinery
Making modern music
Can still be open-hearted
Not so coldly charted
It’s really just a question
Of your honesty

One likes to believe
In the freedom of music
But glittering prizes
And endless compromises
Shatter the illusion
Of integrity

'For the words of the profits,
Are written on the studio wall,
Concert hall –
Echoes with the sounds…
Of salesmen.'"

The first half of the song just reminds me of wandering around listening to my AM radio as a kid, on the go.  When radio was invisible magic...

--Shane
KD6VXI

Very nice! How about the input circuit? Can you enlighten us on that?

Nice work Shane...

Is this an 11m Amp?  :-X

Noticed two propane tanks right behind the Byrd wattmeter. If that oil filled dummy load blows, the propane tanks might go next.  :-[

At that kind of power, especially if running upper HF (20m up), human tissue in the presence of strong RF fields gets hot. At work with 3KW 13.56Mhz RF amplifiers, even with the covers on, RF leaks out of the cracks. I used to sometimes write the test sheet using the amplifier cover as a writing surface. No more, since I got a hot wrist and a lingering pain.

Jim
Wd5JKO

It's not for 11, but it would cover it.  It covers about 19 mhz to just over 30 now, continuously.

He has a low band (160 to 20) meter amp that Is a work of art.  Vacuum relays for each tank tap, servo controlled vac variables, etc.

I have to replace a tuned input cap, and as such, will be yanking the rf deck out again.  At that point, I'll snap a bunch of pics.

I tried a few input circuits so far.  Tried a simple L, a parallel resonant and now it's got a pi. When I used the other, simpler inputs I had a low pass filter at the input, something akin to class F.  Didn't seem to make a whole lot of difference and efficiency was the same or better with a pi input, AND I kept the bandwidth, AND the parts count was lower.  Less things to blow, but the receiver cap crapped out.  So, waiting on the Russians now to deliver a pair for the input.

The output got a 5 to 200 pf vac variable on CTune. The original Henry just didn't have low enough C min to get to the am region on ten.  I'd bottom the cap out.  CLoad is a 500 pf IIRC.  

The 'down the road' upgrade will be servos on CTune and CLoad.  Ultimately its going into a C Train somewhere in the middle of nowhere..   On a remote station.  But that's for another time.

For myself, I bought both a 3cx3000 and an 8877 powered rf generator.  The 3k will be 75 and 40 and the 8877 I have a beautiful roller inductor for the upper bands.

--Shane
KD6VXI

All I see is a little grey box with a meter... wonder what that amp looks like?
Did I miss that post?

Yeah well, I saw the Chambers Brothers in 1971 at a distance of <20ft, with an audience of about 40-50 people
in a recording studio, in Hempsted LI, at a WLIR live broadcast from said studio. They did THAT song. BAMMMM!!!

Whew. They were good, really good!

                   _-_-

But that those tapes were out somewhere... but that they were. :(

Bear,

It's a Henry rf generator conversion.

I have more to post, but here's the chassis and xformer.  
The small xformer seen below the glitch resistor bank is the original 12 volt supply for the relays and led.  I inadvertently ran 220 to it instead of 110, so that xformer smoked.  It now has a DIN rail Meanwell switcher slightly lower.  That one takes 240 in, so all is good.

I'll snap better closeups, etc when I get home.  Going to boil more mineral oil tonight lol.

--Shane
KD6VXI

Mas

Output network.  This was with an 800 ohm plate resistor.

--Shane
KD6VXI

Mas mas

Picture of the rf deck and the front faceplate I built with switching input has TX switch as well as ptt.  Their is an interlock ptt out the back so amp keys before the exciter is sent ptt.

Their is also a switch for GG or completely variable bias.  I can go from 4 to 37 volts of bias on it.. The rf deck has an additional 15 x 10A10 diodes switchable to get truly class c.  Visions of plate modulation in my head, and the owner liked the idea.

That faceplate is completely removable for maintenance via a pair of molex connectors.. One connector for the high voltage input display, the other for all low volt, ptt, etc.

Their is a 3 amp fast blow fuse in the cathode, tested works..  In actual operation, it will be replaced with a 1.5A.  I also have 4 X 200 ohm 200 Watt resistors in series with the HV line as a glitch resistor.  It's also been tested, the amp faults and high voltage xformer contactors shut off 240.  A fuse also blows, but I used a bolt across the hv fuse holder to test the actual circuit breaker.

--Shane
KD6VXI

Their is also not visible,  but to the right side of Cload a choke going to ground, in case CBlock was to fail.

Jim, I was thinking of those propane bottles actually.  They are going to be moved.  I used to use them to solder large wire, strap, etc, but now have a 260 Watt (iirc) gun, so it's somewhat superfluous.

Also, I talked with another prominent Canadian builder who had warned me about tinned and braided strap.  The 3x6 they had built, on 10 and 15, turned into a 'total gong show' as Jim says when hit with any qro.   Flashes and bangs galore.

As you can see, I used untinned braid.  I DID solder about half inch on each end and then drilled through for my attachment points. 

--Shane
KD6VXI

Not bad for an 11 meter amplifier... ;D  Sorry.. I just call them as I see them.. you do need to ditch the brade..  Don't worry.. We know your rf generator will never leave the dummy load.. ;)

Why ditch the braid?  It's working.  What reasoning would I need to lose it. 

As to it being an '11 meter' amp, what the end user does with it is up to him.  I built it to a specification, and it has surpassed almost everything the owner asked for.  Which was, to convert it from a CB amp to something usable from 15 meters on up.

It does this.

Building a CB amp today would be an excessive in futility.  The monster amps are for sale for a song with no sunspots.  Hence the owner picking this one up to convert.

--Shane
KD6VXI

I'm not criticizing you at all.. I understand what your doing.. no big deal to me.. as far as I'm concerned the amount of power you run should only be limited by your electrical service.. no matter what band it is..
Any braid in rf service can cause problems.. BE's know all about it.. I'm just reccomending you change it so that you don't have any future headaches.. especially sense I know how 3x3's end up getting run.. I've seen braid that has arced between the weave..  once again your mileage may vary.. I've just never had any luck with it..

I have some strap on order, I will replace it before this leaves. I was counting on the additional L before this CTune to get it on the am portion of ten.  It worked, but barely. A new cap that has 5pf min fixed it perfect.

Ultimately, this amplifier is going to be run in a pretty hostile environment.. A remote station in a connection box, from what I've been told.  He has another remote amp that is beautiful.  Vac relays for coil taps, steppers on tune and load, etc. It only covers 20 to 160 however. That one was purchased for a song, as the original consignor had paid for it, gone SK and his estate told the builder they didn't want to deal with it, dispose of it. The entire remote contest station in the desert of California ultimately went that way.

--Shane
KD6VXI

Sad to see someone pass and all their stuff get trashed but at least your guy was there to save it.. It still has all the air intake filters correct? I say that because if its truly going to be remote you obviously know as well as I do that dirt/dust and high voltage do not mix..

Hi Shane,

Good decision about ordering the copper strap. I usually buy a big sheet of roof flashing at Home Depot. It lasts a long time. Cut it with snippers or a bandsaw and then sand the edges smooth - and round the ends.


About using braid inside the pi-net circuit....

Remember that the circulating current going thru the resonant circuit - coil, two capacitors, bandswitch if used and connecting leads (braid) is Q * plate current.  Circulating current will be determined mostly by the final L/C ratio you are running. This is circulating RF current that can get ungodly high if the circuit is very high Q.

The point is, why make that beautiful coil so heavy duty if the leads connecting it are smaller or a poor RF conductor?   Use the same effective diameter all the way thru for good practice. (wide copper strap)

Outside of the pi-network, the current is back to normal plate current until it gets transformed down to 50 ohms. So the lead connecting the tube plate cap feeding the  pi-net class B pulses carries smaller current, though, using heavy copper strap thru-out the complete RF plate circuit is a good practice to maintain low inductance and better stability.

The danger then becomes breaking the vacuum cap glass seal due to mechanical strain. The solution is to use a strong ceramic post pillar close to the vac cap that takes the mechanical strain of the incoming unsupported strap. Then run a second copper strap loop supported by the pillar to the vac cap that is formed to be a perfect fit. No mechanical strain and at the same time low inductance copper strap.  

A classy add-on is to send out all the pre-formed RF straps for silver plating.  Or, there are DIY kits for silver plating on the web.

T

With the proliferation of projects, I'm contemplating the brake, shear, etc for 4 hundred at Harbor Freight.  It's harbor freight, but that sure would make a LOT of strap out of copper sheet!  :-)

I know about cool amp silver rub, used it on a couple tanks before.

My initial reason for braid was multiple.  I wanted flexible, so when things expanded and contracted things wouldn't break or be stressed.  And the original vac cap on CTune was too large, so I was hoping for a bit more inductance.  What I didn't take into account was, all those micro edges between each piece of braid, etc. As you and 'www explained.  Thanks, now I have an understanding of WHY it's bad juju.

The Q is designed at 6......  Another reason CTune had to be so small.  An exciter with predistortion  will be used, with the final pa wrapped into it) so the slight increase in imd wasn't too much of an issue.

When I pull the deck again, I'll take pics, but I used 1/4 inch copper tubing from CLoad to the output relay. It supports the back end of the vac variable.  Punched a 3/4 inch hole and routed top to bottom of chassis.

Got ya on the ceramic post to unload the rear of the vac variable on CTune. I was wondering about a piece of strap bent into a U to somewhat support the rear of the cap? As stated earlier, CLoad is supported front and rear.  CTune only by 3 stand offs on the front.  That seal carries a bit of weight now that you brought it up. 


YES the air filter mesh is intact.  I am thinking of a new home AC air return folded filter in front of it. Other than as a spectator on a bunch of mountain top sites, I've no experience with this type of operation...  The owner has done some repeaters....  I'm leaving the wherewithal of getting cooling  etc. Up to him.  I've increased the blower size.  More air means more cooling, but more dust too.

Wish the service panel was bigger here.  I've got a hundred amp service with a 90 amp feed to the shop.  I'm moving the meter main to the rear of my shop, which will drop my service drop about a hundred feet.  At the same time, I'm upgrading to 200A service so the house will get the 90 amp breaker instead of the shop lol.  Priorities!

--Shane
KD6VXI

Lesson learned about the braid.

Strap came in today, sheared it down and replaced the braid in the tank.

The piece from tube to CBlock had discolored.. The high impedance node from CBlock to CTune was fine, but from the end of the tank coil to CLoad was also discolored presumably from heat / current.

Still waiting on the new cap for the underside / Tuned input. Hopefully this doesn't change the tank much, I'm happy with the coverage.

--Shane
KD6VXI

Suggestion: move your light source off to one side so the glare from it won't make the meter unreadable

Al

Yeah, I need to experiment with the lighting.

I just installed a bunch of led shop lights and now it's so bright taking pics is hard.  I may put switches in so I can not light certain areas.  I already did over my primary operating area.

--Shane
KD6VXI

I thought he was going with an ANAN 10 for the exciter  and an 8930 as the driver  in close to  or in class A.

Today, this arrived to set up with the amplifier.  As well as the K3S.

I almost prefer the ANAN so as to be able to wrap predistortion around it, but we will see how this combo fairs......

--Shane
KD6VXI

Got the new capacitors for the tuned input from the past commies.  Wow.  Silver plated, welded plates.  I mean, there has to be as much silver on the caps as what I paid for them!

Anyway, created a couple stress tests.  On the RUSH video, about 3:45 or so, I pan around to the front display and the driver.

The mineral oil was BARELY warm to the touch, but honestly after almost ten minutes of CCS music oil boiling, I don't blame it.

For the RUSH fans:  https://youtu.be/bcViGUR6bS8

And if your a Wall of Voodoo fan, live in California, or otherwise understand the joke:  https://youtu.be/3SHU7G4v0aE  

Some people say the recordings are great, others that it's ragged.  It's a ten thousand Watt bird slug.  

And, for those that say PL259s won't handle legal limit ish.....

Anyway......

--Shane
KD6VXI



(I wanted to record this in the sdrplay...  However, even at 250 watts the sdrplay was full signal.  This was with a clip lead shorting the sma to so239 in a loop or with it using  the same clip to short the antenna input.  I also turned the LNA down as far as possible in HDSDR but that didn't do much, still fully to the right.  Any ideas?  I tried a couple times, since the recording on this cell sounds like a CB radio raped a telephone)

Wrap it in foil ???? ??

klc

I wondered.  There is a guy on ebay selling metallic cases.  Before dropping the money, wanted to see if I could get any real whirl input on real shielding.

I mean, it couldn't HURT.   The rx is in plasdick now.

--Shane
KD6VXI

There's an accessory that does a perfect job of sampling high power.  It's a Coaxial Dynamics 7999 sampler. It gives you roughly 40 dB of attenuation which you can further reduce with something like a KAY step attenuator to get the signal just right.  The price is a little under $200 but well worth the investment - you'll have it the rest of your life.  I have a setup where I can with the sampler (in my case a Bird 4273) along with 90 dB of attenuation connect directly to my receiver.

You can buy these used for around $125 at RF Martin Supply, I think

Al,

I get front end overload with the antenna connector shorted.

Would a sampler actually cut that down?

I was looking to purchase one about the time I got an ANAN, but if that will help this problem.....

I've already added common mode chokes to the USB line.  Maybe I need more, as well as an aluminum enclosure?

The pc speakers hooked up to the computer that was feeding the audio had barely discernable rf ingress.  Not bad, they shared the chassis ground.

The speakers on the bench pc I had to turn off.  Those are older Altec Lansing, and the rf immune where 39 bucks at wal mart.  My reason for bringing it up is, there's not THAT much residual rf.  99 cent store 1/8 inch to rca 6 foot cable with little to no shielding, etc and no feedback problem.

The sdrplay was connected to USB only, and the ant input shorted.

--Shane
KD6VXI

Shane,

I use the SDRplay for my station receiver, and you're right.  During TX, it totally gets wiped out.  It's not good for recording. 

What is great for recording is the REA modulation monitor.  In the software, it allows you to record sampled audio from the inline RF pickup into an audio file. 

I'll do my on-air transmitter experimenting for 75 meters around noon so that nobody can hear me and use the mod monitor for audio checks.

For the price, you can't beat it.  Oh, and it gives you your positive and negative peak percentages. 

Sure is a nicely made amp. Compact as well. It's all about the plate transformer there, and no need for band switching.

I have another one to start, going to be my personal amp.  A 3000 for 20 to 80 and an 8877 for 6 to 20, if all works out.  The owner is picking this one up in two weeks.  I can start mine finally!

Just as the maunder minimum is ready to set in.

--Shane
KD6VXI

Foil is so useful. Makes shields, pie plates, even hats! 


Shane, would you be so kind as to share the 0-40V variable bias circuit? All I can think of is a pass transistor as a bias regulator or a very stout DC supply in series with the cathode.

I can, but it's basically W4ZT bias board. I also have 15 10A10 diodes that I can switch in the cathode return line.  The variable bias is also switchable.

When using the variable, I can go from 4 to 30 some volts.  The additional 10A10s take care of the rest.

The bias circuit has the ability to add additional in.  The bias board was an afterthought, so I left it as is.  

Taking son to Karate, I'll post the schematic when I get home.

--Shane
KD6VXI

Bah, have the Internet on the phone.

http://bias.gs35b.com/docs/060923.html

Link to entire page describing board, etc.

I've used the W4ZT method also... really liked it on a single 3-500.. I have a stockpile of 50w zeners though, so I usually use the quickest method.

Thank you!

Patrick,  no problem.   

Jon,

I've ordered a metal case, purpose punched for the sdrplay.  I'll let ya know what, if any, decrease in stray rf I get.

I want to do some screen shots of the band with, but I can't see it yet myself!  Lol

--Shane
KD6VXI

Shane, have you tried it in class C? since you got that much bias. Anyway after investigation I think a different circuit may be better for me and allow keying via cutoff if the HV transients can be whipped. I'm still impressed with how nice your amp looks.

Pat,

Thanks for the compliment on the amp. 

I have rolled the bias back as far as I can, but on the 6.4 kv DC tap, I still pull some plate current.  I'm going to go back to the 4.2kv DC tap and then I can get it in cut off.  On the high tap, I can only bring it do about 100 mils, iirc.

I have a chart I started at home of operating parameters vs bias.  It sure makes a hell of a difference in efficiency as you roll more bias on!

It also has a 10kohm cutoff resistor in the cathode circuit for cutoff on rx.

--Shane
KD6VXI