GLOBE KING 500B AUDIO

[W8UJX]

I am interested in improving the audio in my Globe King 500B, both frequency response and distortion.

I have done away with the couplates and have bypassed the clipper stage.

I would like to add some negative feedback but am not sure how to do that.  Either NF from the plate of the 6L6B back to the cathode of the 6SJ7 first audio amp stage or from the modulated B+ to the 4-400A back to the cathode of the 6SJ7.

How much feedback should be used and what would the circuit look like?

Would just removing the cathode bypass caps from the cathodes of the first and second audio amp stages be sufficient?

The grid resistor in the first audio amp is only 2.2M, to small for my D104 so I am going to increase that to 6.8M

Thanks,
Jerry, W8UJX   

[WD5JKO]

Jerry,

  There will be plenty of opinions here on which way to go.

The weak link as I see it in that modulator is the 6L6 driver has a high source impedance such that as the 811's are driven harder, the positive drive voltage into each grid will progressively compress since the grid current will increase with increased grid voltage.

The usual way to drive class B modulators is to have a low impedance driver using low Mu triode(s) where push pull is preferred over single ended. Look in the various handbooks, and you will see lots of examples of 2A3, 6A3, or 6B4 push pull triode driver stages for a pair of 811's in class B push pull.

Then go to Bama, and pull the schematic for the Globe King 400:

http://bama.edebris.com/manuals/wrl/king400/

Pick file "f"

Here you will see a 6N7 phase inverter driving a pair of 6F6's triode connected in push pull. A triode connected 6F6 is almost as good as a 6B4, and those modulators sound excellent when 100% stock.

The good news is that on a Globe King 500, they used the same interstage transformer as the 400. They just as I recall, floated the center tap on the primary side.

If you opt to keep that single 6L6, and use negative feedback instead, the results might sound better up to 50% modulation or so, but adding NFB does not increase the drive power available to drive the 811 grids.

Also any NFB crossing transformer boundaries should be limited to no more then 3 db gain reduction for each transformer in line. This is not a hard rule, but realize that across the audio response, say < 100 Hz, and > 4 khz, there will be appreciable phase shift through those transformers, and instability is likely with too much NFB.

If you decide to add more bass, then realize that the power supply capacitors for each audio stage will need to be increased..and that includes the big oil cap for the HV. Then the PTT B+ switching might be hard on those 816 rectifiers..adding the need for a step-start turn on.

Jim
WD5JKO

[W8UJX]

Thanks Jim,

One thing I forgot to mention is that I converted the power supply in the modulator to cap input.  Added a 10 uF 2 KV filter cap.  B+ on the 811A's is 1500 VDC.  Added a bias circuit to supply -4.5 VDC to the grids of the 811a's.

I will definitely check out the circuits in the GK400.  I have built a lot of CW rigs but I have no experience where negative feed back in AF amps is concerned.

Let me bounce this idea off you.  I have a 20 watt HI-FI Heathkit amp that I could use to drive the 811A's.  Instead of using a transformer between the Heath amp and the 811A's what about using capacitive coupling between the plates of the 6L6 output tubes in the amp and the grids of the 811A's.

This may be a totally dumb idea but I won't know unless I ask.  Hi...

Jerry, W8UJX

[WD5JKO]

Jerry,

   Look at the WA5BXO audio driver details. Good stuff. This uses an external Hi-Fi amp as you suggested, and then uses a pair of NPN horizontal output transistors to transform the impedance down to drive those modulator grids.

http://www.qsl.net/wa5bxo/driver1.html

Jim
WD5JKO

[W8ACR]

I agree that NFB is needed on a single ended 6L6 driver. NFB is probably not necessary if you use PP low mu triode drivers such as 2A3's, 6A3's, or 6B4's. If it were me, I'd do one of three things:

Option #1 - I have scanned a schematic from the 1947 Radio Handbook showing a single ended 6L6 with very simple NFB. I have used this speech amp in the past to drive zero bias 811A's at 1000VDC and it worked FB, although not Hi Fi. The driver transformer was a UTC S9. This would be the simplest option, and your audio would be A-OK.

Option #2 - IIRC, both the Globe King 400 and the Globe Champ 175 used triode connected PP 6F6 drivers. They should be able to drive 811A's and they do work well. Although I no longer have the Globe champ 175, I always got rave reviews about the audio which was stock. If you could clone the Globe Champ 175 speech amp, I think you'd have great audio. You would have to find a place to put another tube on the chassis, but it should be do-able.

Option #3 - My personal favorite speech amp is shown in the second schematic. 6N7 phase inverter capacitor coupled to 6A3/6B4 drivers. Again, a UTC S9 driver transformer will work. I have built this speech amp three different times and it has always worked very well. The only transformer is the driver transformer as you can see. Again, you would need to find a spot for another tube. This is basically the same circuit as the Globe Champ 175 speech amp, but using real audio triodes instead of triode connected 6F6's. I am using this circuit in my latest homebrew rig. The microphone is a Heil Classic running through a W2IHY mic preamp. The audio gain pot is only about 1/10th open to fully drive the 811A modulators.

The nice thing about 6F6's is that there are plenty of them around and they are 1/2 the price (or less) of the 2A3/6A3/6B4 family of tubes.

The final picture is my latest homebrew version of the 6A3 speech amp. The transformers are the S9 driver and the filament transformer for the 6A3's. The DC power supply is on another deck.

Ron W8ACR

[W8UJX]

Thanks for the replies guys.  One problem I have is I have no audio transformers or I would have built an amp such as Ron suggested.  I see these transformers on EPay for about $90.

I think I will give the WA5BXO driver that Jim suggested a try.  That way there will be minimal wiring and component changes to the GK.  Also, the parts are readily available.

I do have some questions for Jim.

I converted the HV supply in the GK modulator to cap input.  Removed the filter choke and put a 10 uF 2 KV cap in its place.  This gives 1500 VDC on the 811A’s.  That requires -4.5 VDC bias for the tubes.  I rectified and filtered the 6.3 VAC on the 6L6 filaments to get this bias voltage.  There is no step start so I am a little concerned about the life of the 816’s

Question:  I wonder if I should leave this mod in place or go back to the original 1000 VDC on the 811A’s?

I notice a note on the schematic of the WA5BXO driver that specifies 300 VDC max on the plates of the output tubes in the speech amp.  My Heathkit HI-FI amp has 435 VDC on the plates of the 5881 push pull output tubes.  At 10 watts RMS output into an 8 ohm restive load the P/P voltage on the plates of the output tubes in the amp is over 450 VDC.
The amp is capable of much more output than this.  It is a 20 watt amp.  I’m not sure how much power is required to drive the 811A’s.

Question:  If the input voltage to the WA5BXO driver goes over 300 V P/P will this blow the transistors in the driver?

Jerry, W8UJX

 

[W8UJX]

Jim-JKO,

One other question.  I wonder which would be better to use as the input transformer in the WA5BXO Driver circuit.  The existing 811A grid transformer in the GK with the primary disconnected or a 35 watt output transformer I have here.

I bought this transformer on EPay.  New Sensor Corporation P/N NSCO22842-T.  Specified as 35 watt tube output transformer for guitar amp suitable for 6550, KT88 etc.  No other information given.

I have a few other small output transformers in the junk box but I figure this is a good starting point and nothing should be critical.

Jerry, W8UJX

[W8ACR]

If you change your mind, I can get you a good driver transformer cheap!

Ron

[WD5JKO]

I converted the HV supply in the GK modulator to cap input.  Removed the filter choke and put a 10 uF 2 KV cap in its place.  This gives 1500 VDC on the 811A’s.  That requires -4.5 VDC bias for the tubes.  I rectified and filtered the 6.3 VAC on the 6L6 filaments to get this bias voltage.  There is no step start so I am a little concerned about the life of the 816’s

Question:  I wonder if I should leave this mod in place or go back to the original 1000 VDC on the 811A’s?

I notice a note on the schematic of the WA5BXO driver that specifies 300 VDC max on the plates of the output tubes in the speech amp.  My Heathkit HI-FI amp has 435 VDC on the plates of the 5881 push pull output tubes.  At 10 watts RMS output into an 8 ohm restive load the P/P voltage on the plates of the output tubes in the amp is over 450 VDC.
The amp is capable of much more output than this.  It is a 20 watt amp.  I’m not sure how much power is required to drive the 811A’s.

Jerry,

   You need to ask yourself whether you want to take a chain saw to your Globe King, or do you want to just do some relatively easy mods. Something in-between the extremes would be to revert to the Globe King 400 circuit which did a commendable job with triode connected 6F6 drivers. On the easier side, the 6L6 driver with negative feedback as posted by Ron W8ACR would be an improvement. I did something similar once to a King using an EL-34 instead of the 6L6, and with some NFB, it worked fairly well, and seemed to provide a bit more power then the 6L6 did.

  I would be concerned with those 816's hot switching a CAP input filter. Also the extra modulation power, although nice, might lead to a flash over somewhere on an audio peak. Using a step-start relay would help, as would a solid state conversion to replace the 816's with something like four or five 1N5408's for each 816.

"Question:  If the input voltage to the WA5BXO driver goes over 300 V P/P will this blow the transistors in the driver?"

I am not sure. Lowering the B+ in the amplifier might be a good idea.

One other question.  I wonder which would be better to use as the input transformer in the WA5BXO Driver circuit.  The existing 811A grid transformer in the GK with the primary disconnected or a 35 watt output transformer I have here.

I bought this transformer on EPay.  New Sensor Corporation P/N NSCO22842-T.  Specified as 35 watt tube output transformer for guitar amp suitable for 6550, KT88 etc.  No other information given.

I have a few other small output transformers in the junk box but I figure this is a good starting point and nothing should be critical.

  I would keep what you have and try that first. Remember that stock GK 500 interstage transformer might have an unused CT. tucked away. If so it is likely the same transformer used on the GK 275/400 where there was a pair of 6F6 triode connected drivers.

  Remember that the WA5BXO circuit takes very little power, and has a voltage gain of just under unity (~ 0.9), typical of emitter followers. The output impedance is low, so if the transistor has a gain of 100 at 200ma collector-emitter current, then the base current would be 2 ma. This might correspond to an 811 grid driven positive to 100v where the grid current could peak at 200ma or so.

  In that circuit is a +/- 300V supply using off line voltage doublers. I would consider transformer isolation, and for 811 modulators, I bet +/- 150V would be fine. Then again, doing this would also require further lowering of the B+ on the audio driver tube amp.

  Jim
WD5JKO

[W8UJX]

Cheap???  Did I hear someone say Cheap?  I'm retired, partially disabled, on a fixed income and love to tinker so Cheap is my middle name.  Tell me more Ron, Please...

I have a lot of stuff in the junk box including KW Power Transformers and a 300 watt Thordarson Mod Xmfr but I have no Good Interstage Transformers.  I would like to acquire several good audio Interstage Transformers to play around with so I can build some of these speech amps I see.

I want to end up with something kinda semi HI-HI.  Flat from 100 CPS to 5 KC or so.  Better if possible.


Jim, I want to preserve the GK as much as possible.  If I can find some Interstage Transformers the 400A circuit along with the WA5BXO driver are some that I want to try.  Build them up on an external chassis to minimize changes to the King.

I have several of the K2AW 10 KV rectifier modules so I could replace the 816’s with those.

Jerry, W8UJX

[AB2EZ]

Attached is a proposed design that is similar to the WA5BXO design, but which employs low cost FET's, and does not require a driver transformer. Note that these FETs have to be attached to a heat sink using insulating pads.

It is a variation of the design I have used successfully with my Johnson Ranger to greatly improve the audio by eliminating the Ranger's driver transformer. Ref (including photos): http://mysite.verizon.net/sdp2/id13.html

The circuit should be more than capable of producing the 185V peak AF grid-to-grid signal required to drive the grids of the 811A modulators (as per the 811A specification: http://www.mif.pg.gda.pl/homepages/frank/sheets/049/8/811A.pdf). The two bias inputs can be used to separately adjust the grid bias of each tube. The required voltage at each of these bias inputs, for biasing the 811A grids at 0V, will be a few volts positive (with respect to ground) ... to make up for the gate-to-source voltage drop of each FET.

Stu

[W8UJX]

Tks Stu,

I have not ordered the parts yet so I think I will order parts for both designs and see which one works best for me.

I have increased the B+ on the 811A's to 1500 VDC which will require a grid bias of -4.5 VDC.

Jerry, W8UJX

[WD5JKO]

My Globe King 500B story began at the Manchaca Fire Hall in Manchaca, Texas. This was about 1979, or maybe 1980. It was a rainy morning as the hams went about their way under the covered roof with flash lights in hand trying to find that bargain before the sun rises. I was there too, about 24 years of age. I spotted a pick up truck with a camper shell driving by with something familiar looking inside it. Going closer, it was a Globe King 500B with a price tag on it for $20!! Standing in the pouring down rain, I bought the thing, and somehow got the pieces one at a time to my car. On the way, the modulator deck sure was light, and looking at it, it was stripped of the power and modulation transformers.....

This rig started a long learning process and reconstruction. My youthful exuberance brought me extreme highs followed by extreme lows. I learned a lot. I will post the 30+ year old hand drawn schematic here. I would not say this is a perfect and bullet proof design. Quite the contrary, but it did have its attributes using a technique that I have never seen done before, or after, and maybe for good reason.  

I wanted to drive them 811 grids with a zero ohm source impedance solid state driver. I remember having a lit 808 tube on the bench, and driving the grid with a 200V P-P sine wave until the grid turned red like a toaster filament. The grid current was peaking at 1 ampere @ 100V. Looking at the waveform, there was no difference if I turned the filament off...That was my goal, zero ohm source impedance..at all costs.

The modulator was all solid state except for the 811's, and the +150V regulator tube. It did a decent job, and as best I could tell, it was Hi-Fi for the day, and it really got every last watt of the 811's that might have been lost before with a higher impedance drive setup.

So I achieved my goals, and operated this rig for several years on 10M AM back in the early 1980's. Then it all came crashing down one day when I did something very stupid. I got a 1/2 dozen 811's from a swap meet, and I tried them one at a time in my GK 500B. One of the 811's had the grid touching the plate. I'll never forget the meter pegging fire show, the flashes, and smoke that followed. Every transistor in that modulator was now fused silicon, and even the FET in the condenser microphone was shot.

So my experience should suggest to those that might want to build a direct coupled SS driver similar to what I did, to have a active clamping protection circuit in place such that a tube failure does not cause a catastrophic domino falling disaster.  

After this episode, I went to a pair of 6B4's working with the original GK-500B interstage transformer. I just wasn't up to rebuilding that SS driver again.  

Oh, those HV JFETS used, 2N6449/A5T6449 are still available:

http://www.interfet.com/Datasheet/2N6449/

These are depletion mode, and bias up similar to a 12AU7 tube.

Jim
WD5JKO

[W8UJX]

Hi Jim,

Interesting fireworks story.  I like those hand drawn schematics.  I may experiment with some of those circuits if I can find the parts.   

I bought my GK 500B in 1958 when I was 21 years old.  A year later I joined the Air Force for a four year tour of duty in Spain.  Couldn’t take the King with me so I left it in my Grandmaw’s house.  After the Air Force I worked as a Tech Rep for Philco Corporation.  Every year I moved to a new contract location.  In 1965 I came to CA.  When Philco told me they were sending me to Tehran, Iran I said, “No Ur Not”.  I stayed in Ca and still live here.  Winter WX is better than Ohio.

I dragged that GK around with me every place I went.  In 1975 it started falling apart.  First the HV filter chokes in the PS and Modulator shorted out.  When the HV power transformer shorted out I decided to sell it.  I’m sure most every one that has one will agree component ratings were an issue.  I wish I had kept the modulation tranny.  I imagine those are pure unobtainium now.

Now that I am retired and disabled I decided to buy a King and get back on AM hence the interest in improving the audio.  I wanted a 500B but could not find one.  I bought a 500A from a guy in Oregon but really wanted the VFO in the power supply.

Later I bought a 500C and combined the 500A RF and modulator decks with the 500C power supply.  I like the black meters and the modulator on the 500A better than the B or C.

I want to play with both tube and transistor speech amps but I don’t have any Interstage transformers or knowledge of solid state circuits.  You guys have given me some ideas on transistor circuits and maybe W8ACR can supply a transformer or two???

A note – The modulator in the A version does not have the useless limiter circuit and uses four tubes in the speech amp so I have an extra 8-pin socket over the B model for use in a new design. 

The primary of the Interstage transformer has an unused lead to use in a P/P 811A driver.   

Jerry, W8UJX     

[WD5JKO]

Jerry,

  I received a private email from John, WA5BXO, and he gave me permission to post his reply here.

From WA5BXO:

The circuit that I put together probably has a number of flaws in smoke prevention, as W5OMR will testify.  The idea is sound but it is essential that the input is not over driven, so although the input could be from a low z source to the low Z of the input XFMR it would be quite easy to over drive and of course the fidelity would not be as good anyway.  Capacitor coupling from a amp plate, that has plate supply of about 200 volts is a better option.  I had a Bogan amp with about 450 volts on the plates using capacitor only filtering in its power supply.  I put in a choke input filter, reducing the power supply voltage drastically.  Since this required a re-biasing of the Bogan’s output tubes I decided to make sure they were in class A.  I used the amp as a voltage amp only (no load on the output winding of the Bogan output transformer.  I also changed the input to these tubes by making a resistor dividing network that took a sample from each plate of the output tubes (capacitor blocking the DC of cource)  and fed this back to the respective tubes grids with a resistor in series from the triode drive tube circuit so as not to change the driver load to drastically.  I played with the values of this inverse feed back until I got about 50% gain the result was a very clean output with a wider bandwidth.  I did this all to compensate for the inductance of the output XFMR as it would change the audio had I not.  The ECG177 diodes are fast diodes that will turn on quickly should the input go more negative than the voltage on the emitter supply thereby protecting the EB junction from instantaneous peak reverse breakdown bias.  The extra negative voltage would just be fed to the grid of the class B modulator helping to keep it turned off.  Remember, that particular modulator tubes plate will be going higher in voltage as the Grid is going negative and it should be kept from conduction for proper class B operation during this half cycle.  There is no protection from Base to Collector however.  Therefore unless someone designs a limiting input circuit the plate supply of the driver tubes should be less than the supply for the Xransistor.  In class A the driver tubes will not produce a peak voltage higher than there DC supply  The ECG 164 can handle more supply voltage if necessary.  Be careful though.  I think its CE volt rating is less than its CB rating.   I suspect a limiting circuit could be designed for the input if someone, like JIM, wanted to put it to the test. HIHI.
            A few side notes:
            I originally built this circuit using ECG 198 transistors.  With a 50 volt Collector supply and a 150 volt emitter supply but all it did was drive 813s in AB2 (not a lot of requirement there).  Class B triodes are a lot more challenging.  Also remember that when the modulator HV supply is off the control grids will take the brunt of the unlimited current of the driver transistors.  The driver Xsistors will draw current as is demanded by the load.  The grids will demand a lot when there is no HV supply.  So be sure to kill the audio when not modulating.  There is some small 100 ohm resistors the collectors of the Xsistors to help limit this current but they will burn out.  The ECG 198s and ECG 164 may be hard to find these days and there is a lot of HV MOS FETs that could be made to go in place."

  So I attach a file to an adjustable current limiter that might be applicable to limiting the current to the transistor collectors (or FET Drains). These are carried by Digikey for a few bucks.

I liked your story about the GK following you all around the country in your younger years. How many times did you lift those decks up/down steps, etc. ?

The very true, and not so funny thing is my 6B4 audio driver seemed to get me better audio reports then that solid state driver.

Jim
WD5JKO

[W8UJX]

Hi Jim,

When I was a kid we lived in an old two story farm house and I can remember carrying those GK modules up and down the stairs until I finally settled on a large upstairs bedroom for location of the shack.  Now at 75 it is all I can do to pick up the power supply section and carry it a few feet.  Old age has really sapped my strength. 

Thanks for the additional info.  All of my experience has been with vacuum tube circuits.  For this reason I am leaning more in the direction of the 6B4-G speech amp.  This is a simple circuit which seems easy to build and make operational.   I did an inventory of the parts in the junk box and I have all the parts except the 6B4’s and the interstage transformer. 

I about had a heart attack when I checked the prices of the tubes and transformer on Epay.  I’m looking at between $250 and $300 for three parts.  I had no idea those parts were that expensive.

I wonder if you or anyone else knows of a cheaper source for these parts.  There is a ham radio flea market in San Jose this Saturday but I don’t hold out much hope finding these parts.

Jerry, W8UJX   

[K5UJ]

6B4s have been driven sky high by audiophools.

Mike at hamtubes.com wants $35 each  He is a good guy and may make a deal.  he's at a lot of the hamfests around here (Chicago) call or email him.

[WD5JKO]

Look at this:
http://www.tubelab.com/6AV5.htm

And then check here:
http://www.abcvacuumtubes.com/vacuumtubes_tubelist.html

[W8UJX]

Thanks Rob and Jim. 

I got some tubes and a transformer from RON, W8ACR.  I will also check out the info you gave me.

Jerry, W8UJX