Carrier Shift, FMing, SSbing... Warble.. on AM Transmitters.

[Jim, W5JO]

What about those little glass tube components?  I think they are small h.v. capacitors?   I've wondered about them.  They're the little glass tubes with wire on each end and some kind of fluid dielectric?  I think they were used for h.v. 60 years ago because they were small but I'm not sure about that.  There are a bunch of those in my 755 VFO so I figure they are probably in the 300A VFO also.  

Rob

Rob those should be low value trimmer caps if you are referring to the ones with the screwdriver adjustment on one end used to match the output to the next stage.  I have never seen any glass cap with fluid in it put in a Champ or WRL product.  This is not to say they weren't or were subbed in for a repair of some kind.  There are others that look like small glass tubes with wires on both ends and no adjustment that are ceramic caps used to couple and trim the VFO.  Some of these are temperature compensating caps used to stablize the frequency.

Clark, mechanical stability and clean contacts are very important, so do as suggested.  Clean the piviot points in the variable cap and hit all the solder joints with an iron should one or more of them has crystalized.  That cover is soldered down to help prevent  coupling and for mechanical stability.  If things are right in that vfo and voltage supply circuits, you won't need any extra caps  or components

[WD5JKO]

Clark,

   You have a very common problem from rigs of the same era. Some are worse than others. Getting a tube oscillator to key cleanly without drift after being off for long periods certainly is a challenge. Others have already provided excellent input to consider. Yes, adding more filter capacitance across a gas type VR tube can make a relaxation oscillator. This is not the case if the regulator is a HV Zener diode. That said the VR tube regulator (like OB2/OA2) will be much more stable than any HV zener or zener string. HV zeners all drift upward as they heat up. So if you need better filtering to eliminate hum modulation/FM'ing, a well bypassed zener is a good choice. If you need HV stability at the VFO screen grid, then the VR tube is hard to beat.

   I have wrestled with this many times, and what I once ended up doing with an old Heath VF-1, was to regulate everything. That included the plate, screen, and filament voltage. Another thing I did was to keep the oscillator running all the time. I did this by adding a relay to switch a capacitor across the L-C tank to shift the oscillator downward about 50 Khz while I am receiving. When transmitting the relay opens and the oscillator returns to normal frequency. The idea here is to keep that VFO tube oscillating, and to keep the L-C components warmed up from RF current. Then with constant filament voltage, and screen and plate B+, drift is minimal.

   If you make the changes I propose, you might still have some FM'ing if the VFO sees a variable load when you modulate.
Regulating the element voltages on the next stage 6CL6 might eliminate this effect.

   Instead of hacking up your Globe Champ, you might just keep it stock, have fun, and be proud of owning an SBE rig. 

here is an example of a simple filament regulator:
http://pages.prodigy.net/jcandela/Band%20Hopper/Voltage_Doubler.JPG

Oh, switching the VFO tube from a 6au6 to a 6bh6 cuts the filament current from 0.3A to 0.15A. That is less heat in the VFO compartment, and if you use a 7806 filament regulator, it will be a lot cooler with a 6bh6.

Regards,
Jim
WD5JKO
   You might

[N2DTS]

Yes, a large cap is no good, but you can use a small value without problems.

The big thing is to make sure the current through the regulator tube is correct, and has not decreased because the dropping resistor changed value.
With anything made by wrl, I always suspect they used the minimum they could get away with, sometimes even less!

Brett

[KM1H]

There will be no way to satisfy everyone with a fundamental frequency VFO except for maybe a Collins PTO such as in a 310B exciter.

I wouldnt get involved with making cap changes unless you have a counter, lots of caps and even more time. Even then you will need to SS regulate the DC and filament voltage.

An easier method is use an external hetrodyne VFO and disable either the xtal oscillator or the RF output in TX.

Ive also heard that the WRL 755A VFO is very stable but havent used one. The Heath HG-10B is another good one. They would have to run continuously and leakage would have a noticable signal in the receiver. A decent relay or diode switching in the output line can drop that down into the noise.

Carl
KM1H

[w3jn]

You do NOT want to put an electrolytic after the voltage regulator.  This will make a nice relaxation oscillator and make things a LOT worse 

Johnny,
           I was thinkin about that, but you beat me to posting it. I did that some years ago and it really went to town! ! Depending on the size of the cap, you could actually watch the VR tube pulse on and off.  

Heh, been there, done that with a hammy hamboned HRO60.  Someone tacked a 20 uF cap in there and the VR tube was merrily oscillating away at about a 2 Hz rate.

[Rob K2CU]

I would stay with the 0A2, and check the series resistor, especially if it is carbon comp. The 0A2 has a dynamic impedance of about 80 Ohms, where a comparable 150V 5W zerner runs over 300 Ohms.

The difference?  In a typical circuit, you have, say 300 V for your main plate supply in the rig. The 0A2 is running 20mA of current and the load is 20mA. The dropping resistor is then 3750. IF you have 10V RMS of ripple on the 300 V, it becomes 200 mV of ripple at the load. Changing to a Zener with 300 Ohms dynamic resistance would then give you 700mV of ripple at the load. You would have to add a 33 uF cap across the load to reduce the zener ripple to that of the 0A2 without a cap.

[K5UJ]

There are others that look like small glass tubes with wires on both ends and no adjustment that are ceramic caps used to couple and trim the VFO.  Some of these are temperature compensating caps used to stablize the frequency.

Okay Jim those must be the ones.  I have seen them for sale in those Radio Shack catalogs on-line, the ones that go back to the early 1950s but I forget the brand name.  "Glass"-something.   Yes, my 755 has several in it.  I assumed they are okay and do not need to be replaced because they are sealed and, unlike an old electrolytic, the dielectric has not dried out.  I'd be surprised if they are not found in other WRL VFOs so I figured it was on topic here.  Thanks for ur answer.   If I had a camera that could take clear closeups I'd add a photo of one.

73
Rob

[DMOD]

I am placing the Caps BEFORE the regulator on the VFO plug under the chassis.  That should be just fine right?

Yes. On your posted schemtic C7 IS before the gas VR and before the dropping resistor R4.

Phil -AC0OB

[w1vtp]

<snip>
Ive also heard that the WRL 755A VFO is very stable but havent used one. The Heath HG-10B is another good one. They would have to run continuously and leakage would have a noticable signal in the receiver. A decent relay or diode switching in the output line can drop that down into the noise.

Carl
KM1H

Carl: I noticed that my Eico 722 is very good in the leakage dept.  I cannot hear a thing unless I attempt to couple the beastie to the outside world.  What did you have in mind -- short the output or switch it to a load?  Since I'm going to elect... switch my Tx / Rx anyway, perhaps I should go the diode switch.  Typically, the 722 has a short freq drift when going from Rx to Tx but settles down after that

Al

[KM1H]

Al, it would depend on the VFO to a large degree. If it doesnt have a buffer stage then you have to see the effect of an open, short or load especially if its a yooper at hitting the PTT like the Champ.

Carl
KM1H

[ke7trp]

I ordered the caps.. Will get them tomorrow.  I got the sag down to 2 volts.  Its alot better. It was dropping 6 volts on the intput of that gas tube! A 100UF will probably take most of this away.  We tried for 2 hours to get that Copper plate off.  We tried a 260 watt Trigger iron.. It would not do it.. To much of a heat sink.  We even tried 2 irons at once. 

I am going to have to buy a small Torch at radio shack.  I suspect the job will be easy with the torch.  This has never been apart.  Those tubes have to be toast after all these years. I am going to just wait till I get the cover off and do the job right. New tubes, Tube Dropping resistor, New caps. Also going to stiffen the Main PS.  Because this transmitter has so much audio, The current really soars when you modulate it. This realy sags all the rest of the supplys. 

I am going to do all of the above and stop there. I dont want to hack this radio up.  They all do it. I just talked to a guy on his Champ and it FMs as bad as mine.   I just want to perform a few simple things to get the this a little better.

HAMs are like a pool of sharks.  When one smells blood, they all come running.  One one guy on his fancy $5000 radio says "Hey, Your FMing!" the rest of them chime in "yeah!!! Lets get him!!!"  I will just tell him to piss off.  Vintage AM is supposed to be fun.  Not all Radios are perfection.. I love my old Champ. Its like an old Friend!



Clark

[K1JJ]

HAMs are like a pool of sharks.  When one smells blood, they all come running.  One one guy on his fancy $5000 radio says "Hey, Your FMing!" the rest of them chime in "yeah!!! Lets get him!!!"  I will just tell him to piss off. 
Clark

Sounds like you're making progress.  After doing the stuff you mentioned, get some NPO caps and play around with them as a final touch.

Really, there's no reason to have to live with a VFO that moves around, vintage or not. That is one mod that is worthwhile and doesn't matter if it's "factory" stock or not.

Yep, funny about the ham reports. I notice that trend too. If you mention you might have some distortion and want a report, suddenly everyone starts to hear distortion and that's all you hear about for a while. Then someone mentions it is better and they all jump on that even more. I think AMers like to report good news more than bad anyway...   But we are truly herding animals.

[ke7trp]

I agree. I am going to fix it..   Funny comments about the Sharks..  I never ask anyone how I sound. I wait for a few key individuals and ask them.  Everyone has a different opinion and is using a different RXer and speaker.

I can remember one night on am where a guy bashed my Audio for 30 minutes.. Over and over.  Finally someone keyed up and said 'Turn your damn noise blanker off so we dont have to hear you bitch anymore".............Long pause..   "guys, Real sorry, "My blanker somehow got turned on"  My loud audio was pounding in there and his noise blanker was really playing hell on his RXer.

Sometimes its best to just not mention things like this.  I try to invite everyone to talk.  When it turns to arguments and the pool of sharks, I just turn the set off.

C

[K5UJ]

Just one mod suggestion Clark (although you probably plan to do this already):  drill a few holes, get some nuts, bolts and star washers and fix the cover so it bolts on instead of soldering.  Hope it has a flange for you to drill through.

Rob

[ke7trp]

Its just SIX tabs.. I wonder why the soldered this thing instead of bolting it down?  I was just going to torch it back.

C

[K5UJ]

Its just SIX tabs.. I wonder why the soldered this thing instead of bolting it down?  I was just going to torch it back.

C

my guess is with the right equipment on an assy line it was faster to solder it down back then.

[ke7trp]

Ok.. I am stumped here fellas..  This copper cover on the VFO will not come off.  It looks like its soldered down to the chassis. 

I have tried Two pencil irons at once,  A 260 watt gun and now a BLOW TORCH!   No go..

I held the small handheld torch on the Solder joint for 5 full minutes..  I still cannot scrape the old solder with a tip of a screw driver.  I had to stop as the copper plate was so hot it was burning me through a towel.. I am afraid I will damage caps and resistors within. 

DIRECT heat from the Blue tip of a torch will NOT unsolder this damn cover..  I really hate to cut it off!

Advice???

[K5UJ]

I don't own a 300A and have never been inside one but experience tells me that when it seems near impossible to get into something that will need to be serviced eventually (there's a tube in there after all) then there is probably another way that's being overlooked because it isn't obvious.  I can't believe they'd make this thing in a welded shut container.   Maybe something slides off?   This was way before the days of making something that gets tossed out when it stops working so surely it opens up somehow.

[ke7trp]

Nothing slides off. Its soldered shut. It must be silver solder or some other type of realy high temp material.  I just heard from a guy that said he cut the tabs and made L brackets to hold it back down.  What a mess.  I just want to service the VFO..  haha.

C

[ke7trp]

For those of you that have a GLOBE Champ, 300, 300A, 350 ect..   Add a 100 UF 300 volt cap BEFORE the VFO regulator tube. On WRL gear the VFO is wired in with a 4 PIN plug.  On a Champ 300 this is on the BOTTOM and PIN 4.  Plus of Cap to Pin 4, Minus of cap to any near by Ground lug. 

This will totaly solve the carrier shift, FMing ect.  With the addition of this single cap, The Radio is clear on AM and SSB modes on the reciever at full modulation.

I suggest running this rig on a VARIAC at 115 volts.  If you run it on 125 volt residential power, Your voltages are way up inside the rig.  This is troublesome to the VFO (250 volts turned into nearly 290 on mine) and the main power supply cap. This cap is 1000 volt oil can. I tested 1150 volts across this cap! Filiments are running at 7.5 volts instead of 6.3.  Use 866s to keep the voltage Drop and use a variac.

Clark

[WD5JKO]

Add a 100 UF 300 volt cap BEFORE the VFO regulator tube. This will totaly solve the carrier shift, FMing ect.  With the addition of this single cap, The Radio is clear on AM and SSB modes on the reciever at full modulation.
Clark

   Clark,

  What you propose is a great modification! Maybe go a step further, and put a switch in series with that capacitor. Label the switch on your front panel, 'SBE' on one side, and 'AM' on the other.

   If you want to do it, maybe add a 10K resistor across the switch to pre-charge the cap. That way when you throw the switch, no major sparking should occur across the switch contacts.

Great stuff!
Jim
WD5JKO

[K1JJ]

Good going, Clark.

So, what can we learn from this, technically?  As usual, just taking simple voltage measurements with a VOM while the rig is in operation would have shown the voltage regulation problem. (And that's what you did)    Dynamically changing parameters from the norm, in their many forms, is what identifies most problems.

As for the cover of the VFO, I would cut the top off with a Dremel tool cutting disk and install a tiny set of hinges for future access.

T

[KM1H]

A bucking transformer is usually cheaper than a variac. Find an old 10-13V transformer with a secondary rating at least 1.5X the total current of the rig.

I have a couple of surplus shop big ones of 15-20A that feed each bench so everything runs at 110-113V. Nice and cool, in spec, and no more blown pilot lamps. The big plate transformers for amps and mods run off their own feed.

Carl
KM1H

[Jim, W5JO]

That VFO cover is soldered to the chassis of the VFO to give mechanical stability and, more importantly in 1955, to help with TVI.  Invest in a Weller 80-125 watt soldering iron to get it up.  I took mine off to replace the VFO tube and then drilled a small hole on the lip for a # 4 metal screw. 

You might want to be careful adding to many big caps to the voltages busses.  That power transformer is the weak link in the transmitter.  More than one have gone bad because of caps on the buss that supplies HV to the VFO develop a lot of leakage or short and the ones on the bias winding can do the same.  Of the four Champs I have seen, only one has the original transformer in it.  I do suggest you walk softly with it.  The problems that occur are current related not voltage.