The AM Forum
March 28, 2024, 08:39:12 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Stabilizing VFOs  (Read 4477 times)
0 Members and 1 Guest are viewing this topic.
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« on: December 26, 2011, 10:56:02 PM »

You want a VFO challenge? Try PSK31.

I attempted to use a vintage Hallicrafters SR-150 on PSK31 about a week ago to QSO with a buddy in Massachusetts and caused quite a disturbance. Of course I let her warm up for about 30 minutes before trying. I knew to keep my drive down so I was seeing about 10 watts on the meter so as not to get into ALC. The waterfall display was fairly stable but when I went to TX and then came back to receive, the station was far out of the passband and it took about 10 seconds for the decoder to lock even with AFC. It was worse for the poor soul on the other end. He had to basically attempt to find and track me till I settled in. This is how not to make friends on the band. Now to make a signal the Halli has a VFO and two crystal oscillators in a classic single conversion scheme.

Measurement with my counter revealed that the VFO was slewing more than 100 Hz between TX and RX for the first 10 seconds. Now 100 Hz might not be an issue for AM it is and is probably not a big deal for SSB, but for PSK31 where the signal is only 30 or 40 Hz wide, it is a show stopper. 

The VFO is the pentode section of a 6EA8 and it is regulated off a 150V gas tube.

The Buffer is the Triode section and it is on the unregulated 250VDC.

Most of my problem was traced to the VFO. It was not a pushing problem - varying the load was not the culprit.

All of my mods are ridiculous but reversible and they are all good VFO stuff:

First the Hallicrafters SR-150 was some kind of a milestone rig because it is the first rig to sport a RIT function with a proper varactor diode and pot control. It also uses a diode for frequency correction. I deemed these to be worthy of disconnecting and i got some immediate improvement. Don't hang stuff off the VFO like relays and switches and such.

Then I decided to try regulation on the plates of the buffer and the two XTAL oscillators. This was a couple of zeners and a mosfet for 240VDC. Again, improvement but still I was getting 50 Hz or so of shift.   

Next I did what I should have done in the first place - changed tubes. I saw everything from 150Hz shift to 10 Hz shift! Depending on the tube!

I also put a diode across the relay coil which is a 250V job! This kicked back directly into the medium voltage line.

Checking the fillaments on the VFO I noticed that they put two tubes in series to make up the 12.6VAC. This sagged 2 tenths of a volts 12.6 to 12.4VAC on TX. I did a simple half wave rectifier to a low drop our regulator and set it so I had exactly 6VDC on the fill of the VFO.

BINGO - that was the culprit. All became stable. No shift RX to TX.

Now the 1962 Hallicrafters SR-150 is as stable as those synthesized jobs Smiley Well maybe not quite...

Mike WU2D




 

* 150_VFO.pdf (2868.8 KB - downloaded 177 times.)

* hallicrafterssr150.jpg (52.77 KB, 640x316 - viewed 427 times.)
Logged

These are the good old days of AM
Jim WB5WPA
Guest
« Reply #1 on: December 26, 2011, 11:32:13 PM »

That was a good bit of detective work!
Logged
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« Reply #2 on: December 27, 2011, 09:17:53 AM »

This was a frustrating deal that took longer than expected and I "gave up" twice only to be drawn back in.

VFO Drift and compensation are topics that we see a lot but I had not heard of much about this kind of TX to RX shift issue. Does a VFO have to have regulated DC on the filament? No - but is does need stable voltage either AC or DC if you desire solid state performance.
Logged

These are the good old days of AM
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #3 on: December 27, 2011, 02:18:42 PM »


Mike,

   You touched upon a lot of important issues we face with vintage VFO's. The use of a Low Drop Out Regulator for the filament supply is neat since it limits extra heat buildup from regulator dissipation.
I'm curious though, did you make 12 something volts DC for the series two tube filament string, and what type of LDO regulator did you use?

The crystal oscillator circuits can also be a source of 100's of hertz drift over the warmup period, and that is apart from the drift from filament, plate and screen regulation issues. You will find that those series zener strings getting to 240v will have quite an upward temp drift as the voltage increases when the diodes get warm. Still, you will have good load regulation which I think is more important.

Since your VFO operates all the time, you escaped the usual key up and drift from RF heating of the oscillator tank L-C. That is also a tough nut to crack.

Thanks for posting your experiences.

Jim
WD5JKO
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #4 on: December 27, 2011, 03:13:16 PM »

The only way I could get the vfo in my 75A-4 and the PTO in my VFO (a converted T-368 master oscillator unit) to remain rock stable, both of which run all the time as soon as the a.c. power to the rig is switched on, was to run both units on regulated filament voltage.  The culprit was small variations in line voltage, some of this occurring with the mains voltage, and some due to variations on the load when various units in the shack, especially the transmitter plate supplies, were switched on.

For the receiver, I put the whole thing on a Sola constant voltage transformer that I picked up at a hamfest.  It was a modern one, about 3" X 3" X 5", designed to run a desktop computer, and unlike the older classic versions that were notorious for generating an annoying loud acoustical buzz, this one is virtually silent.  For the VFO, I run the filaments off a regulated +12v DC supply, with a dropping resistor to get the voltage down to 6.3 volts.  I need to build up a dedicated homebrew one to deliver 6.3 volts directly.

On 40m, even less than a volt of line voltage variation is enough  to cause both the receiver and vfo to drift far enough to move outside the passband of my narrow 300~ CW filter.  The B+ to both oscillators was already DC regulated.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WU2D
Contributing
Member
*
Offline Offline

Posts: 1800


CW is just a narrower version of AM


« Reply #5 on: December 27, 2011, 03:28:58 PM »

I really had no idea that the filament voltage was so important. I figured that the thermal changes had to be quite drastic to make a VFO move, but this is not the case. I took the 12.6VAC and did a simple half wave 1N4001 diode to a 2200 uF cap and tried a 7812. With the two 6EA8 tubes in series (0.45 Amps) the regulator wasn't. I mean the regulator did not have enough delta input to output to actually regulate. With a huge 10,000 uF capacitor it did work. So I looked in the junkbox and found a Linear Technology LM1085 which is a 3A "low dropout" version of the LM317. It worked at 12VDC but surprisingly not with much headroom. I guess this device is "lower dropout" but not on par with modern low dropout devices. Anyway, that is working and the VFO does not care about having two tubes in series.

Yes as K4KYV says, even little events like the furnace or sump pump coming on will drive a VFO nuts even if the plate is regulated. You have to do the filament too. 

Mike Wu2D
Logged

These are the good old days of AM
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #6 on: December 27, 2011, 06:25:07 PM »

Yes as K4KYV says, even little events like the furnace or sump pump coming on will drive a VFO nuts even if the plate is regulated. You have to do the filament too. 
Mike Wu2D

   Yes, but as you also described, changing out the VFO tube, and picking the best of several can have a dramatic effect on frequency stability. I like to use a variac and check the frequency when I go from say 110v to 120v and back recording the drift for each tube. With the 1626 in a BC-458 used as a VFO (CE 20A), this test can reveal a 10:1 range between the best (lowest drift), and worst (highest drift) where all test good on an I-77 or similar tube checker. Sometimes a tube with a well used indirect cathode will shift a lot as the filament voltage is reduced, and this will also show up with an emission tester.

   For those into "no-mods" or collectors of high end vintage gear, the VFO tube "pick of the litter" choice can make a huge difference without having to drill holes, or solder anything.

   Oh, for those into Huff & Puff stabilizers (me), a prerequisite with tube based VFO's is to regulate the filament voltage. But we have covered this before on AMFONE.

  OK on your half wave rectified supply freeding the LM1085. You might consider replacing that 1N4001 with a Schottky rectifier since these drop about 1/2 the voltage of an otherwise similar silicon diode. Look at the 1N5818 as one example.

http://www.fairchildsemi.com/ds/1N/1N5818.pdf

Jim
WD5JKO
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #7 on: December 29, 2011, 09:38:55 PM »

I didnt have to go to that extreme but did similar with a Globe Champ 300 that I wanted stable on 10M AM. The biggest improvement was tossing the 6AU6 for a 6AH6 which is also SOP for Heath and Johnson VFO's. Zener regulating plate and screen at slightly lower voltages also helped as I had sufficient drive. Ive done the regulated DC filament thing to many BA receivers that I use on CW and SSB as well as the 6n2M Clegg Zeus and Interceptor.

Luckily i havent had to deal with any interest in digital modes yet but that might occur on VHF/UHF weak signal were a couple of TS-830's drive transverters. They use a SS LC VFO which has been fine for CW meteor scatter but I doubt it for JT.

Carl
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.056 seconds with 18 queries.