Drift

[K9ACT]

I received an advisory note from a hallway monitor, er ah... OO stating that my CW sig was drifting during a QSO on 40 Meters.

The VFO on my CW rig is a homebrew classic 6AG7, gas regulator and of course it drifts.  A few hundred cycles or so but riceboxes drift too. Everything does and nothing is perfect.

The question is, how much drift is too much?  He cited FCC reg 47 CFR 97.101 (a).

A search produced this:

(a) In all respects not specifically covered by FCC Rules each amateur station must be operated in accordance with good engineering and good amateur practice.

So, how much drift is bad engineering or bad amateur practice?

Although brand new and right out of an old handbook, my rig functions about like most boat anchors of the period so are we to accept the opinion that this stuff is no longer good engineering and all good amateurs should go out and buy rice boxes?

Jack

[W3RSW]

Well if it's a single tube vfo your very stable. Almost has to be some drift unless your name's "Art";way less drift than a TR7.  Slow drift should be fine; fast drift or warble probably isn't.  Outright chirp, of course, is objectionable but not unheard for a Hartly direct coupled to the antenna.

Methinks in this day and age of plastic; he's never heard a real rig before or was very bored since most everone is perfectly digitally controlled right down to 0.1Hz.
Tell him that drift's not much of anything compared to chirp.

If it's a crystal osc. then you have every right to reply to him, thank him for his observations but outline your exact circuit, schematic and page no. from the handbook. Tell him he's missing out on classic radio and ought to read the 'Old Radio' section of "QST" to see more examples of what you've built yourself.

I'd write to thank him regardless and mention the joys of homebrewing in this day and age....  and continue to run your rig.
 

[w3jn]

I think Rick is on the money.  He's never heard a classic rig on the air.

I'd just ignore it.

That said, if you want to help the drift, find a 10-20 pf N750 disc cap and put it in parallel with the coil, as close to the coil as possible.  You may have to play with the value a bit but this will help tame the drift.

A 6AG7 VFO, while OK in its day, really isn't up to today's standards.  The 6AG7 runs too hot.  A much better design would use a 6J6 or 12AU7, or 6U8, with 1/2 of the tube used as a buffer.  THEN drive the 6AG7.

[k4kyv]

I don't think the FCC would be concerned about it, unless it was so bad that it was  drifting off your frequency and landed on top of another QSO.

But I still like for my signal to be as drift-free as possible.  That's why my VFO runs all the time, 24/7, as long as it is connected to the a.c. power mains.  I have it shielded well enough that I don't hear it in the receiver so I  don't have to turn it off during receive.  It multiplies up from 160m, so no spurious images from heterodyne products.  I run the tube filament as well as the plate off regulated DC.  I found that even a volt or so of line voltage variation caused enough filament voltage change to make the vfo drift out of the  receiver passband on 40m.

Using a good cw filter, with bandwidth <500 Hz, and a stable VFO on both transmitter and receiver, allows signals to be spaced a few hundred Hz apart and still be QRM-free.  That's why the new 80m allocations are working so well, and the cw ops still have room to operate without undue congestion.

As far as chirp goes, I don't find a slight chirp objectionable.  It makes the signal less boring to listen to, and I have found it even makes some signals easier to copy.  I am talking about a very slight chirp, nothing that would cause the signal to move outside the receiver passband, even at 300~ selectivity.

[W2XR]

I think Rick is on the money.  He's never heard a classic rig on the air.

I'd just ignore it.

That said, if you want to help the drift, find a 10-20 pf N750 disc cap and put it in parallel with the coil, as close to the coil as possible.  You may have to play with the value a bit but this will help tame the drift.

A 6AG7 VFO, while OK in its day, really isn't up to today's standards.  The 6AG7 runs too hot.  A much better design would use a 6J6 or 12AU7, or 6U8, with 1/2 of the tube used as a buffer.  THEN drive the 6AG7.

I agree with John.

Another possibility would be to use a 6AU6 as the oscillator, followed by a 6AG7, etc., as a buffer.

To really gild the lilly and improve the short-term stability (i.e. immediately upon turn-on of the VFO and for perhaps another 15-20 minutes operating time thereafter), consider wiring the filaments of these two tubes so that they remain on 24/7. Heath did this in the Apache and the Marauder transmitters, and it did improve the short-term stability of the VFO to some degree.

If this was chirp (FM) the OO noticed, then that is a problem. If it was simply a slow frequency drift, typical of the older boatanchor-type rigs, then IMHO, that's BS and the OO was clearly incorrect in his judgement. Incidently, I once had 60 hz FM on my carrier; it was a cathode-to-heater short within the VFO oscillator tube. A new tube solved that problem.

I also would not be too concerned about this matter.

73,

Bruce

[WQ9E]

He cited FCC reg 47 CFR 97.101 (a).

A search produced this:

(a) In all respects not specifically covered by FCC Rules each amateur station must be operated in accordance with good engineering and good amateur practice.
 (end quote)

If the FCC ACTUALLY enforced anything remotely resembling "good amateur practice" I think a little bit of drift would be considered minor compared to the undesirable on-the-air behavior. 

As I recall, isn't the OO corp set up by the ARRL?  Perhaps someone who is still a member could get some information on what sort of training they receive and perhaps suggest that their time might be better spent monitoring behavior.  Which causes more problems (both for operators and also for our PR efforts): a little drift or a gaggle of idiots proudly parading their stupidity for all to hear???

Rodger

[k7yoo]

Mental drift is far worse than VFO drift...... perhaps he doesn't like your politics!!

I would politely thank him for his concern, outline the technical measures you are considering, and move on to have a nice day. You might mention the recent "vintage " articles in QST and how you are running a similar setup.
BTW..if you want I will bring you a freq counter to monitor your drift. Will you be at Stoughton?
Skip

[k4kyv]

To really gild the lilly and improve the short-term stability (i.e. immediately upon turn-on of the VFO and for perhaps another 15-20 minutes operating time thereafter), consider wiring the filaments of these two tubes so that they remain on 24/7. Heath did this in the Apache and the Marauder transmitters, and it did improve the short-term stability of the VFO to some degree.

But it will also cause premature tube failure.  Better to let it draw a small amount of plate current as well.  Just leaving the filament on allows an electron cloud to build up around the hot cathode, and the magnetic field generated by the a.c. filament current makes those electrons vibrate at 60hZ, and they literally sandblast away the oxide coating, or in the case of transmitting tubes, the thorium layer.  A fraction of a milliamp of plate current is enough to pull the electrons away from the cathode structure.

Ther R-390A manual cautions against leaving the  receiver in stand-by mode for long periods of time, to avoid shortening the life of certain tubes.

[N0WEK]

To really gild the lilly and improve the short-term stability (i.e. immediately upon turn-on of the VFO and for perhaps another 15-20 minutes operating time thereafter), consider wiring the filaments of these two tubes so that they remain on 24/7. Heath did this in the Apache and the Marauder transmitters, and it did improve the short-term stability of the VFO to some degree.

But it will also cause premature tube failure.  Better to let it draw a small amount of plate current as well.  Just leaving the filament on allows an electron cloud to build up around the hot cathode, and the magnetic field generated by the a.c. filament current makes those electrons vibrate at 60hZ, and they literally sandblast away the oxide coating, or in the case of transmitting tubes, the thorium layer.  A fraction of a milliamp of plate current is enough to pull the electrons away from the cathode structure.

Ther R-390A manual cautions against leaving the  receiver in stand-by mode for long periods of time, to avoid shortening the life of certain tubes.

Would I be wrong in assuming that filaments running on DC aren't affected by this? Maybe another reason to run them on DC.

Greg

[Todd, KA1KAQ]

Well Jack, I stand by what we discussed on the air, based on the amount of drift on your signal: this OO is a Yaecomwood driver and hasn't a clue what drift is, or acceptable levels thereof. It would be different if your signal was drifting out of the passband and required constant adjustment of the tuning, then sure - that's worth mentioning. But I detected next to no drift on your signal, which makes me think your report was based on 'no old or HB tube gear allowed'.

Did he give you a definitive measurement of cycles/second or minute?

[K1JJ]

Hi Jack -

There's some very good ideas here.

I've built several tube VFO's over the years and still use the one in my homebrew 11 tube superhet receiver. Here's how I would handle this situation... :-)


Take the bull by the horns and make the VFO so good there is no issue. Consider it a fun project challenge.  First, as already suggested, set up a freq counter and take measurements of the existing setup . Measure drift from a cold start, drift from 5 minutes warmup on, chirp and any other parameters you think will help to show the existing habits.  Get a baseline for future improvements.

Then do as Johhny N says... put in a dual envelope triode/pentode, or whatever, as an oscillator and buffer, and use the 6AG7 as a simple class C amp driver to the outside world. You need good load isolation.

Keep interstage coupling cap values as small as possible to keep loading to a minimum while still maintaining proper drive throughout.

Measure your regulator to insure it is rock solid stable or make it so. This means plate/screen and filaments. DC and a zenor in the fil should do it.  Also be sure the outside load to the 6AG7 is not pulling down too badly or a second buffer is needed, but unlikely.

Consider running the unit 24/7 as suggested with plate voltage.

Then finally, get an assorted bunch of the 750 NPO caps and contact solder them in one at a time while taking measurments for drift. With time, you can learn the VFO's habits and counter anything. Customization is your friend here.  You can do it while manufacturers cannot spend the time.

I say, "screw these newbie OO's - I'll make it so good it'll be even better than their riceboxes."  You will succeed.... because you CAN make it that good. Just like fine-tuning your FB telescopes, it'll be fun and a great learning project to be proud of.

Later-

T

[WA1GFZ]

It must be quota week at the OO police dept. Ask him if he can tell you the drift rate and what NIST instrument was used to determine you had a problem.
Suggest the guy get a life