Hi Arv,

"Consider H&P as a PLL design that locks on multiples of the reference frequency.? It?

adjusts slower than a PLL but otherwise many of the PLL design formula can be used."

Its not a PLL, I've designed more than a few.? Its a FLL, the differnce is one your
locked to the phase of a reference signal and the other you measure frequency
and correct.? The latter allow for some latitude depending on count resolution.

My first H&P was a signal generator locked to the LSB of a frequency counter.
Handy gadget.? ?

I would later find a HP sychronizer for the HP603 signal generator and study
how that worked.? That was phase locking to harmonics.? By the early 70s
I was doing full out /n and sampling? PLL systems.

My reference is to some of the simpler circuit used short counters and the amount
of drift tolerable would have the frequency of the VFO bouncing back and forth
depending on the filter RC time constant and the count resolution.? Those would?
drift between finite points and those were defined by the count resolution.

A stable VFO can be enhanced but poor and unstable VFO should be fixed first then?
improved.? Most of the time a frequency readout is needed just to understand the drift.

Allison

One more issue with the Heathkit vfo/DX-60b problem is that the vfo outputs 7 mhz for 40-10 meters.? The radio x2 x3 or x4 to get the correct frequency.....so, a 100 hz drift on 40 meters becomes a 400hz drift over the same time period on 10 meters.

The HG-10 can be more stable than it was designed by completely rebuilding the vfo (re: temperature stable caps, removing heat, etc) but an accurate Huff and Puff implementation would fix it completely, once set up properly.

I am starting on a homebuild PTO vfo using a kit from qrpguys, that I hope to use as a test bed for making a stable 5mhz PTO for an a.m. transmitter build.? I might just throw my hands up in the air, in the end, but I would like to see what I can do.

Allison

All good information, but can I politely disagree with a couple of your statements.

1)? Back when many of us were experimenting with H&P design it was common to

obtain stability within 2 Hz as compared with WWV over an hour.? Stable VFO and a

slow correction factor were used to get this level of performance.? Of course stability of

the reference source is as important as is the H&P design itself.

2)? There is no specific "two frequencies" involved with H&P.? The frequency bounds

are determined by amount and rate of drift in the oscillator.? The correction factor being

applied and how often it is applied has a big effect on how well frequency is controlled.?

3)? Apparently there are as many poorly understood and poorly implemented H&P circuits

out there as there are good and well understood ones.? Your mileage will vary.

H&P sometimes gets a poor performance review due to the user not taking time to

understand how it works, and thus not adjusting the control loop properly for the VFO

being stabilized.? Properly done it can be as stable as its reference signal source.?

Consider H&P as a PLL design that locks on multiples of the reference frequency.? It

adjusts slower than a PLL but otherwise many of the PLL design formula can be used.

If doing H&P stabilization you need to measure the non-stabilized amount of frequency

drift in a particular time period.? Then using that information you need to determine how

much compensation and how frequently it is applied to counteract that drift.? A series

capacitor is placed between the varactor and the VFO tank circuit.? Value of this cap

determines how aggressive the H&P correction will be.? It will be different for each

VFO, thus the comment that it could not be a kit system because each application is

different.?

The so-called FAST H&P is a different animal.? It is easier to simply "connect and use"

but the need for fine tuning still applies for optimum performance.? "FAST" refers to the

rate of correction application, resulting in less tendency for jitter than from a poorly designed

traditional H&P system.? FAST H&P systems seem to be more tolerant of varying amounts

of drift and have a larger correction range.? If I was going to make a kit H&P system it would

probably be the FAST H&P version because it is more tolerant of poor installation.

There are computerized versions of frequency stabilizers that are similar to H&P, but not

really H&P in the purest sense.? These usually measure VFO frequency and periodically

apply a correction factor via varactor or magnetic saturation tuning to hold the frequency at

a user-determined value.?

If using a PWM output to drive the tuning varactor you will probably need a separate

integration capacitor (filter) between the PWM and varactor to minimize any jitter introduced

by the PWM pulses.??

If there is enough interest in resurrecting the H&P discussion it might be worthwhile to look

at moving that discussion group from to the newer Groups.io platform.?

Arv

_._

Hi Jim, I would personally not hook the shield to to the QCX or the GPS. I have about 3 GPS units running and have never used a shield. If after using the radios and you think? that a shield would help then I would definitely hook it to the negative lead on the cable on both ends. BUT the big issue with hooking a GPS to the QCX is if the QCX is in CW mode it will key the transmitter solid. So you may want to switch the QCX to WSPR mode to calibrate it. I have only done this one time on each of my QCX radio's. OR make sure you have a dummy load on the QCX, If you do this to calibrate the radio in CW mode. I hooked a GPS to my first QCX? and by chance I had a dummy load on it. But the finals got rather hot. I was lucky.? Good Luck Bill KN8DMK

VFO instability and drift are symptoms of parts failure or implementation errors.
Sometimes its just plain broken and needs to be fixed.

If it does not have a regulated supply it will be unstable even with vacuum tubes.
If the circuit use poor quality or non temperature compensated parts they are
also trouble for drift.

I've also seen cases where just changing the oscillator tube improved matters.

HG10B with a OA2 added to stabilize the VFO volage does get better and if?
some of the caps are replaced with COG and a mix of N750 parts it will be stable.
Allowing the heat to escape doesn't hurt.? Doing this is more a matter of time
paying attention to the drift then chaning from zero temp coefficient to
-750temp coefficient until drift is near zero or going the other way. Iterate
as needed.

How stable is needed varies with application for CW or SSB less than
100hz/hour is attainable.??Getting to good enough for AM anything
under 1khz an hour will do but its easy to do much better than that.

All the same things apply to solid state VFOs.? I have built many and 1hz
an hour is not that hard with patience and maintaining a stable temperature.

Primary offenders is heat always.? Toroids like FT50-7 are about the best
at +35ppm for temperature. Varicap diodes are terrible and can easily
exceed +100PPM and also that varies with what applied votlage so
stabilizing by temperature compensation is tricky at best.? In some cases
junction transistors operated at the less an ideal operating point will heat
internally and cause drift. Most published designs are at best sloppy
and rarely was the effort made to even mention how to make it stable.

The problem with H&P is its always bouncing between two frequencies and
usually that is anywhere from 20 to 80hz.? Narrower than that and the VFO
can move out of lock to the next range up/down as the time to sample goes up.
So that required the VFO to be at least marginally stable.? Also it needs a way
to turn it off while tuning or it will fight you.

Allison?