In particular, say I took a garden-variety 20MHz fundamental
microprocessor crystal and instead used it at its fifth overtone,
trying to hit 100 MHz. The LC network is there to make sure that it's
on its fifth overtone. Will this "misuse" mean that the oscillator
will be harder to start up, less stable, more noisy, ???, than a
crystal oscillator made out of a real overtone crystal? I don't mind
if I "miss" 100 MHz by a several tens or hundreds of ppm, as long as
it's stable there.
If anyone knows of a place that ships off-the-shelf 100 MHz fifth or
seventh overtone crystals, I can avoid this whole exercise.... :-)
Overtone crystal cuts are not fundamentally different from fundamental
crystal cuts, so to a 1st-order approximation they'll work. Crystals do
have spurious responses that can cause mode jumping, and these responses
don't necessarily map the same way the overtones do, so using a 20MHz
crystal at 100MHz may or may not work, depending on the luck of the
draw. Other than that I don't know of any differences.
IIRC Digi-Key has 100MHz crystals, but I may be remembering 100MHz
oscillators. YMMV. IDNKWTFIAS. Caviat Emptor (so _that's_ what CE
means! Here I thought it was a quality mark). Etc.
Wescott Design Services
>Are "overtone" crystals cut differently than "fundamental" crystals?
>Or are they just specified differently?
>In particular, say I took a garden-variety 20MHz fundamental
>microprocessor crystal and instead used it at its fifth overtone,
>trying to hit 100 MHz. The LC network is there to make sure that it's
>on its fifth overtone. Will this "misuse" mean that the oscillator
>will be harder to start up, less stable, more noisy, ???, than a
>crystal oscillator made out of a real overtone crystal? I don't mind
>if I "miss" 100 MHz by a several tens or hundreds of ppm, as long as
>it's stable there.
You can use a fundamental mode crystal as an overtone oscillator, but
even if you can get it to oscillate, it won't be generating an
overtone at 100MHz, since overtone modes of oscillation aren't
harmonically related to the fundamental. It's more like the slab of
crystal is vibrating like the drumhead of a steel drum with small
areas of the slab vibrating at higher frequencies, instead of the
entire slab virbarting at just one frequency.
Check out "Chladni patterns" if you're interested.
Here's some pattrens for violin tops and circular plates:
>If anyone knows of a place that ships off-the-shelf 100 MHz fifth or
>seventh overtone crystals, I can avoid this whole exercise.... :-)
Anybody who makes crystals ought to be able to help you out; here's a
In an AT cut crystal the overtone modes are close, but not exactly on,
the odd harmonics of the fundamental. Furthermore, all of the
literature that I've read on AT cut crystals reports that they vibrate
in the bulk of the crystal, in shear mode -- see figure 7 here:
Perhaps you're thinking of SAW devices?
>Furthermore, all of the
>literature that I've read on AT cut crystals reports that they vibrate
>in the bulk of the crystal, in shear mode -- see figure 7 here:
>Perhaps you're thinking of SAW devices?
No, I was thinking they vibrated in thickness compression. Thanks for
One effect to watch out for with use of unspecified
overtone modes is that the behavior of the resonator
is not ideal; the presence or size of nearby spurs and
the Q depend on how uniform the thickness is that
determines frequency and the placement and size of
contact metal. The wavelength is typically much less
than the dimension along the non-shearing axis, so
having a single mode of resonance near the nominal
frequency or its overtones is not guaranteed, except
by careful construction and verification. So, clearly,
a guarantee about the behavior near the fundamental
resonance cannot be extended to the overtone modes.
If I was trying to build a stable and pure oscillator
operating at a crystal overtone, I would buy the
crystal specified for the overtone I would be using.
Above views may belong only to me.
I pointed that out in a previous post. But hey -- wouldn't it be fun to
have an oscillator that yodels?
a CB xtal will probably operate on 100MHz, althouth I've only seen
applications for 45 and 81MHz
J. M. Noeding, LA8AK, N-4623 Kristiansand
to get optimum performance one would grind the 100MHz 5.OT finer or
even polish it, and the thickness of the electrodes might be different
to get optimum Q.
But you should be ok by using a 20MHz fundamental in its 5th.
There are also manufacturers that make 100 in fundamental (up to about
200MHz), and many should have 100 in 5th as standard part...
> IIRC Digi-Key has 100MHz crystals, but I may be remembering 100MHz
> oscillators. YMMV. IDNKWTFIAS. Caviat Emptor (so _that's_ what CE
> means! Here I thought it was a quality mark). Etc.
Dialog news reader's tip popped up to say that CE is "creative
editing". It doesn't KWTF IDNKWTFIAS is, but I got everything but
the IAS part.
> I don't know why off-the-shelf crystals are needed when Jan Crystal (Ft.
> Myers FL) will make the crystal to your specifications in a few days for the
> same amount of money. They can do fifth ot at 100 MHz. quite easily.
Jan's what I was about to suggest. I thought the rock I needed would
have been off the shelf, but they made it and sent the test results.
IDNKWTFIAS: I Don't Know What I Am Saying.
As Douglas said the frequency may be a bit off unless you get a crystal
made for 5th. An alternative for the 20MHz garden variety would be to
make a 20MHz oscillator, run it into a fast gate and fish out the 5th
the old fashioned way, with an LC circuit. Then run that through a gate
again if needed.
I Do Not Know WTF I Am Saying?
Now Rich, that's being awfully harsh on yourself.
OK, silly I know, but I cant figure out YMMV...
OTOH, WTF, IAAR
sorry dude, 50 years of IEEE UFFC papers suggest *you* are wrong. I was
surprised when I learned this too.
Your Mileage May Vary
Beware of those who post from srvinet.com!
Michael A. Terrell
>OK, silly I know, but I cant figure out YMMV...
Year 2005. (;-)
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
I remember being told by a crystal 'expert' that with some cuts the
difference can be much larger than that. Is that so?
Honestly John! That should be AMMV.
not pull with external reactance change perhaps a byt
Reason for the post, I think you've changed ISP's on me again, my mail to
you gets bounced. Would you pse address me a short note to the e-mail
address and give me the current one? That is, if it's not me you're trying
to get rid of!
There is a related phenomenon in the field of piano tuning. It has
long been known that overtones (called "partials" by piano people) of
piano notes are not exactly related to pitch of the fundamental
frequency by whole numbered ratios. Instead they are related by
The amount by which this series deviates from the ideal whole-numbered
ratios is called "inharmonicity" and it differs from one string to
another. The stiffer the string, the more inharmonicity. Long thin
strings, as are found on harpsichords, have almost no inharmonicity.
Short strings in the highest section of the piano have the most
inharmonicity. Since one of the goals of piano tuning is to make
partials of different notes come out the same, this phenomenon of
inharmonicity makes piano tuning inherently more difficult than
instruments that have no inharmonicity, like pipe organs.
What is perhaps more like quartz crystals is carillon bells. They are
tuned at the factory, and each partial is tuned independently and
separately by grinding away metal from different levels on the bell.
In view of these related phenomena, it is no wonder that overtones of
quartz crystals are independent of each other and from the
There is a whole science of crystals all on its own. When they make a
crystal they make a thin disk of material. You would normally expect the
edge of the disk to simply be at right angles. Instead it looks like
The exact angle and depth of that chamfer is how they control which
overtones are selected for and which are supressed. In fundamental
crystals, the maker usually grinds the chamfer so as to reduce the 3rd
kens...@rahul.net forging knowledge
No, its more like a jello when you jiggle the dish side to side. The main
action of an AT cut is shear mode. In the harmonics, the motion looks
kind of like this:
If you think about the top two lines of text in my little drawing. I
think it is obvious that if the maker thinned it down by one line of text
just as you come to the edge, that portion of the crystal would not work
well at this harmonic. This is what they do in crystals intended for
fundamental operation. It knocks that activity down by several dB at the
overtone. This makes it very unlikely that a simple oscillator will take
off at an overtone.
In the ideal AT cut crystal "c mode" shear is the only activity. In the SC
cut, the "b" and "a" modes appear. The extra complexity of the mode
selection circuit is part of the reason that SC based OCXOs cost so much.
Will the harmonic be precise? No. Will it be "close", which is what the
original poster asked? You bet. Depending on the oscillator circuit, can
it be "pulled" on frequency? Perhaps.
But to say that the crystal doesn't resonate anywhere near the harmonic is,
as I said, bullpuckey.
"Terry Given" <my_...@ieee.org> wrote in message
> RST Engineering wrote:
>> That is total and absolute bullpuckey.
>Sorry, dude, 50 years of designing with crystals, right from when I ground
>my first surplus WWII rock on a piece of glass with toothpaste as the
>abrasive says that what the original poster asked is correct.
>Will the harmonic be precise? No. Will it be "close", which is what the
>original poster asked? You bet. Depending on the oscillator circuit, can
>it be "pulled" on frequency? Perhaps.
>But to say that the crystal doesn't resonate anywhere near the harmonic is,
>as I said, bullpuckey.
Sorry, dude, no matter how much time you've got in, if you go back
and read my post, you'll find that I wrote:
"You can use a fundamental mode crystal as an overtone oscillator, but
even if you can get it to oscillate, it won't be generating an
overtone at 100MHz, since overtone modes of oscillation aren't
harmonically related to the fundamental."
and that you replied with:
"That is total and absolute bullpuckey."
Notice that I didn't say "near", I said "at".
If you can find fault with anything I wrote in that post, I'd
appreciate specific criticism instead of that broad brush you painted
>"You can use a fundamental mode crystal as an overtone oscillator, but
>even if you can get it to oscillate, it won't be generating an
>overtone at 100MHz, since overtone modes of oscillation aren't
>harmonically related to the fundamental."
When you look at older (pre-PLL) VHF communication gear of the more
professional kind they didn't use 5th or higher overtones but employed
frequency multiplier stages. For good reason, one being the offset you
had mentioned. I'd never run a crystal on its umpteenth harmonic and
always designed in multiplier stages like the radio folks did. With
today's cheap logic chips that doesn't even cost much in extra parts.
>and that you replied with:
>"That is total and absolute bullpuckey."
Look on the bright side. Some of us, including me, didn't know the
expression "bullpuckey". I got a kick out of it.
> I'm currently
> using a SAW at 660 MHz for the clock in a 9951 DDS.
Interesting factoid: I was looking to experiment with 100MHz
oscillators largely as a clock source for my own AD9951 experimentation
(using the AD9951's built-in PLL multiplier at 4x). I was hoping to
experiment a bit with 20 MHz crystals I already had in hand before
ordering some "real overtone" crystals cut just for me. I've been
looking at AD app note AN-419 and it's Butler oscillator, in
particular, although the clock input of the AD9951 probably has
different requirements than the AD9850 targetted in AN-419.
Does the AD9951 really work at 660MHz? I thought it was only good to
So far my experimenting has used the on-chip oscillator at 25MHz and
the PLL at 16x to get to 400MHz.
> Actually, it's better
> than my 200 MHz 7th overtone tripled to 660
We bandied about "non-harmonic" relations here but how you get from 200
to 660, I don't know.
The built-in multiplier is quite noisy and makes the 9951 run terribly hot.
> Does the AD9951 really work at 660MHz? I thought it was only good to
Yes, if you DON'T use the on board multiplier. I've had it to 750 MHz just
to check it since I had heard of some DL's overclocking it to that
frequency. Properly heat sunk to the eval board, and without the multiplier,
it's cool as a cucumber. AD rates it only to 400 MHz but a sample of 6 units
all operate well at 660 MHz.
> So far my experimenting has used the on-chip oscillator at 25MHz and
> the PLL at 16x to get to 400MHz.
> > Actually, it's better
> > than my 200 MHz 7th overtone tripled to 660
> We bandied about "non-harmonic" relations here but how you get from 200
> to 660, I don't know.
Well, this one is a 220 MHz 7th overtone from ICL specially surface treated
for low noise and operating in a Stephensen bipolar/FET Butler. But as I
mentioned to Doug, afraid my MMIC tripler makes a bad job of the 660 output
despite a 3 pole final filter. The SAW is not near the Q of the crystal, but
the SNR is much better.
> Often cheaper to multiply up than buy an expensive 5th overtone that
> was difficult to pull onto frequency and fussy to set up.
And these special cuts can indeed be fussy. They can also be a
> The exception would be current and size saving for some portables.
Even then it could be done. Besides the discrete solution there are
blazingly fast logic inverters such as the ALVC series. These are
usually under 20 cents and come in the super tiny TSSOP format. Now I
just wish they had unbuffered versions to do the oscillator part with.
If a 74HCU04 is needed for other jobs on the board it could run the
oscillator but for any reasonable speed these require more than 4V.
Now suppose someone makes a crystal oscillate in some overtone mode that
the crystal manufacturer recommends against and is predicted to be
"inharmonic" but turns out to be only a few hundred or even sometimes a
few 10's of KHz from a multiple of a frequency that results from being
used as directed?
As I said in different words in a different post - correctly predicting
that $#!+ (AKA "slop") will spatter does not necessarily that much will
spatter nor that any will spatter far, and maybe in many cases it is
doubtful that both much will spatter and that much will spatter far.
- Don Klipstein (d...@misty.com)
If it's not an integer multiple of the fundamental then it won't be a
> As I said in different words in a different post - correctly predicting
>that $#!+ (AKA "slop") will spatter does not necessarily that much will
>spatter nor that any will spatter far, and maybe in many cases it is
>doubtful that both much will spatter and that much will spatter far.
If the prediction came true, then it came true.
It can also oscillate in one of several mechanical modes, eg., longitudinal,
breadth-wise or in torsion. And in shunt or series-resonant electrical
The circuit it is embedded in can encourage a preferred frequency. It is
easy to select harmonics. Self-preference is also given to the frequency
which has the highest Q, ie., the least mechanical loss. This is usually the
It does not oscillate EXACTLY at multiples simply because it has three
dimensions and Length, Breadth and Thickness slightly 'interfere' with each
A poorly cut crystal, eg., lack of parallelism, at which there may be no
strong preference may jump erratically between two non-harmonically related
Frequency versus temperature curves depend on oscillation mode and on the
angle at which the slab is cut relative to the direction of the individual
crystals in the bulk material lattice as found by optical means. Cubic
curves are best because they contain a flat horizontal portion.
have been examining some surplus mobile telephone base station
equipment and discovered that the 70MHz to 455kHz mixers consists of
2x SA602. Since I've never seen an application using two such items,
my guess it for an image rejection type mixer.
Could somebody please guide me into some notes describing such mixer,
possibly using 2x SA602 (and a crystal osc)
J. M. Noeding, LA8AK, N-4623 Kristiansand
Frank GM0CSZ / KN6WH
An image rejecting mixer requires quadrature inputs (both signal and
LO), two mixers, and summation of the outputs i.e.
sin(A+B) = sin(A)cos(B) + cos(A)sin(B)
Also - you're unlikely to have image problems at the second mixer.
J M Noeding wrote:
> have been examining some surplus mobile telephone base station
> equipment and discovered that the 70MHz to 455kHz mixers consists of
> 2x SA602. Since I've never seen an application using two such items,
> my guess it for an image rejection type mixer.
Could it be for dual diversity receive?
there are 4 receivers, two on each board (with 2x SA604 and 4x SA602),
so I think it is a lot of diversity if it was interesting