Following up on my post from yesterday that got no
resonses, I am assuming that everybody uses crystals
at 20MHz because they tried a resonator once and it
didn't work.
That's what I did, but to keep parts cost down, I'd
love to use a resonator in a new design if it's
possible. It also appears that there are lots of new
"high frequency" resonators to choose from.
Do any of them work on PICs? Inquiring minds want to
know. I guess I'll have to get an assortment and be
the guinea pig.
Cheerful regards,
Bob
P.S. I'd still love to hear from anyone who has tried,
with part numbers if you've got 'em.
I've used 20MHz resonators with built in capacitors in some projects.
These products ship in pretty large volumes and there were unacceptable
failure rates (something like 0.5% to 1.0%). In some of those products,
we've gone to 5MHz resonators and use the PLL to get 20MHz. In others
we've gone to a 20MHz crystal. In another product line, I used 16MHz
resonators with built in capacitors. Later (when I moved from 16C OTP
parts to 18F parts), I changed to 10MHz resonators and the PLL clock
multiplier. Much more reliable.
> Following up on my post from yesterday that got no
> resonses, I am assuming that everybody uses crystals
> at 20MHz because they tried a resonator once and it
> didn't work.
>
> That's what I did, but to keep parts cost down, I'd
> love to use a resonator in a new design if it's
> possible. It also appears that there are lots of new
> "high frequency" resonators to choose from.
>
> Do any of them work on PICs? Inquiring minds want to
> know. I guess I'll have to get an assortment and be
> the guinea pig.
>
> Cheerful regards,
> Bob
>
> P.S. I'd still love to hear from anyone who has tried,
> with part numbers if you've got 'em.
> Following up on my post from yesterday that got no
> resonses, I am assuming that everybody uses crystals
> at 20MHz because they tried a resonator once and it
> didn't work.
>
> That's what I did, but to keep parts cost down, I'd
> love to use a resonator in a new design if it's
> possible. It also appears that there are lots of new
> "high frequency" resonators to choose from.
>
> Do any of them work on PICs? Inquiring minds want to
> know. I guess I'll have to get an assortment and be
> the guinea pig.
>
> Cheerful regards,
> Bob
>
> P.S. I'd still love to hear from anyone who has tried,
> with part numbers if you've got 'em.
piclist-bounces@mit.edu wrote:
> P.S. I'd still love to hear from anyone who has tried, with
> part numbers if you've got 'em.
I recall trying ECS 20MHz resonators with no sucess. I also recall a problem with a SMD oscillator that stopped functioning at 30 Celsius (also from ECS).
Murata ceralock i've had better success with, but I haven't tried at that frequency. I think 16Mhz was the highest i've ever tried. Please keep us posted!
There is a very small price difference between a 20MHz
SMD crystal $0.7and a similar resonator $0.5 (Digikey prices)
I've used 16MHz trough hole resonators, unknow manufacturer,
very low price (about $0.05) but I will never recommend it.
However almost all computer CDrom units are using 24Mhz and 33MHz
SMD resonators with great success.
greetings,
Vasile
> Following up on my post from yesterday that got no
> resonses, I am assuming that everybody uses crystals
> at 20MHz because they tried a resonator once and it
> didn't work.
>
> That's what I did, but to keep parts cost down, I'd
> love to use a resonator in a new design if it's
> possible. It also appears that there are lots of new
> "high frequency" resonators to choose from.
>
> Do any of them work on PICs? Inquiring minds want to
> know. I guess I'll have to get an assortment and be
> the guinea pig.
>
> Cheerful regards,
> Bob
>
> P.S. I'd still love to hear from anyone who has tried,
> with part numbers if you've got 'em.
At 10:07 AM 6/15/2007, Bob Blick wrote:
>Following up on my post from yesterday that got no
>resonses, I am assuming that everybody uses crystals
>at 20MHz because they tried a resonator once and it
>didn't work.
Harold H has had good results with a particular brand of resonator at
20 MHz. IIRC, ECS was not reliable but Murata was. I'll have a look
through my archive to see what I can find out.
I don't do much at 20 MHz myself - projects with small (12c, 12f)
parts use internal RC (sometimes with a 32 KHz crystal if I am doing
time-keeping), most projects with larger PICs use a 4 MHz resonator.
>Subject: [PIC] OK, nobody has tried a 20MHz resonator?
> Following up on my post from yesterday that got no
> resonses, I am assuming that everybody uses crystals
> at 20MHz because they tried a resonator once and it
> didn't work.
>
> That's what I did, but to keep parts cost down, I'd
> love to use a resonator in a new design if it's
> possible. It also appears that there are lots of new
> "high frequency" resonators to choose from.
>
> Do any of them work on PICs? Inquiring minds want to
> know. I guess I'll have to get an assortment and be
> the guinea pig.
>
> Cheerful regards,
> Bob
>
> P.S. I'd still love to hear from anyone who has tried,
> with part numbers if you've got 'em.
We've been using surface mount & through hole 20MHz
resonators with built-in caps in several PIC-based products
for at least 6 years now with no problems.
> Following up on my post from yesterday that got no
> resonses, I am assuming that everybody uses crystals
> at 20MHz because they tried a resonator once and it
> didn't work.
>
> That's what I did, but to keep parts cost down, I'd
> love to use a resonator in a new design if it's
> possible. It also appears that there are lots of new
> "high frequency" resonators to choose from.
>
> Do any of them work on PICs? Inquiring minds want to
> know. I guess I'll have to get an assortment and be
> the guinea pig.
>
> Cheerful regards,
> Bob
>
> P.S. I'd still love to hear from anyone who has tried,
> with part numbers if you've got 'em.
Hi, Bob, I've used Murata through-hole parts with no
problems. Not so much for cost reduction, but for increased
shock resistance. I can dig out the p/n if you would like.
> At 10:07 AM 6/15/2007, Bob Blick wrote:
>>Following up on my post from yesterday that got no
>>resonses, I am assuming that everybody uses crystals
>>at 20MHz because they tried a resonator once and it
>>didn't work.
>
> Harold H has had good results with a particular brand of resonator at
> 20 MHz. IIRC, ECS was not reliable but Murata was. I'll have a look
> through my archive to see what I can find out.
>
> I don't do much at 20 MHz myself - projects with small (12c, 12f)
> parts use internal RC (sometimes with a 32 KHz crystal if I am doing
> time-keeping), most projects with larger PICs use a 4 MHz resonator.
>
> dwayne
Following up on this, ECS generally says if the resonator works with their
inverter based circuit, it's fine. muRata will, however, characterize a
resonator to your particular circuit. You send them a board, and they run
a bunch of tests and come back with a recommended part number. They also
have a "resonator finder" on their web site where you specify a chip and
it comes back with a part number. It doesn't cover every chip and does not
consider layout issues, but it's a start.
piclist-bounces@mit.edu wrote:
> We've been using surface mount & through hole 20MHz
> resonators with built-in caps in several PIC-based products
> for at least 6 years now with no problems.
May I ask what resonator mfg and their part number? what PIC was used with these resonators?
That reminds me that I once had a problem with a resonator (ZTT 20.00MZ
thru-hole 3-pin resonator) and since then I almost always put a guard ring
around the resonator/crystal on the PCB.
On Friday 15 June 2007 12:59, Harold Hallikainen wrote:
> ... It doesn't cover every chip and does not
> consider layout issues, but it's a start.
That was the part that soured 20MHz resonators for me.
I had trouble with those failing to start on a
PIC16F74 many years ago.
Maybe the newer PICs have better oscillators?
Thanks to everybody for the input. I have several
different brands and part numbers on the way and will
test them all on a PIC16F884 at different temperatures
and report back.
--- Vasile Surducan <KILLspampiclist9KILLspamgmail.com> wrote:
> There is a very small price difference between a
> 20MHz
> SMD crystal $0.7and a similar resonator $0.5
> (Digikey prices)
Hi Vasile,
Thanks for replying.
The resonator can come with caps builtin. The board
area and assembly of one part compared to three is
some of my consideration as well.
>> 3-pin through-hole versions w/caps are ECS also.
>> Part # ZTT-20.00MX.
>
> That was the part that soured 20MHz resonators for me.
>
What oscillator modes are you using? Do resonators need to
change mode with speed like crystals? Do they need to NOT
change modes?
--- William Chops Westfield <RemoveMEwestfwTakeThisOuTmac.com> wrote:
> What oscillator modes are you using? Do resonators
> need to
> change mode with speed like crystals? Do they need
> to NOT
> change modes?
Of course I tried both XT and HS modes. But neither
started reliably.
The real pain is smd footprints - 20MHz resonators
with builtin caps come with pitches of 1.0, 1.2, 1.35,
1.5 and 2.5mm, so I limit myself to a smaller group of
parts after I lay out the PCB.
On 6/15/07, Bob Blick <spamBeGonebbblickspamBeGonesbcglobal.net> wrote:
> --- William Chops Westfield <TakeThisOuTwestfwEraseMEspam_OUTmac.com> wrote:
> > What oscillator modes are you using? Do resonators
> > need to
> > change mode with speed like crystals? Do they need
> > to NOT
> > change modes?
>
> Of course I tried both XT and HS modes. But neither
> started reliably.
Even I know you know, one suggestion:
Keep the SMD area of the resonator very clean.
The OSCOUT route should have minimal parasitic capacitance
(add that to the 7pF of the PIC IO pin and to the resonator built in
capacitor value)
Measure and store the output level of the oscillation on your TEK
using a 10:1 proble and see the amplitude variation for all your
tested resonators.
The problem is the internal capacitors value and the Q factor of the system.
Check the OSCIN IO trigger level for your particular Vdd, and see the
lower limit of the amplitude oscillation (which is decreasing with
temperature increase).
The SMD resonators with good Q are packed in some white ceramics.
Heating the package above 250C for more than 1 minute will split the
package in two, so you may see how looks inside, maybe taking a look
under the microscope will be interesting. Also if you may cut from the
internal capacitor surface (the one connected on OSCOUT) you will see
interesting results.
Of course those operation are good only for understanding the problem
and maybe considering for the future that using 3 components is safer
than using one.
sbcglobal.net>
