How to measure crystal ESR

[Jeff]

I want to sort a bunch of 20MHz crystals by ESR. Is there a lab setup
someone can suggest that will work?

Thanks,

Jeff

 

[Andy]

Andy writes:

I have used a simple setup to do this. It uses the same principle
as
crystal resistance meters.

Take a signal generator and pad it down to give about a 1 ohm
source resistance. From this source, put the signal thru a crystal in

series and then a 20 ohm resistor to ground.
Tune the generator freq for MAXIMUM voltage on the 20 ohm
resistor.
This will be very very very sharp. Then substitute a resistor for the
crystal to give the same voltage reading. The resistor value that
gives
the same reading as the crystal did is equal to the crystal esr.

You really need a tuneable generator with a good vernier to do
this,
but it works.

The reason I chose a 20 ohm resistor was because this was a
ball park value of what I expect a 20Mhz crystal esr to be.

The reason I padded the generator down to 1 ohm was becaues
I didn't want any interaction with the generator output impedance, and
I had a lot of signal to work with anyway......it just seemed like a
good
idea at the time.

I used a spectrum analyzer across the 20 ohm resistor as a
detector,
cause it was handy, and plenty sensitive for the padded down generator
signal. A Booton RF meter would work well with good sensitivity,
also.
Actually, since I was using a substitution method, almost anything
would
work..

Since I used the substitution method, all of the extraneous
variables
and impedance mismatches cancel out.......

Good Luck,

Andy

 

[doug dwyer]

I suggest a series resonant oscillator such as a butler tuned to the
crystal frequency with a AC coupled resistor of say 50 ohms in place of
the crystal. You will observe that as the supply is reduced the
transistor current reduces and the gain reduces. Replace the resistor
with a crystal reduce the supply till the oscillation stops Find the
resistor value that gives the same starting voltage as the crystal. Note
that the technique will show up crystals with high start up resistance
and low running resistance (second level)


In message <1122874338.311620.74020@g44g2000cwa.googlegroups.com>, Andy
<andysharpe@juno.com> writes

Andy writes:

I have used a simple setup to do this. It uses the same principle
as
crystal resistance meters.

Take a signal generator and pad it down to give about a 1 ohm
source resistance. From this source, put the signal thru a crystal in

series and then a 20 ohm resistor to ground.
Tune the generator freq for MAXIMUM voltage on the 20 ohm
resistor.
This will be very very very sharp. Then substitute a resistor for the
crystal to give the same voltage reading. The resistor value that
gives
the same reading as the crystal did is equal to the crystal esr.

You really need a tuneable generator with a good vernier to do
this,
but it works.

The reason I chose a 20 ohm resistor was because this was a
ball park value of what I expect a 20Mhz crystal esr to be.

The reason I padded the generator down to 1 ohm was becaues
I didn't want any interaction with the generator output impedance, and
I had a lot of signal to work with anyway......it just seemed like a
good
idea at the time.

I used a spectrum analyzer across the 20 ohm resistor as a
detector,
cause it was handy, and plenty sensitive for the padded down generator
signal. A Booton RF meter would work well with good sensitivity,
also.
Actually, since I was using a substitution method, almost anything
would
work..

Since I used the substitution method, all of the extraneous
variables
and impedance mismatches cancel out.......

Good Luck,

Andy

--
dd