# RE: [SI-LIST] : PECL oscillator termination

From: Ingraham, Andrew ([email protected])
Date: Thu Nov 09 2000 - 20:04:22 PST

> Here's a question on 3.3V PECL osc. termination.
>
> #1 Is there any recommended value of the coupling cap?

Large enough to pass the lowest frequency without waveform tilt. Since it's
an oscillator, you have a nicely defined "lowest frequency". The reactance
should be well under 50 ohms at that frequency.

But not so large that the capacitor's resonance gets in the way of your
signal at the (edge-rate-dependent) harmonics.

> #2 How to size the resistors to DC-bias the PECL outputs? Is there also
> any recommended range of size?

If each output switches by 0.8V, and if the load is 50 ohms differential,
and the traces are matched to the load, then 2*0.8V/50ohm = 32mA must be
switched by each output. Size the pulldown resistor to provide more than
32mA. If it's less, or very close, then the transistor turns off on the
falling edge and might cause a plateau. You want a few milliamps out of the
transistor even when it switches low.

If the load is 50+50 = 100 ohms, the current being switched is only 16mA.

Stay within the current limits of the output transistors, and the power
capability of the resistors.

More pulldown current does shift the levels down slightly, OK because it's
differential.

> #3 In the case of #2, the resistors are supposed to be grounded, then
> what
> if I make them in series?

In series with what? I don't understand this either. They must go to
ground, or the circuit doesn't work.

If you mean add two more resistors, each one in series with a line to the
caps, then you get a voltage divider and less signal reaches the load. You
would be terminating the wires at both source and load ends, which might be
good, but (P)ECL works just fine with the source end NOT terminated, and the
loss of signal amplitude might outweigh the benefits.

Andy

**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected] In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****

This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:30:02 PDT