From: Sunil Kumar (email@example.com)
Date: Fri Oct 13 2000 - 08:15:09 PDT
George and Michael.
If probe is picking up noise then I whould see same amount of increase
everywhere, I mean in the whole clock tree. But I am getting almost
negligible increase at the top of the tree. While before PECL-to-TTL
buffer, the increase is quite a lot. That's why, I think probe is not
picking noise, it's due to noise generated in the planes.
When I first observed the effect, I thought of two reasons: current drawn
by the fans; and EMI genarated by the fans. To isolate the first cause, I
used separate power sources for fans and the system. I found the same
amount of increase in the jitter. Hence I concluded that this effect is
not due to the current drawn by the fans.
I used one external fan and switched off the internal fans. The fan was
facing the front side of the system, which was open. Now again I observed
increase in jitter. Hence it is not due to vibrations/mechanical coupling.
On Thu, 12 Oct 2000, Michael Nudelman wrote:
> I tend to agree with somebody here, who suggested your probe actually
> picks up noise. I don't see how EMI from fans can badly affect twinax
> cable, unless your Ground Connection is re-e-e-eally screwed up (and it
> takes some screw-up plus some ingenuity and luck....).
> George Borkowicz wrote:
> > One more per cent. Try to replace the fans with a resistive load
> > of the same power (bulbs, power resistors, what have you) right
> > at the fan end of the fan wires.
> > gjb
> > -----Original Message-----
> > From: Sunil Kumar [SMTP:firstname.lastname@example.org]
> > Sent: Thursday, October 12, 2000 2:42 AM
> > To: email@example.com
> > Subject: [SI-LIST] : EMI due to fans (fwd)
> > Hello everybody.
> > I am measuring cycle-to-cycle jitter in clock in a system, where
> > clock
> > (PECL, 50MHz) is going from one PCB (card) to other through
> > twinax
> > cable. In the second card clock is converted into TTL and then
> > distributed
> > to various ICs. I am looking at the clock just before the
> > PECL-to-TTL
> > convertor. Both the cards are in the same backplane. The system
> > is
> > equiped with a set of fans for cooling, which are kept below the
> > card
> > frame. Now what I have observed is, when the fans are not on,
> > jitter is
> > much much less in comparision of the jitter when fans are on. I
> > think, the
> > fans are generating EMI, which is disturbing the planes in the
> > backplane
> > and the cards, and hence jitter is increased. Any explanations?
> > One more thing which I have observed is, the PECL-to-PECL clock
> > fanout buffer (no PLL inside) is reducing jiiter (with fans off),
> > i.e.
> > jitter at input is more than that at output. Can anybody explain
> > this?
> > Is there any PECL-to-TTL clock fanout buffer which adds very low
> > jitter? The PECL-to-TTL buffer which I am using, is adding lot's
> > of
> > jitter.
> > Thanks
> > Sunil Kumar
> > Senior Research Engineer
> > ATM Group
> > C-DOT
> > Bangalore-560052
> > INDIA
> > **** To unsubscribe from si-list or si-list-digest: send e-mail
> > to
> > firstname.lastname@example.org. 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
> > ****
**** To unsubscribe from si-list or si-list-digest: send e-mail to
email@example.com. 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 - 14:29:45 PDT