[SI-LIST] Re: How to couple noise to the PDN of oscillator
- From: Jory McKinley <jory_mckinley@xxxxxxxxx>
- To: Istvan Novak <istvan.novak@xxxxxxxxxxx>, Zhangkun <zhang_kun@xxxxxxxxxx>
- Date: Thu, 15 May 2008 09:03:14 -0700 (PDT)
Also, You can also think about a parallel device (switch) connected to
the power supply that is DC coupled. The device can be a single FPGA
with outputs connected to the power supply (DC coupled) set somewhere
in the 50 ohm range. You can tune the amount of noise by turning off
and on the outputs.
-Jory
----- Original Message ----
From: Istvan Novak <istvan.novak@xxxxxxxxxxx>
To: Zhangkun <zhang_kun@xxxxxxxxxx>
Cc: si-list@xxxxxxxxxxxxx
Sent: Thursday, May 15, 2008 6:58:51 AM
Subject: [SI-LIST] Re: How to couple noise to the PDN of oscillator
Zhangkun,
The 0.1-ohm impedance on your oscillator PDN requires 1A to create 0.1V
injected noise. You
can generate 1A AC current with a suitable opamp, or by a transformer or
by a separate power
amplifier. Each will have its own challenge. Opamps down like to work
across low load
impedance, and if you you a transformer to raise the load impedance
presented to the opamp
output, you likely to run into slew-rate limits at higher frequencies.
Regardless of how you
inject the noise, a major limiting factor is the inductance connecting
to your final 0.1-ohm rail.
If my math is correct, a 25MHz corner frequency with 0.1-ohm requires a
nanoHenry inductance
or less. As long as you stay with sinewaves and dont need a flat
transfer response from your
noise source to the 0.1-ohm PDN point, you can use brute force (a power
amplifier), but it
takes a lot of power: to generate 1A with a 50-ohm source (without
transformer), you need
50V. Assuming rms voltages and currents, this corresponds to 50W.
Another option to
consider is a two-stage step-down: first use a transformer to step from
50 ohms say to one ohm,
followed by a resistive step-down from 1 ohm to 0.1 ohm. The 1W
resistive divider can be
small size and wide band. Similarly the 50:1 impedance matching
transformer can afford
up to 10nH inductance on the output connection, probably allowing you to
create a flat
transfer response over the entire frequency range you are interested
in. This would allow you
to create a constant output level without manual or automated level
corrections, plus you
could use waveforms other than sinewave without major distortion.
Regards,
Istvan Novak
SUN Microsystems
Zhangkun wrote:
> Dear all
>
> Now we are studying the effect of power noise on jitter. We meet a problem,
> how to couple noise to the PDN. For example, we want to couple noise of
> 25MHz into the PDN of oscillator. However there is impedance dismatch. The
> impedance of signal source is about 50ohm and that of PDN is about 0.1ohm.
> The energy would be reflected.
>
> Any advice would be helpful
>
> Best Regards
>
> Zhangkun
> 2008.5.14
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
