# Shared power and data pair

Discussion in 'Electronic Design' started by Greg Neff, Dec 9, 2003.

1. ### Greg NeffGuest

I have to supply both power (48VDC @20ma) and data (1.25Mbps) over a
single pair. The requirement I'm having trouble with is that the 48V
supply shall provide a minimum 12K output impedance in the 0.5-2MHz
range. A 3.9mH choke is the obvious answer, but the SRF of chokes
this size is way too low. Does anyone have any ideas to help me out
here? I could string a bunch of small chokes together in series, but
I would like a better solution. I am thinking virtual inductor, but I
need some pointers on this...

================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.

2. ### John WoodgateGuest

I read in sci.electronics.design that Greg Neff <>
Can you not use an electronic current source?

3. ### John PopelishGuest

Putting an inductance of 4 millihenries (with a self resonant
frequency at or above 1 MHz) in series with the 48 volt supply would
do the trick. Finding such an inductor might be tricky. If the data
is differential (and it would be good if it were) would allow the
inductor to be split into two 2 millihenry units, one in series with
each line.

4. ### Spehro PefhanyGuest

That was my first thought. Depends on the swing of the signal and how
much voltage drop he can stand to the power.. but if we make favorable
assumptions a discrete current source/sink could do it easily.

Best regards,
Spehro Pefhany

5. ### Greg NeffGuest

I have to supply 48VDC +/-10%. The maximum load will be 20ma.

================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.

6. ### Greg NeffGuest

The data is differential and transformer-coupled to the pair. A
capacitor will keep the transformer dry.

A SRF of 1MHz is right in the middle of the 0.5MHz to 2MHz range that
I'm worried about. This is what I'm trying to avoid.

================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.

7. ### John PopelishGuest

A 4 mHy inductor with a self resonance at 1 MHz has 12k ohms impedance
at both .5 and 2 MHz (inductive at .5 MHz and capacitive at 2 MHz),
and a higher impedance at all frequencies between, with a peak at 1
MHz. Is there a problem with this?

8. ### Greg NeffGuest

Well, I'm not sure. The data is Manchester encoded, so it has
1.25MBps and 0.625MBps frequency components. I don't want the
inductor to introduce any peculiar effects due to the change in type
of impedance at these frequencies. Maybe I'm being paranoid, but I
want to be safe.

================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.

9. ### John PopelishGuest

The minimum impedance spec should cover this possibility. If the
transmission line has an impedance of a few hundred ohms or less, and
the differential receiver terminates it near that over the necessary
spectrum, the resonance of the series inductance should produce
acceptable distortion as long as its impedance is high enough compared
to the transmission line impedance.