[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

FW: Power supply quality (Leif Åsbrink Oct 15 2002)




-----Original Message-----
From: owner-linrad@xxxxxxxxxxxxxxxxxxxxx
[mailto:owner-linrad@xxxxxxxxxxxxxxxxxxxxx]On Behalf Of Leif Åsbrink
Sent: Tuesday, October 15, 2002 2:13 AM
To: linrad@xxxxxxxxxxxxxxxxxxxxx
Subject: [linrad] Power supply quality


Hi All,

I do NOT agree here:
  
> I agree with both of you on the importance of power supply quality.
> What I can get is a vey high grade medical equipment commercial power
> supply with 0.05% regulation, less than 5 mV ripple, and essentially no
> leakage.  It is of course a linear supply and supplies + and - 15 volts.
> For this application I don't want a switcher. 

No need for high quality here. A supply with 5% regulation is fine and 
you may even use unregulated supplies. 

If there is ripple on the DC voltage, you will get two kinds of problems.

Firstly the ripple will couple directly to the audio output. If the
ripple voltage is 100mV peak-to-peak on one or the other of the supply
voltages, the spurs symmetrically around 48kHz will be at -110dB compared
to a near saturating signal. 

Secondly, the ripple will modulate strong signals. The sidebands 
surrounding a really strong signal, one that is 1dB from saturation, 
will also be at -110dB with 100 mV on one or the other DC supplies.

If ripple is applied to both supplies the spurs around 48 kHz double
(+6dB) or cancel depending on the phase of the ripple voltage. 
The modulation sidebands of a strong signal increase by about 12 dB
regardless if the ripple on the +15 supply is in phase or out of phase
with respect to the ripple on the -15V supply.

Ripple up to 100mV is unlikely to ever cause any problem. Real signals
have more bandwidth than 120Hz at the -100dB level:-)

If you want to use Linrad as a high quality spectrum analyzer to look for
close range unintended modulation you should make sure the ripple is well 
below 10 millivolts.

I suggest you use something rugged but cheap for power supply. If there 
is more than 10mV ripple, a really BIG capacitor 47000uF should fix it. 

 73

Leif  /  SM5BSZ


LINRADDARNIL