smps heater supplies

[pre65]

My 12v 30A SMPS have an earth connection to the AC input, and the 833a cathode is grounded, well nearly, just a 1R resistor. But I have used them grounded on earlier iterations.

[Nick]

Yep, the frame type SMPS I have used have been isolated (and the spec gave a rating for that isolation) but reusing a existing supply all bets are off.

[Paul Barker]

Somebody(maybe Paul) has cut the earth wire at source on the unit that I have.....

Yes I cut them all as they are for stacking.

[ed]

did a bit of measuring today. The only consistent thing I could find was that they all seem to have 2 signatures......

in order of photos:

HP laptop 19v 3.5amp
at 40 hz 9mv RMS
at 99 khz 17mv RMS

Samsung netbook 19v at 2.1amps
at 430 hz 2.8mv RMS
at 67 khz 10.0mv RMS

Samsung 5v 700ma(used for AVR chip testing)
at 8khz 2mv RMS
at 12 mhz 6mv RMS

all frequency and voltage figures reported by UNI-T scope

[ed]

........

[ed]

.....

[JamesD]

Brilliant - and thankyou.. I'll have to think about what it means...

J

[jack]

did a bit of measuring today. The only consistent thing I could find was that they all seem to have 2 signatures......

in order of photos:

HP laptop 19v 3.5amp
at 40 hz 9mv RMS
at 99 khz 17mv RMS

Samsung netbook 19v at 2.1amps
at 430 hz 2.8mv RMS
at 67 khz 10.0mv RMS

Samsung 5v 700ma(used for AVR chip testing)
at 8khz 2mv RMS
at 12 mhz 6mv RMS

all frequency and voltage figures reported by UNI-T scope

9mV on 19.5V is 0.05% ripple or 67dB
2.8mV on 19V is 0.15% ripple or 77dB
2mV on 5V is 0.04% ripple or 68dB

Need to keep these figures in perspective - though there is *some* ripple in the audio band, is it audible at all? Its very VERY low level noise.

A simple LCR or super-capacitor arrangement could reduce the ripple further.

(from data on the web...) The typical peak to peak output voltage ripple for a linear supply might be 1000x or 60dB less than the output DC level. So for a 5V output supply the typical peak to peak ripple voltage might be 5mV. A switching supply typically sees about a 100x or 40dB reduction. With careful work, this can be improved to 10,000x or 80dB reduction.

I'd de more concerned about RFI/EMC - i.e. radiated noise - and any cr*p put back down the mains by the SMPS - remember, they are directly coupled to the mains - just a full bridge rectifier after the mains input - very good (HP/Cisco etc.) SMPS supplies will have symmetric multi-stage filtering between the mains and the input bridge - this is as much to stop cr*p getting onto the mains from the SMPS as it is stopping mains-borne cr*p getting into the SMPS. Budget supplies have the absolute minimum they can have to scrape through their certification, if indeed that "certification" is genuine and not just stamped on for good looks... I know of folk who have had electrical goods manufactured in China and who have been asked what stamps/labels they want printed on the cases - all fake certifications of course...

If you are using a "budget" SMPS, I'd look closely at the noise its putting back onto the mains and consider using a nice multi-stage common & differential mode filter between the supply & mains input - I like Schaffner units, e.g. the FN2020-3-06, often found on eBay... remember, these filters are symmetric so they stop noise going either way through them - you can get decent multi-stage ones built into IEC connectors too but you may want to have a stand-alone unit to prevent noise coming back from the SMPS getting into your standard xfrms for the HV supplies etc.

[nayaanika]

Hi my computer's SMPS is a making sound at regular intervals, how can I rectify the same. And also the sound diminishes slowly.
_______________________________
fashion police ~ fashion n you