Non isolated buck power supply

[dlmuns]

After reading thread 248-226082 I am concerned about a design I am working on. It is a non-isolated off line buck converter for LED lighting. I understand the dangers of such a design, the system it powers has interlock switches to power it off when the unit is opened and is fully isolated from earth ground. It uses a full wave bridge rectifier to create the DC bus voltage. Member Skogsgurra replied "Using a single diode rectifier will make the Neutral available to the user. Using a bridge rectifier will not." Could someone explain why this would be so ?

 

[ScottyUK]

If you use a bridge rectifier the neutral will alternately be connected to the positive and negative rails depending on the polarity of the half cycle. If you aren't isolating the supply then it is difficult to use the DC supply for anything involving the neutral because they have no common reference.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

Draw a bridge rectifier circuit.

Look at it.

Try to picture where a ground could be.

You'll see the problem. It's pretty graphic.

If you proceed in this exercise to calling one of the AC lines the neutral and then bond it to earth you'll see neither of the rectified side's + or - can be successfully earthed.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[dlmuns]

Thank you for your replies. This system runs from 120/240 with a ground and a neutral. The neutral is bonded to earth at the distribution transformer (center tapped) for the system. The bridge rectifier of the LED power supply is connected between one phase of the 120 VAC and neutral. The negative side of the bridge rectifier floats. It is not connected to earth in any way. If you follow the rectifier current paths neutral is either sourcing or sinking current depending on the phase of the line voltage. If you measure between neutral and earth it is 0 volts AC, well not exactly 0, since there will be a small voltage drop across neutral depending on the amount of current flowing and the size and length of the neutral wire. Which is why bonding neutral to earth at the system when its already bonded somewhere else is verboten. Anyway other than the power supply being a shock hazard I still don't see a problem. I have tested this system in the lab and in the field and haven't seen any issues, am I missing something ? Are there potential issues ?

 

[IRstuff]

That's not necessarily an easy question to answer. However, in some cases, hot and neutral can be miswired, in which case, you'd have hot attached to chassis, for example, which would be REALLY bad.

I nearly got my eye clobbered by an exploding LM741 op amp when I was getting ready to scope a circuit by connecting the ground lead to the circuit ground. BAM! Turned out the circuit had one leg of the AC tied to chassis, and it wasn't the neutral that day.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[itsmoked]

Of course you're not supposed to bond the neutral to earth at your power supply!! That's why I called it an exercise.

At this point I'm not sure what you're asking.

We explained you can't ground either side of the rectified output when not using isolation. In those cases the entire product needs to be insulated. Often TVs do this.

A non-transformer design has another issue. Normally the inrush to charge the caps is limited by the transformer saturating. Without the transformer you may need to add something to protect the diodes or use vastly larger diodes than otherwise.

As Scotty mentioned using a FWB is royal PIA for measuring and scope work.

Other than those points there isn't any particular problem.

Keith Cress
kcress -

http://www.flaminsystems.com