Transformerless power supply: Can I lose the opto?

[NogginBoink]

I'm sketching out a circuit that uses a resistive power supply as described in Microchip's Application Note 954.

One of the advantages of a transformerless power supply is that it eliminates one of the most expensive parts on the board: the transformer. One of the big disadvantages is that even the low-voltage parts are referenced to mains voltage, which could be lethal if you touch a part while the doohickey is running.

My circuit would be a DMX node with integrated SSR's. As I looked at my schematic, with the traditional SSR circuit as the output, I started asking myself... why do I have the optoisolators in there? If my low-voltage stuff is going to be referenced to mains voltage, would it be possible to trigger the TRIAC directly from an I/O pin of the PIC chip? That would save on parts count, too!

[mrpackethead]

What you are talking about doign its entirely possible.. Look at microchips application note TB094 and they do exactly what you are looking for.

How are you planning on providing isolation between the dmx input and the parts at mains potential?

[NogginBoink]

That's a good question. Got any good ideas for me?

Lemme read through this appnote and see how closely my schematic matches theirs. Thanks for pointing this out to me; I didn't know it existed!

[RJ]

Be careful if you go that way!

I also gave quick thought about this at one point on the SSR4 but decieded agianst it.
I was not so worried about the shock hazard as the SSR's are a shock hazard themselfs.

But it would only take some moisture to get in your unit and it could be a bad day for you. And you will have cost involved in isolation for the DMX that makes a $2 Transformer become cheaper IMHO.

If you wish to think in that direction I have found these to be small and cost effective and avalible from digikey.

http://www.pulseeng.com/products/dat.../LT2006_04.pdf
The one I am using in the SSR4 is $3 in singles and with just 10 they drop to $2.16 and 25 gets you under $2

I am not saying you can not do it cheaper by doing it this way.

It is possible but it is a design you should think hard about before you go that route. I say this in honest good will for you with no other motive.

I am not saying do not do it, but be careful doing it.

RJ

[P. Short]

I do not recommend a transformerless design. My theory is that if you need to ask these questions, you do not know enough to do this safely. And attempting to learn it through q&a over a forum is not a reliable to get this knowledge.

--

Phil

[NogginBoink]

Yeah, I'm thinking you guys are right. I hadn't thought through the RS485 bus isolation; that's kind of a brick wall here.

RJ, thanks to the pointer to the cost effective transformers. That price point takes some of the sting out of moving back to a design with a transformer.

Still, though... the cost savings, if it could be done, would be nice. A transformer, bridge rectifier, and voltage regulator add quite a bit of cost.

Rats.

[Wayne J]

Sometimes you have to draw the line between safety and savings. I personally am not interested in saving a few dollars while pushing things toward a potentially hazardous design. Remember we want to be around to watch our displays and enjoy everyone else's.

Lets try and keep to designs that are safe for all of us blinky flashy nuts. ;)

[omzig]

Sure it can be done!

[NogginBoink]

Sometimes you have to draw the line between safety and savings. I personally am not interested in saving a few dollars while pushing things toward a potentially hazardous design. Remember we want to be around to watch our displays and enjoy everyone else's.

Lets try and keep to designs that are safe for all of us blinky flashy nuts. ;)

While I understand and agree, the flip side is that saving a few dollars per board can become quite a few dollars when multiple boards are involved. I want to be as cost sensitive as I can (although I have abandoned the transformerless design for now).

[mrpackethead]

What makes the 'transformerless' design any more or less safe.. Both designs have mains potential on a board.. Both would have isolation between the input side (either dmx or triac control), and the mains potential.

In both designs, if you put your fingers in the wrong place, its all badness!

My point is
"BOTH DESIGNS ARE POTENTIALLY DANGEROUS"

So shall we

Lets try and keep to designs that are safe for all of us blinky flashy nuts