a little help with capacitive power supply values

[Byte_Butcher]

Yep, like anonymouse said, fuse in the neutral is bad!

If you really *must* build a transformerless supply, goto the document I referenced in post #17, scroll down to page 10, figure 10 and build it just like that. They've already figured out the right way.

Correctly designed transformerless supplies are plenty safe, especially since the commercial devices that usually contain them are invariably well sealed in a well insulated (plastic) housing to keep curious fingers away. Lots of insulation between the device and the user is the key.

But most of my projects involve a lot of testing, fiddling, finger poking, and experimentation (with the guts open and exposed), so most of the time I much prefer a good isolated supply.
I buy little transformers 10 at a time for under $3 each and don't have to worry about what parts are "hot" when I connect my 'scope ground or poke at a part with my finger.

If the size and weight of your project is REALLY critical (but if you're sending it up in a rocket, it probably won't be mains powered), or if you're building a product by the 1000's where saving $1 worth of parts becomes a pretty big deal, then maybe xformerless is the way to go. But for most applications (especially hobbiest) I have to ask... Why? What's to be gained that offsets the risks?

So... why do you want a xformerless supply? What benefits does it have for your project?


steve

[Melanie]

You're getting there...

Loose the Neutral Fuse as suggested.

Your Triac isn't completely wired... it's a three legged device. The LIVE goes through the TRIAC from the Supply Connector and goes out to the LIVE pin on the LOAD Connector. The NEUTRAL is common between the Supply and the Load, correct as you have it.

For those keeping an eye on this thread, if you use an isolated Supply for the PIC, it's low-voltage side will become LIVE the moment you connect it to the TRIACs gate. If you want your PIC to be isolated, then you will need an Opto-Isolator (or Pulse Transformer) of some sort between the Gate and the PIC.

Again, going back to an earlier question, yes, zero-cross is important - for many reasons (even if it's just to keep the peace with your neighbours). Go read up about it.

[dhouston]

If you really *must* build a transformerless supply, goto the document I referenced in post #17, scroll down to page 10, figure 10 and build it just like that.

Steve, isn't that the same document Dennis referenced in post #1 of this thread?

[dhouston]

Dennis,

You might also find Microchip AN236 helpful. It has a capacitive supply, a ZC detector and a triac.

• http://www.microchip.com/stellent/id...pnote=en012050

And there are more schematics of X10 modules on NASA engineer Dr. Ed Cheung's site.

• http://www.edcheung.com/automa/circuit.htm

[Dennis]

Hi All
I'm still busy trying to get my head around all the valuable info you have offered and definitely am taking the time to read through as much as my already overloaded mind can take !

Melanie and anonymouse thanks for the tip regarding the fuse on the NEUTRAL line :-) , I have updated the schematic to reflect only a single fuse on the LIVE line !
And thanks again Melanie for the info regarding the OPTO and the TRIAC,
the reason the TRIAC is open ended is because I have not decided what value to use for the TRIAC, since I think this may be dependent on the choice of PIC (and PORT) ? The designs I have seen so far seem to indicate somewhere between 470R and 1000R for the TRIAC resistor (R3 in my diagram).
Two things should me taken into consideration here :
1. the TRIAC gate currents (MIN and MAX)
2. the PIC pin allowable current
PLEASE CORRECT me if I am wrong!
In the meantime I am reading through this as a start!
http://www.microchip.com/stellent/id...pnote=en011202

What's of concern is that Microchip have made no mention (that I can see) of the differences between a COMMON NEUTRAL or COMMON LIVE design in the appnotes !!

Steve,thanks again for the tip especially for the transformer info ... READ the posts to date very carefully please... we have already established that the Microchip design is a COMMON LIVE design which 'floats' 5V below either 120VAC or 220/230/240VAC and that it would be a whole lot 'safer' to rather 'float' 5V above NEUTRAL, hence the changed design thus far.
Also as Dave pointed out, I did mention in post 1 that I would like to use a transformer-less PSU.

Dave .. thanks so much for the continuous valuable x10 references as well as the tips thus far.
That appnote sure has a lot of info and someone went through a huge amount of work to setup the application to get it to that stage I'm sure (especially looking at the section for the tuned circuit replacement of the usual x10 120KHz tuned transformers), is that mains injection being done purely with a resistor and a cap ?
I saw a project a few days ago where the designer is injecting a 135KHz pulse into the carrier and was wondering if that could be done with the pulsin pulsout command instead? Will see if I can find it for you since I think it may interest you as well.

Dave and Melanie (and anyone else)... regarding a zero-crossing sensor (which may be needed for dimming), where would you best situate it in this circuit ?
I have seen various designs indicating it in different places, and most have one of the PIC pins connected with a resistor in series directly to

For all those following this thread, please note that nothing has been built as yet and the attached schematics are merely used for the discussion.
Please also take note of ALL the safety warnings!

Thanks to everyone for all the info so far!

Kind regards

Dennis

[Dennis]

I have been reading through this appnote http://ww1.microchip.com/downloads/e...tes/40171a.pdf and noticed the presence of two extra capsacitors namely C1 and C2 in the appnote schematic ..what is the function of these in this instance?

I have updated my design (see attached file) to include a zero-crossing sensor and have connected the TRIAC.
My concern is the value of the TRIAC resistor and that's the reason for the ???.
If there are any suggestions/concerns or changes to be made please let me know.

Kind regards

Dennis

[Dennis]

Forgot to attach the update.

While checking out various design for PSU's I noticed one person remarked that he might move the Varistor put the varistor after the 47R, so it doesnt have to short out the power station on spikes, same with the 20M resistor.
He also suggested he would not use a 20M but a rather an attenuator, say/guess 220K and 5k to ground.
Any thoughts around this statement ?

PLEASE NOTE - I have also added a WARNING for anyone who might stumble across this !!

Kind regards

Dennis

[Dennis]

Trying to attach the schematic again (renamed this time)


Dennis

[dhouston]

I saw a project a few days ago where the designer is injecting a 135KHz pulse into the carrier and was wondering if that could be done with the pulsin pulsout command instead?

Search the forum for Anand Dhuru's posts on building his own X10 modules for 230V. I believe he is using the hardware PWM to generate 120kHz. Also, Darrel Taylor made some modifications to the PBP XOut command that might be of use.

[Dennis]

Hi all

Dave .. thanks once again for the extra info and tips :-) they always provide good reads and info :-)

Melanie,Steve and anoymouse...

While re-reading the Microchip appnotes, this one in particular

I noticed some interesting points...
Firstly this one regarding the current draw,

The PIC12/16/17 microcontrollers draw a maximum of 10 mA, even at the highest frequency and voltage of operation, therefore low current availability is not an issue. AC line voltage isolation can be addressed by using MOVs or transient suppressors on the PIC12/16/17.

and then something strange about the FUSE on the NEUTRAL LINE ???

In most applications, the output voltage should be reg-
ulated. Figure3 shows a diagram for a practical circuit
where a +/- 5V regulated output is generated. Note that
the neutral is connected to ground through a fuse. This
+5V
would guard against improper AC wiring.

Any comments on these statements?

Kind regards
Dennis

[mackrackit]

Where did you get the second quote? The one about the neutral.

Have you checked your local electrical codes about the proper way to wire something for the system you have? Might clear up the neutral thing for you. Here the ground and neutral are "bonded in the box" not at the device. I can not see yours being any different, I am sure you have a transformer with a center tap providing power to you.

BTW, you never said why you want to do this or what you think the benefit will be.