a little help with capacitive power supply values

[Dennis]

Adam Thanks for that :-)
Saw it when it was first posted ... I don't want to use a bride rectifier or transformer and 1 amp is way more that I need :-)

All I want the pic to be able to do is act as and intelligent switch

AC IN --->> PIC >>> LOAD

Make sense ?

Kind regards

Dennis

[amgen]

the capacitor reactance (resistance for ac voltage) is
1/(2 X pi X freq X Cap(micro-farad)) ohms,

for 50MA from power supply, 220V/.050 = 4400 ohms....
the required resistance provided by cap for 50HZ at 220V.

then,
Cap= 1/(314 X 4400) equal aprox .7 micro-farad cap (>250 VAC rated)
works out to about .15 micro-farad per 10MA.
See attached diag.

[Dennis]

Amgen

Thank you that's exactly what I'm asking for !

I have been through the appnote and what's confusing me is how you got he value for the resistor R1 ??
You said ....

the capacitor reactance (resistance for ac voltage) is
1/(2 X pi X freq X Cap(micro-farad)) ohms,

For 50MA from power supply, 220V/.050 = 4400 ohms....
the required resistance provided by cap for 50HZ at 220V.

How did you get the value for this ...

Cap(micro-farad)) ohms

Kind regards
Dennis

[Melanie]

Just Transpose the formula that you've got... if...

Xc=1/(2 pi F C)

then...

C=1/(2 pi F Xc)

Xc is the AC Resistance (Reactance) that you want in Ohms, C=Capacitance in Farads and F is the Frequency in Hertz.

Say you wanted 50mA at 110v 60Hz (way too high for driving a PIC and a TRIAC say for Lamp Control - more like 15mA required), then V/I=R (where in our case R is actually Xc)...

110/0.05=2200

So we know that Xc must equal 2200... therefore applying the transposed formula...

1/(2 x 3.14 x 60 x 2200) = 1.2uF

Theoretically, that's great... but practically you'll discover that Capacitors don't come in all the myriad of values that Resistors come in, so you probably only have easilly available 0.1uF, 0.22uF, 0.47uF and 1uF. Work out which one best suits your current requirements and design your circuit to match it. I would strongly discourage you from desgining Transformerless Power Supplies with Supply Capacitor values greater than 1uF.

[Byte_Butcher]

Melanie pretty much said it all, but there's schematics and math formulas for transformerless supplies in this Microchip appnote:

http://ww1.microchip.com/downloads/e...tes/00954A.pdf



steve

[Dennis]

Hi all

Thanks to everyone for their help so far ..

Steve for the tongue-tip-test tip and for the latest one
but I am already referring to that document in post number 1 in this thread .. and that's the source of the confusuion.

Amgen thanks for the bridge-rectifier based sketch and formulae !

Dave Houston I will remain eternally grateful for the X-10 info pages and other links.

Adam those links you pointed me to where a great read and it seems I may have opened a can of worms with this thread ??

Dave, Mackrackit aka "CODE EYES" thanks for the forewarning and all the help thus far.

Melanie not only did you clear up my confusion regarding the calculation but have also given some fantastic tips to take into consideration which also triggered some more thoughts/questions.

So far I have the following calculated.

For 220V @ 50Hz to drive a MAX load of 30mA, I would need the following values (or have I messed up somewhere and is it just under 15mA?)

MOV(VDR1) = 275V
C1 = .47uF (suggested is X2 rated 275VAC based on other forums posts and internet sources)
R1 = 100 OHM (1/2 Watt)
D1 = IN4007

C2 = 470uF 25V (electrolytic)
D2 = 5V1 Zener

If all is correct, this output a half-wave rectified, smoothed 5V supply, right? If full wave were required I would need to a
If there are any further suggestions/changes I would appreciate them.

Is it true to say that the circuit ground 'hovers'(for lack of a better word) below HOT/LIVE and is not reference to EARTH GROUND and this is why all the warnings and danger signs ?

My questions are ...
(And while I ask I am busy reading/checking up on the following)

1.Whats happens when the circuit wants to draw more than max Ohms ?
2.How do I check the current draw of my PIC (assuming the PIC is standalone and not connected to anything other than power ?)
3. Once I start connectinng things like a push-button or triac or relay and perhaps an LED and maybe a POT for dimming, what would the ball-park usage start looking like and why ?
4.Which component is taking the most strain as the circuit draws current, is it C1 ?
5.Is C1 the current limiting factor of this circuit ?
6. I was thinking of using a BT139 Triac (since I have quite a few lying around), is it a good choice? I read that
7.What should the value for R3 be if I am to drive a triac ?
8. Is it feasible/possible to add a wireless TX module to the PIC circuit and if so what safegaurds (if any) like noise reduction should I put in place and where ?
9.In post #14 here amgen points out that we are looking at 0.15UF per 10MA and this seems to run true for both bride rectifier (full-wave) based as well as for the non-bridge based (half-wave) supplies. So the question is why choose to use a rectifier or not ? Is it cost ? Less of a smoke factor ? Space ?
10.To test the voltage on the output side could I just pop my good old Fluke meter on 5V and circuit GND with it set to read DC volts.
(and yes I will make sure I am insulated - rubber sole shoes, no jewellery, rubber gloves - hell I will wear a rubber suit if needs be!)
11. Are there any tips for connecting things up on the PIC side ?
12.And for taking measurements on the PIC side ? Apart from the app note SAFETY PORTION which refers to using an isolation tranformer.
Anyone have a schematic/block-diagram for measurements please?
13. Unless I am planning to use zero-crossing point to control the triac (can't think of a use right now, can you ? ) ,should I be concerned about zero-crossing issues as I have read somewhere that this circuit may cause slight delays in switching the load on or off based on Wattage and type either resistive or inductive) because the zero-crossing signal is phase shifted from the actual zero-crossing?

Oh and last but not least I can think of at least two applications that might be a no-no
1.Touch-sensitive pad(s)
2.ICSP
But what about things like USB or Connection to the PC serial port?
What sort of buffering or isolation would be required if is possible.

Apologies for the lengthy post and any questions which may seem obvious or silly to anyone reading this post (and the other posts too!)


Thanks once again

Kind regards

Dennis

[amgen]

Dennis,
Very through work !, and good questions.
Your diag had diodes in wrong places.
1) too much current draw will drop voltage to PIC to a point of no-operation,
thats why my ckt had 12V ZD and 5V reg, it provides smoother supply.
2)you have to insert DC ammeret (can be digital meter) in wire supplying PIC and other circuit.
3)Probably getting toward your 30MA, keep draws as small as possible, example---2MA led
4) C1 drops most voltage but most all parts are critical for operation
5)C1 is the current limiter, could use 600VAC rated part which is larger than say 250V part
-R1 developes heat by the amount of -- Current squared times Resistance
- also Z diode makes a little heat -----Z voltage times current through it (can measure or estimate
6) use an OPTO triac driver then any triac of the amps you need is OK
7) use 200 to 500 ohm to drive the triac driver depending on the opto chosen
8) consider added power used (more current) by added parts, within range of designed supply components
9) full-wave bridge makes use of pos and neg cycle of ACV, where 2 diodes only uses 1/2 cycle and the other 1/2 cycle is just wasted. bridge rectifiers come in dip package up to 1000 volts.
10,11,12,13,1,2))) If not using Z crossing, you could hook 10 or 20 volts from an ac or dc supply to test,adjust, ICSP then after all working hook to 110/220 power.
I think touch-sense sw's would be possible shock hazzard due to lack of higher voltage isolation, at some point Transformer isolation makes more sence.

I made some push-button 110V fan timers with this circuit, fits in wall switch box, never skipped a beat or burned one up, so I was addressing the same issues as you.

good luck,
don
amgen