Already had this worked out ...
Since your load is only 60W, there's no reason or benefit to use a snubber with higher capacitance. At 240V/50Hz and 0.1µF capacitance, the steady-state current through the snubber will be about 7.5mA.
Here, the capacitive reactance plus snubber resistor is 31831Ω + 100Ω. (240V/31931Ω = 7.5mA AC)
Oh, the resistor should have a sufficient working voltage rating to handle 240VAC continuous ... usually its sized 1.6W or higher with at least 500V rating because it has to handle the voltage spikes and surge current.
Cool cheers for that. But .... 1.6w resistors bit high isn't it? Been looking at decent brand pricey snubbers 0.1 uF 100 ohm but still only seem to be 0.5w. e.g https://docs.rs-online.com/1bb6/0900766b813e543c.pdf
These snubbers seem to be manufactured specifically for mains 250v ac so thought they'd had a beefier resistor in them?
I don't really get how to work out the resistor wattage in this situation because it's only dissipating the capacitors charge over the discharge time, plus the ac cycle in the mix to confuse me more. Looking online such discussions normally end in a argumentitive thread! lol
Cool cheers for that. But .... 1.6w resistors bit high isn't it?
If you're designing your own ... no, not too high, probably too low. It's not the steady-state current that's at issue, just the capacity to handle surges/spikes. Note that 240V RMS is equivalent to 339 V peak, or 679 V peak to peak. (that's for a clean sine wave). Really, the RSF2WS type would be better. (2W, 600V). In normal working conditions, the capacitor will drop most of the voltage, until a spike comes along and the capacitor temporarily becomes (almost) a short circuit.
Been looking at decent brand pricey snubbers 0.1 uF 100 ohm but still only seem to be 0.5w
Yep, lots of epoxy, larger size and surface area significantly improves surge withstand voltage. Wattage isn't so much of an issue when thinking of the steady-state current - works out to 0.0056W.
Do the math.
V=IR=0.0076A 100 ohms = 0.76V.
P=IV= 0.0076A0.76V=0.005776W !
1/4 W=0.250W
1/2W=0.5W
which means 1/4W resistor is 0.250W/0.005776W (43) times LARGER than it needs to be !
This is ALL due to the small cap having a LARGE
REACTANCE value which should be obvious acts
to limit the current which limits the resistor power
dissipation.
All you have to know is Xc= 1/(2PifC) | where
Pi=3.14159265359 (yes, it's a phone number on the right of the decimal point.) It's an unassigned number for SLAC (Stanford Linear Accelerator Center) (1 (415) 926-5359) (put it in your contacts)
f=frequency (50Hz)
C= Capacitance in Farads (F)
Bottom line is by keeping C small you limit the resistor size. The larger the C, the higher the resistor wattage because the current increases with cap value.
If you had calculated the power dissipation for the
29 ohm resistor in my original calculation you would have realized I did something horribly wrong
because the current 240/29.5 ohms = 8.13A so
V= IR = 8.13 * 29 =235.77V
P=IV=8.13A235.77V = 1,916W !!!
The 29 ohm resistor would have to be 2000W !
That's called a 'sanity check' . If you get insane
values you screwed up big time.(like I did)
The takeaway is that when a calculation blows up
in your face like that it's hard to forget.
("The worst calculation I ever did was ...")
1/2W /100Ω / 600VDC / 250VAC as a discrete through-hole part is non existent.
When the relay contacts open, an arc is generated by the motor's winding and contact bounce ... the cap sees fast rising spike(s) which naturally pass through the capacitor as if it were a short circuit. The resistor limits the peak current which is much, much higher than 7.5mA (could be several amps peak).
I found RC snubber calculation on the web: NOTE: NOTICE THE SAME CALCULATION FOR THE CAP ? (C=** I2/10****)** ** ?**
** Notice that this website shows the units for the cap (uF) (NOT F !)**
** Now , if we go back to my previous calculations : **
C=** I2/10**
= (0.250 A)2/10 C** = 0.00625 F**
and CHANGE the units to uF, C=** I2/10**
** = (0.250 A)2/10**
** = 0.00625 F** ** =>**** 0.00625 uF** C= (0.250A)2/10 C= 0.00625 uF **
** C = 6.25 nF
** Now to calculate the series current, we first need to find the Reactance of the cap:*.**
__Xc= 1/(2PifC)__
Let Pi = 3.14159265359 Let f = 50 Hz XC = 1/(23.14159265359506.253064001360667e-9 F)
And last but not least , we have to find the wattage of the resistor:
P = IV
Let I = 0.000471 A
VR = I * R
= 0.000471 A * 29 ohms
VR = 0.013659 V
PR =I2 R
= (0.000471 A)2 * 29 ohms PR = 0.000006433 W (6.43 uW) (PASSES SANITY CHECK)
Somewhat different from my first attempt, eh ?
1/2W /100Ω / 600VDC / 250VAC as a discrete through-hole part is non existent.
When the relay contacts open, an arc is generated by the motor's winding and contact bounce ... the cap sees fast rising spike(s) which naturally pass through the capacitor as if it were a short circuit. The resistor limits the peak current which is much, much higher than 7.5mA (could be several amps peak).
I'm interpreting you Dlloyd and Raschemmel as giving conflicting info. I'm probably being stupid so forgive me. I mean Raschemmel reckons a 1/4 w resistor is 43 times larger than necessary and you reckon I need 1.6w from the maths and 2w to include safety factor.
Also I don't know why you say "1/2W /100Ω / 600VDC / 250VAC as a discrete through-hole part is non existent". Its right there in the list, part no. 104M06QC100, highlighted means available. Whilst on the subject of components, when the manufactures state these voltages like the 250vac 104M06QC100 one, is this 250vac rms? They don't seem to say. Obvs my voltage peak is 339vac from a 240vac rms.
This calculation is based on C = I^2 /10, C in uF. Yesterday your calculation was based on C = I^2 /10, C in F. Big difference. This is also for an across the switch snubber, rather than across load.
People including manufactures seem to advise across load snubbers to across switch. One reason being across switch "leaks" current, whilst across load doesn't because obviously it definitely doesn't see any power when switch is open. I don't get how though, I know the across switch rc snubber circuit is physically complete even when switch is open, but the capacitors dielectric isolates them. Do they mean a small current passes through the load as opposed to leaks through the dielectric?
Read Reply #24 (found the reason for the error in my first calculation)
This calculation is based on C = I^2 /10, C in uF. Yesterday your calculation was based on C = I^2 /10, C in F. Big difference. This is also for an across the switch snubber, rather than across load.
I see you caught that too !
FYI, yes, it does leak current , but with the correct units for capacitance, the current comes out to 471uA.
Using the calculations here I got 80 ohms for the resistor using rms voltage and 118 ohms using peak voltage.
Also I don't know why you say "1/2W /100Ω / 600VDC / 250VAC as a discrete through-hole part is non existent". Its right there in the list, part no. 104M06QC100, highlighted means available. Whilst on the subject of components, when the manufactures state these voltages like the 250vac 104M06QC100 one, is this 250vac rms? They don't seem to say. Obvs my voltage peak is 339vac from a 240vac rms.
That part number is for a complete snubber circuit embedded inside epoxy.
The epoxy helps achieve the 600V peak rating ... just like conformal coating on a PCB can more than double the voltage rating between traces on a PCB. This is factory made, so they were able to use or create 1/2 watt rated resistance.
If you are building your own snubber, you can't find a 600V (peak or DC) rated 1/2 watt resistor. They're usually just 250-350 volts peak and are not suitable for 250VAC working voltage.
Why build your own when there are so many off the shelf RC snubbers. I only did the calculation because I wanted
to learn how to do it, not because I would build my own , although there are times when you can't wait , even for
Digi-key and you need to throw something together on the spot, otherwise you lose a day or two..
dlloyd:
That part number is for a complete snubber circuit embedded inside epoxy.
The epoxy helps achieve the 600V peak rating ... just like conformal coating on a PCB can more than double the voltage rating between traces on a PCB. This is factory made, so they were able to use or create 1/2 watt rated resistance.
If you are building your own snubber, you can't find a 600V (peak or DC) rated 1/2 watt resistor. They're usually just 250-350 volts peak and are not suitable for 250VAC working voltage.
OK got it thanks. So to cut it short.. if I get one of these pre made epoxy snubbers, X2 rated for safety, minimum 250vac continuous rated (even if no RMS rating mentioned) and 100 ohm, 1/2 watt, I'll be good to go?
raschemmel:
Why build your own when there are so many off the shelf RC snubbers. I only did the calculation because I wanted
to learn how to do it, not because I would build my own , although there are times when you can't wait , even for
Digi-key and you need to throw something together on the spot, otherwise you lose a day or two..
I don't want to build my own, never did or said that. More about making sure I get the right rated one for the job and also interested in learning. So https://uk.rs-online.com/web/p/rc-network-capacitors/6167682/ this would be ok yeah? 250vac, doesn't say rms but 1000v pulse rated, x2 rated, 100ohm, across the load, bosch... job done?
Perfect.. I've actually already included one of those TDK MOVs in my design to use alongside snubber. I read best practice dictates using them with RC snubber as snubber reacts quicker for the opening switch voltage spike (or deals with it better somehow), but sounds like you think overkill for my small load which I suspected all along.
FYI, I think snubbers are calculated based on inrush current. Do you know what it would be for your aquarium pump ? (maybe 10x operating current or 2.5A ?)
raschemmel:
FYI, I think snubbers are calculated based on inrush current. Do you know what it would be for your aquarium pump ? (maybe 10x operating current or 2.5A ?)
BTW, is it salt water or fresh water aquarium?
I used to think it was inrush current on switch closing too. When I was researching snubbers it seems inrush current is more of a secondary concern, primary being voltage spike welding/wearing the contacts on switch opening due to voltage increasing from the remaining energy in the induced load trying to maintain the current. Theoretically the voltage can increase to infinity (and beyond)! So you need to snub the potential voltage spike by absorbing the current in the capacitor until the contacts have opened to a big enough gap that the current can't arc across and/or the induced loads magnetic field has fully collapsed and run out of energy. I might not be technically correct in that explanation but hopefully you get what I'm saying. I reckon 2.5 a inrush current is about right yeah, that's not a problem, got a decent Omron relay.
Pump is for hydroponics not for aquarium, but that's actually another hobby of mine.
