Further to the discussion at https://avtalk.co.za/threads/pc-board-design-software.1412/#post-10970
Anexus asked me to show what my effort looks like.
I saw this circuit in Elektor or ETI or one of those magazines years ago, in something like a "Design Ideas" column.
Perhaps informative if one starts closer to the beginning to describe how the thing functions. No education here, so no Math. Promise.
The basic idea is to make this
While the relay is off, Mains AC can only pass to the transformer via the 100 Ohm resistor. Once the relay switches on the transformer gets its mains feed unimpeded. So the huge current surge that the transformer (especially toroid types) wants at switch on is limited by the soft start circuit.
After switch on, depending on details, the relay takes a while (something like 10 mains cycles at 50Hz works OK) to switch on. During that time the 100 Ohm resistor passes something like 2.2 Amp to the transformer (220V/100 Ohm) and initially dissipates something like 484 Watt (220V x 2.2A) which may seem like a not so cool idea for a 20 Watt resistor. But as the transformer begins to "become alive magnetically" the current drops and the 100 Ohm resistor dissipates less.
One can abuse things, like resistors and Mosfets, slow blow fuses, transformers and even some caps etc quite terribly, for a short while. They will remain friendly.
Choosing a fuse for this is a bit tricky. With a 180-220 Ohm 20 Watt resistor limiting surge current to the transformer (225VA toroid) and a 1A slow blow fuse, I have had good success. After some 15 years of using this scheme I've never seen the circuit fail with up to 1.5kVA toroids (bigger fuse, smaller value 50W resistors etc then). If there are brown-outs or lightning strikes one usually has to replace the fuse.
I think of fuses as fire prevention, not equipment protection.
So what's the cheapest way to make a delay of some few mains cycles for the relay to come on?
Back to the beginning again:
An AC signal going into a capacitor can get through the capacitor, depending on the capacitance of said cap. Much like DC going into a resistor passes through depending on the resistance. But with AC we think in terms of "reactance" of the cap, and include freq of the AC. With DC there is no freq, just the steady applied voltage, so the calculation is easier.
If we have two resistors connected in a row, in series, with voltage applied across the ends then the voltage that appears at the junction is predictable given the voltage applied and the resistor values. Think volume pot.
If we have two caps in series and apply AC across the ends, we get an "AC volume pot" with the voltage appearing at the junction of the caps being determined by the AC voltage, AC freq, and the 2 capacitances.
So that's the approach when one makes this type of soft start circuit. You use a cap to pass AC from the wall to a relay. The cap helps to attenuate the AC. Being a reactive component, like a transformer (but cheaper and smaller) the cap doesn't dissipate power and doesn't get hot. Cool.
So the AC that passes through the cap, gets rectified by the 4 diodes, and stored on the 220uF cap. The relay coil is connected to that stored voltage. So a while after you switch the mains on, the relay is able to switch on as the charge on the 220uF cap builds up.
The 470nF cap functions as the upstream part of a volume pot, where the relay, diodes and 220uF cap functions as the "lower section" of a volume pot. You want to scale things to eventually get 12V (or 24V) across the relay coil when 220V Mains is applied. The ground in the schematic is not Mains Earth, it's just there to provide SPICE with a reference.
This is the populated board. I used to make the smaller ones. In the mean time I've run out of the smaller 470nF caps and could only find really bulky ones cheaply. So the layout here is for two big 470nF caps (it becomes useful for 110VAC too, if you fit the link.) There are extra holes on the PCB now, so whatever suitable caps you can get hold of can be used.
Assemble like so. Component side view.
A pdf of the layout is at http://www.ludo-diy.co.za/images/avtalk/Soft_Start_Big_X2_Caps.pdf
Print to A4 paper size with no scaling/fit to page or your parts won't fit. I put two of these side by side in the pdf, spaced about 1 jigsaw blade-width apart. I'm assuming you are a monoblock kind of guy. There's some text at the top of the page, a paranoid remnant from the bad old days in DOS when I used Tango and a utility called negjet to manipulate PCL and had the stangest issues sometimes with mirror image printing...
Here's a simulator pic. The green line is the voltage across the relay coil after switch-on (24V relay 1100 Ohm coil) and the blue line is the excess current dumped by the 24V Zener. With a 360 Ohm 12V coil on the relay the zener is not necessary, and the 220uF cap can be a 470uF cap to maintain a similar type of timing for the soft start. The relays usually switch on at about 75% of their rated voltage.
Important note. The whole thing as it stands here should be regarded as connected straight to the mains. Think of the entire board as being "live" The cap may limit current from the mains but the circuit is not "galvanically isolated" as it would be if we used a small transformer to power the relay. So sticking your fingers in there can be lethal. Don't forget this. The human physique, unlike power Mos, is not made for abuse. The 470nF cap has to be an X2 type, in other words rated for "across the mains" use. These are usually polyester or polyprop types with 275VAC rating.
Anexus asked me to show what my effort looks like.
I saw this circuit in Elektor or ETI or one of those magazines years ago, in something like a "Design Ideas" column.
Perhaps informative if one starts closer to the beginning to describe how the thing functions. No education here, so no Math. Promise.
The basic idea is to make this
While the relay is off, Mains AC can only pass to the transformer via the 100 Ohm resistor. Once the relay switches on the transformer gets its mains feed unimpeded. So the huge current surge that the transformer (especially toroid types) wants at switch on is limited by the soft start circuit.
After switch on, depending on details, the relay takes a while (something like 10 mains cycles at 50Hz works OK) to switch on. During that time the 100 Ohm resistor passes something like 2.2 Amp to the transformer (220V/100 Ohm) and initially dissipates something like 484 Watt (220V x 2.2A) which may seem like a not so cool idea for a 20 Watt resistor. But as the transformer begins to "become alive magnetically" the current drops and the 100 Ohm resistor dissipates less.
One can abuse things, like resistors and Mosfets, slow blow fuses, transformers and even some caps etc quite terribly, for a short while. They will remain friendly.
Choosing a fuse for this is a bit tricky. With a 180-220 Ohm 20 Watt resistor limiting surge current to the transformer (225VA toroid) and a 1A slow blow fuse, I have had good success. After some 15 years of using this scheme I've never seen the circuit fail with up to 1.5kVA toroids (bigger fuse, smaller value 50W resistors etc then). If there are brown-outs or lightning strikes one usually has to replace the fuse.
I think of fuses as fire prevention, not equipment protection.
So what's the cheapest way to make a delay of some few mains cycles for the relay to come on?
Back to the beginning again:
An AC signal going into a capacitor can get through the capacitor, depending on the capacitance of said cap. Much like DC going into a resistor passes through depending on the resistance. But with AC we think in terms of "reactance" of the cap, and include freq of the AC. With DC there is no freq, just the steady applied voltage, so the calculation is easier.
If we have two resistors connected in a row, in series, with voltage applied across the ends then the voltage that appears at the junction is predictable given the voltage applied and the resistor values. Think volume pot.
If we have two caps in series and apply AC across the ends, we get an "AC volume pot" with the voltage appearing at the junction of the caps being determined by the AC voltage, AC freq, and the 2 capacitances.
So that's the approach when one makes this type of soft start circuit. You use a cap to pass AC from the wall to a relay. The cap helps to attenuate the AC. Being a reactive component, like a transformer (but cheaper and smaller) the cap doesn't dissipate power and doesn't get hot. Cool.
So the AC that passes through the cap, gets rectified by the 4 diodes, and stored on the 220uF cap. The relay coil is connected to that stored voltage. So a while after you switch the mains on, the relay is able to switch on as the charge on the 220uF cap builds up.
The 470nF cap functions as the upstream part of a volume pot, where the relay, diodes and 220uF cap functions as the "lower section" of a volume pot. You want to scale things to eventually get 12V (or 24V) across the relay coil when 220V Mains is applied. The ground in the schematic is not Mains Earth, it's just there to provide SPICE with a reference.
This is the populated board. I used to make the smaller ones. In the mean time I've run out of the smaller 470nF caps and could only find really bulky ones cheaply. So the layout here is for two big 470nF caps (it becomes useful for 110VAC too, if you fit the link.) There are extra holes on the PCB now, so whatever suitable caps you can get hold of can be used.
Assemble like so. Component side view.
A pdf of the layout is at http://www.ludo-diy.co.za/images/avtalk/Soft_Start_Big_X2_Caps.pdf
Print to A4 paper size with no scaling/fit to page or your parts won't fit. I put two of these side by side in the pdf, spaced about 1 jigsaw blade-width apart. I'm assuming you are a monoblock kind of guy. There's some text at the top of the page, a paranoid remnant from the bad old days in DOS when I used Tango and a utility called negjet to manipulate PCL and had the stangest issues sometimes with mirror image printing...
Here's a simulator pic. The green line is the voltage across the relay coil after switch-on (24V relay 1100 Ohm coil) and the blue line is the excess current dumped by the 24V Zener. With a 360 Ohm 12V coil on the relay the zener is not necessary, and the 220uF cap can be a 470uF cap to maintain a similar type of timing for the soft start. The relays usually switch on at about 75% of their rated voltage.
Important note. The whole thing as it stands here should be regarded as connected straight to the mains. Think of the entire board as being "live" The cap may limit current from the mains but the circuit is not "galvanically isolated" as it would be if we used a small transformer to power the relay. So sticking your fingers in there can be lethal. Don't forget this. The human physique, unlike power Mos, is not made for abuse. The 470nF cap has to be an X2 type, in other words rated for "across the mains" use. These are usually polyester or polyprop types with 275VAC rating.
