-
Inductive Load Kickback: need help fo find the right protection
Posted by goriath on 21 May, 2023 15:29 -
Hi everyone!
First post here after some time lurking (my bad).
Well, I hope to be enough IT here since it's more an electrotechnical matter than pure electronics, but electronics are involved at some point.
I'm working on a bench drill press which I'm recabling and making some improvements which I'm going to list below. In the meanwhile the motor specs are the following (I live in Europe):Quote230-240V~/50Hz - 550W
Class B - IP20
ȵ: 1450 minˉ¹
Variable speed through pulley: 440-2580RPM
It's a single-phase AC motor which is equipped with two capacitors:
100uF 250VAC 50/60Hz start cap
16uF 450VAC 50/60Hz run cap
I know that data is poor, but more in depth info are not available unfortunately.
The motor is powered ON and OFF through a electromagnetic switch from KEDU (KJD20-2), 4pins which cuts Line and Neutral at same time. On the chain, always in ON state as you plug the cable in the main, there is a device board which sense and display RPMs through a photo diode as soon as the motor is started. There is another switch under the pulleys cover which cuts down the feed if the lid is accidentally opened while motor is running. As last thing there are two laser diodes powered by two AA 1,5V batteries, on/off switch, which draw two lines used as reference guide like many other electrical tools are used to nowadays.
What I've already done is the wirings to lasers from RPM board since the PCB allows that through a low power LDO driver which output 5V (laser diodes accept 3v to 12V) and get rid of batteries; similar machine equipped with laser guide power on laser this way w/o batteries
What I would like to do now is add a switch upstream of the chain since I don't like that the RPM board is always ON while not using the drill and it's constantly humming; so I can leave the plug in while not using the machine
During this long journey started months ago I learned a lot and at some point of this very entertaining project I started to worry about the inductive kickback from the motor while shutting down the machine, so I'm wondering now about the following things:- Life contacts: prevents contacts of the switches from arcing
- Devices safety: since there are other devices on the chain like the RPM board and the little AC to 12VDC transformer, which I'm planning to add for led light working area and driven ON and OFF by a little switch
- Other things I may not have considered yet and someone is going to point me out eventually...
In the beginning I thought about an RCD snubber, but I found very hard to design one w/o knowing more about the motor and inductance values; someone suggested me to use a device called Quencharc made by Cornell Dubilier which is an all-in-one package solution, but sold with different ohm values and I don't know what to pick. Another thing to consider in general about a snubber (selfmade or not) could be that more likely I not have enough room to place one.
Later I came across the idea of using a TVS diode. To make a long story short (lol), according to the rule Vso>=Vrms x 1,56 I need a TVS with standoff voltage of 360V or above. What I can find are TVS with 400Vso but rated for only 1.5KW, while concerning is that I would easily need more! 6KW? Even more? I don't know.
Found some options from littelfuse but excluding AK series which are very expensive, there are a bunch of alternatives, 5KW but Vso 350V or even 30KW with 360Vso, but I can't find them within EU.
In conclusion, I'm a bit lost now
Sorry for long posting, I hope this didn't come out overhelming but I wanted to give more informations as possible
Every hints, suggestions, new idea, experience or whatever from anyone who wants to help me is really welcomed!
Thank you in advance
-
RC snubber 100nF ~100R good enough for beginning experiments
-
RC snubber 100nF ~100R good enough for beginning experiments
Do you mean that using simply this could stop all my headches? -> 104M06QC100 -
Totally unnecessary. Induction motors have no "inductive kickback" and are generally extremely well-behaved.
Use your effort elsewhere. Eg, how about power fail prtection? Normally you'd have a self-holding relay that will prevent the motor from spinning up when power is reapplied.
-
Totally unnecessary. Induction motors have no "inductive kickback" and are generally extremely well-behaved.
Use your effort elsewhere. Eg, how about power fail prtection? Normally you'd have a self-holding relay that will prevent the motor from spinning up when power is reapplied.
Since when does an inductive component have no kickback?
OP can easily look at the contactor contact arcing to determine if he wants to "protection". A RC snubber does well enough. -
Totally unnecessary. Induction motors have no "inductive kickback" and are generally extremely well-behaved.
Use your effort elsewhere. Eg, how about power fail prtection? Normally you'd have a self-holding relay that will prevent the motor from spinning up when power is reapplied.
I could intend your observation as I'm more likely over thinking; worrying too much probably. Maybe this is what you are saying me and so you could be right, but from my understanding every *device* which involves the rising of an EMF, as soon as the current flow is abruptly interrupted, voltage across the load increase indefinitely in a bunch of milliseconds and this is what I refer to "inductive kickback".
Even the coil of transformers or relay could have a minimal kickback.
In my case, the motor is started and stopped by an EM switch, so if a power fail suddenly were occurr the switch should turn NO -
Totally unnecessary. Induction motors have no "inductive kickback" and are generally extremely well-behaved.
Use your effort elsewhere. Eg, how about power fail prtection? Normally you'd have a self-holding relay that will prevent the motor from spinning up when power is reapplied.
Since when does an inductive component have no kickback?
OP can easily look at the contactor contact arcing to determine if he wants to "protection". A RC snubber does well enough.
Do you think that the RC snubber posted above could do the job?
And a further more question: does it exist any factual concernings about the risk that any electronic device on the chain could be affected someway by the inductive kickback in general? -
Since when does an inductive component have no kickback?
An induction motor is not an "inductive component", it's a bit more involved.
It's normally modeled as a transformer with variable parameters (changing with speed/load/slip), and just as a normal mains transformer does not "kick back" when turned off, neither does an induction motor. Remaining field is absorbed in the (shorted) rotor.
You might have a point with VFC drive (high frequency snubbing), but that would be totally irrelevant to a single-phase motor with start/run caps and centrifugal switch. You can't do speed control there in any sensible way.
-
Since when does an inductive component have no kickback?
An induction motor is not an "inductive component", it's a bit more involved.
It's normally modeled as a transformer with variable parameters (changing with speed/load/slip), and just as a normal mains transformer does not "kick back" when turned off, neither does an induction motor. Remaining field is absorbed in the (shorted) rotor.
You might have a point with VFC drive (high frequency snubbing), but that would be totally irrelevant to a single-phase motor with start/run caps and centrifugal switch. You can't do speed control there in any sensible way.
That is counter to my knowledge and experience in working with electric motors. Energy storage in the inductances and capacitor, I see the contactor arc and go with that instead of a theoretical approach. You do get a small bit of generator action from the rotor's remanance and spin down.
OP the Kedu is designed for small motors and is a latching contactor/switch. You're not going to activate it 10,000 times so I would not fuss over any contacts arcing; compared to a small relay, it's made to tolerate this.
I would put the worry into safety - the emergency stop easy to reach, good face shield/PPE and some good lighting on the work.
When I am done using a drill press, I unplug it and drape the power cord over the top so everyone knows "it's off".
If you find troubles with the drill electronics then I would go further and consider a MOV or RC snubber. -
Many cheap drill-press / mill / lathe RPM boards have non-isolated PSUs and all wiring associated with them is mains 'live'. If this is the case for your board, tapping its 5V supply to run the laser diodes may be unsafe, unless all wiring and insulating parts are rated for mains voltage, and tools are required to remove any covers or housings.
Assuming it is safe, why not simply put a low current mains on/off switch in the mains feed to the RPM board after it branches off from the motor feed? -
Many cheap drill-press / mill / lathe RPM boards have non-isolated PSUs and all wiring associated with them is mains 'live'. If this is the case for your board, tapping its 5V supply to run the laser diodes may be unsafe, unless all wiring and insulating parts are rated for mains voltage, and tools are required to remove any covers or housings.
I see what do you mean. No, no tapping, JST connectors are on board providing the needed connections already:
(CON1 is 5V supply and left empty, but almost ready to being used! Only CON2 is used originally connecting the photo diode probe in order to sense RPM; plastic receptacles temporarily removed of course)
I crimped already and wired to the lasers; I made all JST connections because is a tiny system which I had available, so I can turn lasers on and off through a little switch and preserved original connections so you can still use batteries for any reason...
If you watch at the second pic you will notice an empty space on the PCB on the path from LDO to 5V out, besides the 47R. It's look like is reserved for an smd diode. I should populate the PCB with the missing diode?Which one and what would it be for?. Since it does seem in parallel to 5V out it should be to stabilize the output, isn't it? Is it recommended find the right one and solder it in place?QuoteAssuming it is safe, why not simply put a low current mains on/off switch in the mains feed to the RPM board after it branches off from the motor feed?
Not sure to understand but I'm going to use this switch upstream the chain so it will switch on/off the RPM board but it will be also mandatory to turn on all the rest: KS-1700
EDIT: I can't understand how to attach inline pics
-
In the beginning I thought about an RCD snubber, but I found very hard to design one w/o knowing more about the motor and inductance values; someone suggested me to use a device called Quencharc made by Cornell Dubilier which is an all-in-one package solution, but sold with different ohm values and I don't know what to pick. Another thing to consider in general about a snubber (selfmade or not) could be that more likely I not have enough room to place one.
From what I'm gathering here is you need to have a main switch that turns off the entire system when not in use. A simple magnetic switch at the mains line is really all you need. If power is lost during operation then that would be the safest approach. Magnetic switch will open on power loss from Mains supply.
For your Induction motor you can use a Arc Quencher either across the contact points of your motor switch or across the motor. It doesn't really matter. Needed them for 600V 50HP motor for irrigation systems but these were more complex and heavy duty. For your case the rule of thumb would be to have a resistance value equal your the supply voltage and a capacitance equal to current in mfd (uF) micro Farads. You have a 550W motor at 240V that gives you around 2 to 3 amps . So the resistor would be around 240 ohms with a 2 to 3 uF capacitance. You don't need to over think it. If the values are a little higher that would be fine.
You can use this as a reference to sizing your Quencher. https://hornerautomation.com/wp-content/uploads/2014/12/MAN0962-01_SparkQuenchers_Datasheet_01.pdf -
In the beginning I thought about an RCD snubber, but I found very hard to design one w/o knowing more about the motor and inductance values; someone suggested me to use a device called Quencharc made by Cornell Dubilier which is an all-in-one package solution, but sold with different ohm values and I don't know what to pick. Another thing to consider in general about a snubber (selfmade or not) could be that more likely I not have enough room to place one.
From what I'm gathering here is you need to have a main switch that turns off the entire system when not in use. A simple magnetic switch at the mains line is really all you need. If power is lost during operation then that would be the safest approach. Magnetic switch will open on power loss from Mains supply.
For your Induction motor you can use a Arc Quencher either across the contact points of your motor switch or across the motor. It doesn't really matter. Needed them for 600V 50HP motor for irrigation systems but these were more complex and heavy duty. For your case the rule of thumb would be to have a resistance value equal your the supply voltage and a capacitance equal to current in mfd (uF) micro Farads. You have a 550W motor at 240V that gives you around 2 to 3 amps . So the resistor would be around 240 ohms with a 2 to 3 uF capacitance. You don't need to over think it. If the values are a little higher that would be fine.
You can use this as a reference to sizing your Quencher. https://hornerautomation.com/wp-content/uploads/2014/12/MAN0962-01_SparkQuenchers_Datasheet_01.pdf
Let's see if I got it. You suggest me an EM switch at main, but in the way things are working now the EM switch from KEDU, from my understanding, is already doing this job in the case of a power loss; cable plugged in, push the KEDU, motor starts, power loss occurs, motor stops and KEDU turns NO again, main power come back but motor stays OFF. And this is still true if I simply put the lever switch posted above at main cabling inside the machine, whatever if the latter is left open or closed, so I don't understand the need of another EM switch, but I may miss something now...
About the snubber, I would need an RC network made of 2 or 3 uF (microF) cap + 240ohm resistor (how many watts?) but neither Cornell Dubilier nor Horner from your datasheet have that, so you are suggesting for a self made?
BTW thank you all guys, I making more progress in the last two days than in months asking here and there!
Seems this is not a so popular and intersting topic which didn't collect the feedback I need
-
Your RPM board has a transformer so should be safely isolated, and OK to tap 5V from. However after running for half an hour with the lasers on, please switch off and feel how hot its transformer is getting as it would be a PITA if the additional load from the lasers burnt it out!
-
Your RPM board has a transformer so should be safely isolated, and OK to tap 5V from. However after running for half an hour with the lasers on, please switch off and feel how hot its transformer is getting as it would be a PITA if the additional load from the lasers burnt it out!
I already made a test outside the machine, but I'll try again to be sure. The LDO has a constant output of 100mA, while the lasers are <= 50mA together so I feel pretty confident
What I am wondering is if I should place that diode there to protect the output from DC overvoltage (I suppose would be that the reason) -
Totally unnecessary. Induction motors have no "inductive kickback" and are generally extremely well-behaved.
leakage inductance ?
switch contacts arcing -
I'm searching for 2,2/3,3uF cap 400VAC rated to mate with 240ohm resistor (maybe 2 watts) for RC snubbing network. WIMA make these in PET. Would be a right choice?
The problem is that they are bulky and I not have so much space available
-
min 800Vdc or X class capacitor (self healing properties allow them be smaller) X class capacitors can be paper and therefor not last but cheap X class capacitors dont last as well
sometimes general 400Vdc capacitors break at 230Vac line -
You might also consider getting a 3 phase equivalent of your KEDU emag start switch (it will have have 6 terminals, 3pst) - use the two phases as your are now and use the third phase to switch just a hot to your little Aux supply- this keep the two circuits somewhat isolated.
-
min 800Vdc or X class capacitor (self healing properties allow them be smaller) X class capacitors can be paper and therefor not last but cheap X class capacitors dont last as well
sometimes general 400Vdc capacitors break at 230Vac line
Those from WIMA I found are 400VAC or 630/1000VDC
I'll see for a x class, they are general 275VAC -
I disagree with the need for such huge capacitors for an RC snubber: 2.2-3.3uF is monster.
You also need a 1MEG 1/2W safety discharge resistor across each one, as they can hold a charge and give a shock at the power cord end or when you are working on this. Be careful.
I use smaller snubbers like 100Ω 1W and 0.1uF 275VAC X2 to reduce contact arcing, but again this motor big switch is built for it anyway.
Does the tach have a mains fuse? -
I disagree with the need for such huge capacitors for an RC snubber: 2.2-3.3uF is monster.
You also need a 1MEG 1/2W safety discharge resistor across each one, as they can hold a charge and give a shock at the power cord end or when you are working on this. Be careful.
I use smaller snubbers like 100Ω 1W and 0.1uF 275VAC X2 to reduce contact arcing, but again this motor big switch is built for it anyway.
Does the tach have a mains fuse?
Indeed is very bulky and I really don't have enough room to store it, so I was about to give up on this and come back to TVS diode
I'm thinking that 1.5KE400 bidirectional from littelfuse can do the job afterall:
Vso=342V
Vbr window=380 to 420V
Max Ipp=3.7A
Max Vc 548V@Ipp
Pppm is 2kW for bidirectional
On the other hand if I can pick more conservative values for RC network components I could implement that more easily:
275VAC X2 100nF cap + 100ohm resistor 1 or 2W in series and another 1Mohm (1/2W) across cap leads
I was wondering: does it exist any chance that I can combine the above RC network and TVS diode to improve transient suppression?
EDIT: about the tach, if you mean the RPM board there is no fuse -
Part of engineering is weighing the complexity, cost and safety aspects. Including asking "do I even need this?"
Without going in and doing some science, measuring the inductive kickback, you can use your eyes to observe the switch arc.
But speculation never ends unless you have the measurements and/or experience. How much energy is involved is a guess.
There's a zillion ways to lower a transient's voltage to something that will not do damage.
The TVS can work but if the spike energy is too much for it, it will short and burn, stink up the lab.
A large MOV can take more energy, i.e. using a 300VRMS-rated part would keep it simple.
A RC snubber will also absorb and softly dissipate the transient energy, but it's 4 parts.
The drill press isn't always energized, not switched on and off a lot.
I dislike cheap chinese gear that is mains-powered with no fuse in it. Illegal for sale in any country but I like the import price and melted traces and shrapnel are the fuse lol. So for the dollars spent, I would add a fuse for it. -
I disagree with the need for such huge capacitors for an RC snubber: 2.2-3.3uF is monster.
You also need a 1MEG 1/2W safety discharge resistor across each one, as they can hold a charge and give a shock at the power cord end or when you are working on this. Be careful.
I use smaller snubbers like 100Ω 1W and 0.1uF 275VAC X2 to reduce contact arcing, but again this motor big switch is built for it anyway.
Does the tach have a mains fuse?
Sorry it's a typo I made when sizing a quencher for the motor. This is most likely the cause of the error. Your absolutely right that a 3uF capacitor is excessively large and my post should have read .2 to .3uf. And giving the motor size and voltage a 2W resistor would be adequate.
Seems a little late to correct my previous post.
-
Ahh it's ok, that makes more sense.
The (snubber) mathematical theory is very difficult and ultimately a rabbit hole. For formal product development I use a scope and HV probe to watch how a (relay/switch/triac etc.) RC snubber is doing, because they occupy a lot of space and cost.
It's a tradeoff between the perfect values and the smallest values that still are good enough.