Inspired by the top Sorensen line PSUs, I'm planning on building a variable output, industrial level PSU around a 5kW EI trafo. It's going to run on a dedicated 30A, 120V, 60Hz line. I suspect I can manage the cap inrush, there are many strategies and circuits for that but my concern is the trafo itself tripping the breaker as it energizes. What is the best overall solution to this problem (thermistors, AC relay circuits, etc)? Which is the best bang for the buck? What are the pros and cons of each? I'm not planning output greater than 3kW so a 30A line should be sufficient to keep it operational.
Maybe this would be of interest: Power Tips: How to limit inrush current in an AC/DC power supply - Power House - Blogs - TI E2E Community
Markw4
The page linked doesn't deal with the transformer energizing, it's focus is on the current draw from the trafo outputs and not unconventional supplies made around a trafo this large. In my case, the energy capacity of the trafo exceeds the line output ~3.6kW. It's tempting to just throw thermistors before the line primaries but I was hoping for something more sophisticated.
The page linked doesn't deal with the transformer energizing, it's focus is on the current draw from the trafo outputs and not unconventional supplies made around a trafo this large. In my case, the energy capacity of the trafo exceeds the line output ~3.6kW. It's tempting to just throw thermistors before the line primaries but I was hoping for something more sophisticated.
What can trip breakers at turn on may be referred to as transformer turn-on inrush current. It depends on the state of the transformer core that last time it was turned off and state (phase) of the AC line when it is turned back on.
In addition, initial charging of filter capacitors is part of what can cause large inrush currents at turn-on. Either way, the same fixes are normally used, which is usually to soft-start the transformer and filter somehow until the turn-on transient has time to mostly settle out into some kind of low level steady-state operation. At that point the rest of the input voltage can be applied, and another small bit of settling occurs. By this means the maximum inrush current can be limited to a suitable amount so as to not trip breakers.
You can use thermistors, resistors and relays, MOSFET, etc., your choice. For 100 kW 3-phase and larger supplies I have usually used resistors and relays.
The thing about simple circuits using thermistors is they heat up and their resistance drops starting from turn on, thus slowly applying full power to the transformer input. But, if you turn off the supply and turn it right back on, the thermistors may not have had time to cool off and reset fully, but the filter caps may still be charged too. If the problem is more with transformer core stored energy, then a non-thermistor solution might be preferable in that case. In other words, it helps a lot to clearly understand what the problem is before you start doing a lot of design work to fix it.
In addition, initial charging of filter capacitors is part of what can cause large inrush currents at turn-on. Either way, the same fixes are normally used, which is usually to soft-start the transformer and filter somehow until the turn-on transient has time to mostly settle out into some kind of low level steady-state operation. At that point the rest of the input voltage can be applied, and another small bit of settling occurs. By this means the maximum inrush current can be limited to a suitable amount so as to not trip breakers.
You can use thermistors, resistors and relays, MOSFET, etc., your choice. For 100 kW 3-phase and larger supplies I have usually used resistors and relays.
The thing about simple circuits using thermistors is they heat up and their resistance drops starting from turn on, thus slowly applying full power to the transformer input. But, if you turn off the supply and turn it right back on, the thermistors may not have had time to cool off and reset fully, but the filter caps may still be charged too. If the problem is more with transformer core stored energy, then a non-thermistor solution might be preferable in that case. In other words, it helps a lot to clearly understand what the problem is before you start doing a lot of design work to fix it.
Last edited:
I designed a triac based soft start circuit.
It ramps up phase angle slowly until full AC mains cycle is applied.
Mine only works up to 2KW so not much use to you but you could find a more powerful triac and bigger heats sink for your transformer.
It ramps up phase angle slowly until full AC mains cycle is applied.
Mine only works up to 2KW so not much use to you but you could find a more powerful triac and bigger heats sink for your transformer.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Here is the solution I used for my lab transformers (3.8kVA in total):
External/Autonomous Soft-Start
External/Autonomous Soft-Start
Power NTC bypassed with a time delayed relay are near enough perfect as a soft start for transformers.
Just use a close rated fuse and rely on the NTC to current limit during start up.
For a 120Vac supply and 30A fuse you need a total primary circuit resistance of 2r0. Yes that 2r0 will allow a transient of 60Aac to pass initially, but I and others have found that the close rated fuse does not rupture.
AndrewT
Sounds sensible, as usual
Will a 5W rated cement resistor hold up over time? What's your favorite relay for this kind of thing?
TonyTecson
In my opening post, I stated I needed variable output.
NigelWright7557
Very nice, I was considering a triac solution but I think I'll take AndrewT's advice. It's simple, cheapish and gets the job done.
Elvee
Very clever device but since this is going to be a design from scratch it's more sophistication than I need. I have bookmarked it. If oil goes to $300/barrel, the cost premium of using your circuit would pay for itself quickly. Thanks for sharing.
Markw4
Noted.
Thanks to all.
Sounds sensible, as usual
Will a 5W rated cement resistor hold up over time? What's your favorite relay for this kind of thing?TonyTecson
In my opening post, I stated I needed variable output.
NigelWright7557
Very nice, I was considering a triac solution but I think I'll take AndrewT's advice. It's simple, cheapish and gets the job done.
Elvee
Very clever device but since this is going to be a design from scratch it's more sophistication than I need. I have bookmarked it. If oil goes to $300/barrel, the cost premium of using your circuit would pay for itself quickly. Thanks for sharing.
Markw4
Noted.
Thanks to all.
you can use a fixed power resistor or a power NTC.
The advantage of an NTC is the variable resistance giving protection against a non closing bypass relay. And this leads to a further advantage the time delay on the bypass can be extended beyond the normal 100 to 300ms. 1 to 2 seconds would be OK for a 2r or 2.5r NTC.
5W fixed resistor is overloaded by too much for this duty.
I use 30w to 50W of low value (thick wire) wire wounds.
Instantaneous power from 120Vac (170Vpk) through a 2r resistor is 7200W !
Peak instantaneous is 14kW
The advantage of an NTC is the variable resistance giving protection against a non closing bypass relay. And this leads to a further advantage the time delay on the bypass can be extended beyond the normal 100 to 300ms. 1 to 2 seconds would be OK for a 2r or 2.5r NTC.
5W fixed resistor is overloaded by too much for this duty.
I use 30w to 50W of low value (thick wire) wire wounds.
Instantaneous power from 120Vac (170Vpk) through a 2r resistor is 7200W !
Peak instantaneous is 14kW
AndrewT
I have no intention of using a fixed resistor for this application unless it gets completely shorted out of the circuit after it does its job. I'm even a little annoyed at the losses to the thermistor but simple is better to start with. It's interesting that 30W resistors survive that kind of slamming. Given such short intervals, it's not implausible for even a 5W wire wound to survive a few cycles. I suppose it depends on the speed of the relay. A short list of the preferred heavy duty relays would be appreciated if you or anyone else can provide it.
TonyTecson
I should elaborate on the trafo isolation. The unit has two input options, 120/240AC so I should be able to use those two segments for the isolation. Isolation is a good idea for a universal supply. You never know what you're going to hook up to. Is there a circuit trick that takes advantage of isolation to curb the inrush?
I have no intention of using a fixed resistor for this application unless it gets completely shorted out of the circuit after it does its job. I'm even a little annoyed at the losses to the thermistor but simple is better to start with. It's interesting that 30W resistors survive that kind of slamming. Given such short intervals, it's not implausible for even a 5W wire wound to survive a few cycles. I suppose it depends on the speed of the relay. A short list of the preferred heavy duty relays would be appreciated if you or anyone else can provide it.
TonyTecson
I should elaborate on the trafo isolation. The unit has two input options, 120/240AC so I should be able to use those two segments for the isolation. Isolation is a good idea for a universal supply. You never know what you're going to hook up to. Is there a circuit trick that takes advantage of isolation to curb the inrush?
if only isolation, then inrush should not be a big concern, as you are just taking care of magnetizing currents if the traffo is unloaded..
different story when switching on that traffo under load..
or are you using this traffo as dc psu? or merely isolation for other amps....
Attachments
The soft start bypass does not need a heavy duty relay.
This relay does not switch off the amplifier and thus does not see the worst of the wear/deterioration that the ON/OFF switch/relay sees.
The ON only has the easy duty of starting up the resistor limited current.
jcx
Thanks for your attention. I was concerned about this too but I was assured it could be done safely. Are you saying the heat generated in a failure would breakdown the insulation? Isn't there a way to add some kind of fuse for failure protection? I think I have to get into the details of how the windings are situated. I don't have it in my hand at the moment but it's a GE house transformer with taps on the secondary and a removable bus bar connecting the two primary halves.
TonyTecson
I'm planning a universal bench supply so amps are only one of the things it will see. Are you saying the soft start only needs to be on the secondaries when the transformer is isolated? Why wouldn't the same issues apply when the trafo isn't ground isolated? Does the remanance reduce when you isolate? I know it's not a toroid but it's still a 5kW trafo! I've never worked with such a large supply before so I'm a little hesitant to jump in without being absolutely sure of what I'm doing.
Thanks for your attention. I was concerned about this too but I was assured it could be done safely. Are you saying the heat generated in a failure would breakdown the insulation? Isn't there a way to add some kind of fuse for failure protection? I think I have to get into the details of how the windings are situated. I don't have it in my hand at the moment but it's a GE house transformer with taps on the secondary and a removable bus bar connecting the two primary halves.
TonyTecson
I'm planning a universal bench supply so amps are only one of the things it will see. Are you saying the soft start only needs to be on the secondaries when the transformer is isolated? Why wouldn't the same issues apply when the trafo isn't ground isolated? Does the remanance reduce when you isolate? I know it's not a toroid but it's still a 5kW trafo! I've never worked with such a large supply before so I'm a little hesitant to jump in without being absolutely sure of what I'm doing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.