Soft Start for a Power Supply

[bsikar]

I am in the process of designing a power supply. The exact specifications shouldn't matter, but I am currently trying to implement a soft start.

I was reading this article: https://sound-au.com/articles/soft-start.htm

I attached a relevant schematic from the site, but I this makes the circuit not floating anymore since they attach the output to the neutral of the input. Is there any way to solve this problem? My transformer does not have an extra tap so I cannot use that as a reference.

I was thinking that I could do a transformer less soft start that then feeds into another transformer, so the output is floating, but I am not sure if this is ideal.

If possible, please send a good soft start circuit or even a completed power supply with a dedicated soft start circuit could be valuable. I have seen some soft start circuits that use the LM317 part of a powersupply to do the soft start, but this still has the bridge rectifier and some filtering capacitors experience large in rush current.

[inse]

But you wouldn’t have to ground the secondary side.
Maybe this is because the circuit originates from audio equipment.
As primary and secondary are isolated via the relay, there shouldn’t be any problem.
But still you are softstarting the entire control circuit as well - if that creates a problem, you also could place the circuit in the secondary power rail before the filter capacitor.

[bsikar]

The primary is isolated via the relay, but the neutral will allow for the possibility of shorting to earth in some cases. At the breaker box I think neutral is tied to earth. If I implement their design and probe with my oscilloscope, I will have the possibility of shorting to earth which can cause damage right?

My understanding could be 100% wrong though, please let me know if I am mistaken.

[inse]

I don’t get it, you don’t have to ground the secondary of your power supply.
The lab supplies I own have an additional clamp that allows to do so if necessary.
I always prefer to have it floating

[bsikar]

Maybe my understanding about how transformers are wrong, but I was thinking that I would use the top of the secondary winding as let's say REF+ and the bottom as REF-. Then my load would flow only through those references on the secondary winding (never in contact with the primary windings). In their schematic they are using the neutral reference from the primary windings, how would I do this same circuit, but without using that neutral primary winding reference?

[inse]

Do! Not! Ground! It!

[bsikar]

I don't understand, what do you mean by dont ground it? I don't want to ground anything; this is why I was confused with their schematic. When they are using the neutral line I was thinking that it would allow for a path to ground since neutral is tied to ground(earth): https://diy.stackexchange.com/questions/157478/why-is-the-neutral-tied-to-ground-in-my-homes-electrical-panel

[inse]

Are we talking about the circled symbols?
Or did I get it totally wrong?
Green circle is advisable, red circle not.
But you would have to wire it up a little differently as you are combining both transformers.
Your power transformer would be hooked to the SA and N terminals and you probably would have to experiment with the delay capacitor as the softstart circuit itself is being softstarted.
As well you will have to determine the right resistance and power capability of the resistor bank.
I made some softstart circuits myself, one for the electric chainsaw which blew up three times until the right and final dimensioning had been found.
The other for a shop vac that started with the table saw - this one worked faultless from the beginning.

[bsikar]

I wasn't referring to the green or red circles, but rather SA and N.

I only have one transformer, so I am unable to connect SA and N to another transformer. I was wondering if its advisable to either:

1) make this without a transformer (since I only have 1 power transformer which i need for the psu)

2) make this with only 1 transformer which I can also use for my psu.

[inse]

In principle you could do it with something as simple as showing in the drawing.
The relay coil voltage would have to match the secondary voltage.
Problem is that the hold voltage of a relay is rather low, meaning if you do only a short power cycle, the relay may stay energized and not softstarting.
You can also add the circuit you found but might need to adjust the timing as the softstart circuit itself is being softstarted.
Using a PTC as soft start resistance is a good idea as in case the relay won’t energize in time, the resistor would protect itself.
You can do a lot of calculations on this, determining the energy the resistor has to accept.

[bsikar]

I see, I will try and simulate/build and test this. If you or anyone else has other thoughts let me know.

Thank you for the help 

[inse]

The manufacturers (e.g. TDK) have graphs /tables to find the right PTC for the capacity to be charged.
I changed my constantly exploding inrush current limiter to a PTC array, but this has to swallow a couple of kW for ms

[Benta]

Yeah, that kind of circuit works most of the time.
It doesn't work with an unstable mains supply or a child playing with the power switch (takatakataka...).

[bsikar]

What circuit are you referring to? And in your opinion, what circuit do you think would be best for my use case and why?

[Andy Chee]

I was reading this article: https://sound-au.com/articles/soft-start.htm

If possible, please send a good soft start circuit or even a completed power supply with a dedicated soft start circuit could be valuable. I have seen some soft start circuits that use the LM317 part of a powersupply to do the soft start, but this still has the bridge rectifier and some filtering capacitors experience large in rush current.

Elsewhere on the site, there is this circuit, powered from the existing secondary, no auxiliary transformer required:



https://sound-au.com/project39.htm

[bsikar]

I see, there is the warning of:

WARNING: In the event of an amplifier fault at power-on, the fuse may not blow immediately with this circuit installed, since there may be no power to operate the relay.  The current is limited to 200% of that at normal full power, so the fuse may be safe for long enough for it to destroy the resistor(s)! The ballast resistors will overheat very quickly, and if you are lucky they will fail.  If you don't like this idea - Use The Auxiliary Transformer.

Since I would have to do this without an auxiliary transformer I will have to test it.

I do have one question about it, in this configuration I see Vcc which would be the Hot line from the wall, would the ground symbol they put be the neutral line? A and SA are the inputs to the transformer right?

[Andy Chee]

I do have one question about it, in this configuration I see Vcc which would be the Hot line from the wall, would the ground symbol they put be the neutral line? A and SA are the inputs to the transformer right?

No!!!!

Vcc is your low voltage secondary @ 50 volts DC or whatever.  Ground symbol is also connected to the secondary ground system.

If the secondary is still unsuitable, then below is the direct mains powered version:

[inse]

This is only a delay circuit, the softstart resistor would be placed between A and SA.
The circuit is powered by the main transformer as in my sketch.
Ground and earth should be kept separate on a lab power supply.

[jonpaul]

simple:

thermistor or Amperite thermal  time delay relay

j
 

[bsikar]

I am having a hard time simulating this. Could someone please help me out by providing a circuit from ltspice (a .asc file).

I found on this forum: https://groups.io/g/LTspice/filessearch?p=name%2C%2Csoft+start%2C20%2C1%2C0%2C0&jump=1

the attached .zip

There is so much happening and their schematic is really confusing for me, but if someone is able to help, I think its a good starting place.

[bsikar]

I dont want to use thermistors because if a bypass relay fails and excessive current flows through the thermistor, it may explode. Additionally, thermistors can be problematic in applications without consistent high power draw, as continuous thermal cycling can reduce their lifespan. Thermistors also have a wide resistance tolerance, making their behavior less predictable than resistors.

[inse]

The take PTCs, they will protect themselves in case of relay failure.

[Benta]

Designed this a couple of years ago. It works, and will take care of kids and unstable mains supply:
It uses a phase-angle triac control, which makes it independent of the load.

[bsikar]

At first glance, that looks really good. Thank you so much for the circuit. I will spend some time testing and simulating it

[Benta]

Thinking back, I think it was a bit of a nightmare simulating it, but I'm no expert there.
I think I separated the circuit into two sections:
1: the left part up to the output of the optocoupler, replacing the relay coil with a resistor.
2: the right part, which is really simple.
Both sections simulated correctly (but independently), and after building the circuit it worked as expected.