Yet Another Soft Start Circuit

After looking at a bunch of soft start circuits I decided to design one according to my beliefs as to what is the best way to do it. It may be completely different from your beliefs but there it is.

I wanted a separate power supply for the soft start circuit so it could be used to turn the amp on and off and the powering up of the amps large transformer would not interact with the circuit. The transformer has a dual primary so it can be used for 115 and 230 VAC. There is a surge suppressor, R1, across the transformer input as well as a small high voltage capacitor to filter out RFI on the line from coming into the amp. The power supply always draws a few milliamps, it powers a multicolor LED that emits red with the amp in standby.

The on/off switch only switches 15 volts DC at very low current so you don’t need a switch that can handle large AC current in-rushes. When switched on, the DC power is sent through a regulator for consistent voltage and time of circuit operation. Relay K1, a small current relay, pulls in and turns on a triac, AC current flows through a couple thermistors in series to limit in-rush current, two 5 Ohm parts for 115 volts to limit the current to no more than 12 amps, or two 10 Ohm parts for 230 volts also limiting to 12 amps. There is a current limiting resistor, R2, at the triac gate and a RC network,R3/C2, over the triac for some protection of the triac.

The circuit uses a voltage comparator, U2, to delay turn on of a relay, K2, which bypasses the thermistors, I do not like to have a current limiter staying in the circuit or the heat they generate inside the amp. The voltage comparator has a voltage divider, R8/R9, to set the voltage at 10v as the reference voltage on the positive input. That voltage is compared to the negative input which has an adjustable resistor, R7, which limits the current charging the capacitor, C6, and the rate of charge therefore voltage ramps up quickly, a couple seconds to more than 10 seconds, for the delay. When the voltage on the negative input matches the positive input, the output is turned on. There is a diode, D4, across R7 to give a quick discharge path for C6 so the circuit can reactivate quickly, as in the case of a temporary power loss. R2 limits the current through the comparator which drives the gate of a small MOSFET through a gate resistor, R11. The MOSFET pulls current through a large relay K2 to bypass the thermistors and another small relay, K3, which changes the LED from red to blue to show status ON. There is a Zener diode, D5 and a small capacitor,C7, at the gate of the MOSFET to protect and stabilize it, probably not needed but cheap insurance.


So far, the circuit has been very reliable in a couple power amps with 2400VA and 4000VA toroid transformers with .78 farads to 1.2 farads of storage caps. In small part quantities the soft start board costs around $34 USD.
Attached is the Schematic, BOM and Gerbers. Enjoy.
 

Attachments

  • softstart.zip
    15.3 KB · Views: 489
  • softstart_Schem.pdf
    175.6 KB · Views: 1,655
  • SoftStartV3.zip
    14.7 KB · Views: 415
Last edited:
  • Like
Reactions: 1 user
I have used a lot of soft starts with bypassed thermisters- have you measured the current/ voltage across those thermisters over time as this starts up and they are removed from the circuit?

I only ask because I have not seen measurements of this before in a soft start application.
 
To Lgreen: the thermistors heat up with the current through them and they drop in resistance with the heat, when the relay pulls in and bypasses them the current through them drops to almost zero and they cool off. Since they are only in the circuit for 3 to 10 seconds they don't get that hot.


If the relay to bypass the thermistors were to fail to pull in, the thermistors will stay warm and remain at a lower resistance and since they are sized for much greater current nothing will be hurt.
 
Actually, if you take it by section it is very simple. The power supply is nothing special.
The voltage comparator as a timer/switch is one of the simplest way to power up. And a relay is a simple mechanical device. The circuit has been very reliable in every amp produced for the past six months since I designed this soft start. It is really the best of many designs combined into one.
 
Account Closed
Joined 2018
The amount of components are not "simple", sorry.
When a delayed turn-on can be accomplished with much less fuss.
Asides from all that, a properly designed amp, or any equipment, for that matter, doesn't need such things.
Tubes, for instance, have their own "delay" due to controlled filament heating.
Solid state has capacitor-related surge suppression.
If.... they are designed properly.
 
I am looking at removing the power supply from the new case that I made for the power amp. With the amplifier being 24" long and weighing 135 pounds with the upgrade of the toroid from 1kVA to 2.4kVA and the capacitors from 2 at 60K mf to 8 caps with a large inductor in between for a CLC and a total of 632K mf. The original Amp was 16" long by 18" by 90 pounds mono-blocks. Moving the PS and amps to separate case will make it so I can move them easily, if I ever have to after setting them in place. Will replace the Soft start with my new boards as well.
 
After looking at a bunch of soft start circuits I decided to design one according to my beliefs as to what is the best way to do it. It may be completely different from your beliefs but there it is.
Enjoy.
There is a soft start working scheme, you can collect it yourself or buy.
HTB1sblKX5zxK1Rjy1zkq6yHrVXaJ.jpg

AIYIMA 1000W Power Supply Delay Power Soft Start Protection Board High Power For Class A Amplifier DIY 30A Relay Protection 220V-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
 
There are a few things about that circuit I do not like. 1. It is overly complicated, but to each their own on that. 2. You have to turn on a switch AND push a button to start the circuit. I did away with that by using the voltage comparator and the controlled cap charge rate. 3. It has the thermistors in both parallel and series, and yes I have even used them that way once but with each parallel device able to carry the whole load. So why parallel them and according to the manufacturers of thermistors it is a no-no to parallel them. Using one higher current rated device instead of two smaller ones in parallel may be more expensive but is far safer, and since I put my design into the public domain I would not share something that could be hazardous.
 
I developed some time ago SoftStart with push button ON/OFF and delay of cca 2sec with current limiting resistors (4x180R/5W in parallel gives 45R/20W resistor but it better to use 10W resistors to have 40W power dissipation of current inrush limiter).

It have fuse to protect toroid from damage, and it is small and compact..i build three unit until today and all worked perfectly.

[NEW] SoftStart+Push Button V1.0
 
For an in-rush current limiter I prefer a thermistor over a resistor. The resistor may be slightly cheaper but if something happens and the bypass relay doesn't pull in then the thermistors with heat up and decrease in resistance allowing the current to flow where as a resistor or set of resistors will just get hot and may limit the current to the amp.
Your thoughts and/or preferences?
 
how about something simple?
You want simple ?
That relais comes in at around 175V~ (without the resistor).
As the NTC gets hot the resistance goes down and the output voltage up.
After a while, when the voltage gets high enough, the relais close the circuit bridging the NTC and holding itself.
With the resistor the threshold can be adjusted to a higher voltage and longer wait.
Mona
 

Attachments

  • KalmIn~Vrel.jpg
    KalmIn~Vrel.jpg
    37.4 KB · Views: 870
  • Like
Reactions: 1 user