I happen to need a soft start circuit. There are several designs out there but I just want something simple while I am testing out different combinations of load to add during the startup.
It finally dawned on me that I could just implement a "manual soft start" consisting of an SPST switch and a couple of NTC thermistors in series. The switch is used to short across the thermistors, removing them from the circuit. Unlike resistors, thermistors can be left in the circuit indefinitely without danger, although it is better to remove them after their soft-start role has ended.
Using thermistors, the end of the soft start can be several seconds after main power is first applied. A click of the mains switch, a lazy pause... count to 10, then throw the switch to remove the thermistors from the circuit, and voila! you have the simplest ever soft start circuit.
Of course this leaves open the door for problems: if you shut down the mains power to the amp and forget to re-engage the soft-start before re-starting the mains power there is no soft start action. I am sure that one could dream up an interlock scheme, but that is straying from the KISS principle.
This is really only for bench testing purposes. But, it's darn simple!
It finally dawned on me that I could just implement a "manual soft start" consisting of an SPST switch and a couple of NTC thermistors in series. The switch is used to short across the thermistors, removing them from the circuit. Unlike resistors, thermistors can be left in the circuit indefinitely without danger, although it is better to remove them after their soft-start role has ended.
Using thermistors, the end of the soft start can be several seconds after main power is first applied. A click of the mains switch, a lazy pause... count to 10, then throw the switch to remove the thermistors from the circuit, and voila! you have the simplest ever soft start circuit.
Of course this leaves open the door for problems: if you shut down the mains power to the amp and forget to re-engage the soft-start before re-starting the mains power there is no soft start action. I am sure that one could dream up an interlock scheme, but that is straying from the KISS principle.
This is really only for bench testing purposes. But, it's darn simple!
Member
Joined 2009
Paid Member
There's another problem to consider: Without the switch, the thermistors will be hot and not protect against inrush if power is removed and then reapplied without cool down time. This is why every circuit I've seen (other than cheap consumer equipment) uses a relay to shunt the thermistors.
I'd use a 555 timer and a relay (or triac, but that might be asking for trouble). Since you're already not isolated from the mains, a cheap cap dropper supply should be fine (and I shouldn't be violating the rules for suggesting it), otherwise the smallest low voltage transformer you have should be plenty good enough.
Another idea: Use a push button to trigger a relay connected to latch when the button is pushed. An extra set of contacts will do this. This will shunt the thermistors, but automatically drop out when power is removed. You have to be the timer and push the button, but it's simple.
Last edited:
I have used a relay with a 120VAC coil in parallel with the primary transformer winding to shunt the series resistance. When the voltage on the primary gets up to 85 - 90 volts the relay contacts close and you have full power. Starting up under a heavy load could be an issue with this however.
I need a delay of several seconds (e.g. 5sec or more) with the additional load in the circuit to cause a slow charge up of the main caps and other circuitry. Most ebay sourced soft starts don't connect the additional load for that long. The typical soft start that adds the load to the circuit for only the first few hundred milliseconds to prevent the initial large inrush won't work for me.
I would never leave a thermistor in the mains circuit continuously. I don't care what other "well known" designers do. It's bad practice IMO.
This just for bench testing purposes. The switch on the thermistors (separate from the mains power switch) has to be manually opened and closed when appropriate. This would not work for a permanent application.
I'm aware of how to design a proper soft start, e.g. with a 555 timer or a simple RC delay circuit powering the coil of a relay. I will eventually design/build something along these lines. The manual switch approach just gives me something to play around with that is easy to throw together and change out components as needed until I get to that stage.
I would never leave a thermistor in the mains circuit continuously. I don't care what other "well known" designers do. It's bad practice IMO.
This just for bench testing purposes. The switch on the thermistors (separate from the mains power switch) has to be manually opened and closed when appropriate. This would not work for a permanent application.
I'm aware of how to design a proper soft start, e.g. with a 555 timer or a simple RC delay circuit powering the coil of a relay. I will eventually design/build something along these lines. The manual switch approach just gives me something to play around with that is easy to throw together and change out components as needed until I get to that stage.
I like FoMoCo's suggestion to (a) make the SPST switch a normally-open, pushbutton type; and (b) use a two pole relay with mains operated coil, to short out the thermistor and to keep it shorted out even when CharlieLaub removes his finger from the pushbutton.
Operating sequence:
Operating sequence:
- Initial condition: amp is off and thermistor is cold
- Activate the On/Off switch to turn the amp on
- Wait at least 3 seconds, and then
- Push the pushbutton switch labeled "Press after warmup"
- Stop pushing the pushbutton
- Listen to music
- When ready to shut the amp off, switch the On/Off switch to Off position.
- WAIT AT LEAST 30 SECONDS BEFORE RETURNING TO STEP 1!!
Attachments
OK, I too have to admit that this is a nice concept. I happen to have some 120VAC coil 12A DPDT ice cube type relays. I haven't bought the switches yet, so maybe I will just go for this option because power on without the thermistors in circuit is not possible unless you happen to also hold down the "press after warmup" button. That is much less likely than forgetting to reset the manual switch in my two-manual-switch concept circuit.
It's also pretty simple, and I like that, too.
Using a single thermistors isn't the best solution for amplifiers.
They warm up on power up and the resistance goes low.
The trouble is if the amp is idling the thermistor can cool down again and become high resistance.
You really need a thermistor or resistor with a relay in parallel.
...or a switch, which is what I proposed initially.
But in either case, yeah, get it outta there after its job is done.
Simple is relative. If super simple is required, then a manual switch might be just the thing. Off hand, I'm thinking for that kind of delay, a 555 astable circuit set at a slow time contstant and a shift register used as cascading flip-flops. Make the circuit so that when the output is active then the timer stops.(The CMOS 555's can work with a higher impedance on the timing pins.) Use this off line circuit to drive either relays or a photo-triac/triac combination. This allows for isolation as well if needed.
BTW, Triac used at 360 degrees will have a small amout of dead time each half cycle and this may effect a large toroid due the core having no gap. You can use the triac to connect the power resistor in series with the primary, and use a relay to short out the triac/resistor combo after the slow start time has passed.
BTW, Triac used at 360 degrees will have a small amout of dead time each half cycle and this may effect a large toroid due the core having no gap. You can use the triac to connect the power resistor in series with the primary, and use a relay to short out the triac/resistor combo after the slow start time has passed.
I recently had the same problem, I needed a very simple ( read cheap ) reliable soft start circuit for an amp with a large power transformer.
The way I did this, was a resistor of 270 ohm 5W in series with the primary winding of the large power transformer ( note: we have 230V~ mains here ). This resistor is shortened by a relay which is powered from one of the main DC voltage rails in the amp.
The resulting delay was about 1 sec and the power-on surge significant reduced.
The way I did this, was a resistor of 270 ohm 5W in series with the primary winding of the large power transformer ( note: we have 230V~ mains here ). This resistor is shortened by a relay which is powered from one of the main DC voltage rails in the amp.
The resulting delay was about 1 sec and the power-on surge significant reduced.
a CL60 in series with a high power 110/120Vac primary.
If you have two high power primaries then each needs a CL60.
If you use a 220/240Vac primary then a high power transformer needs two CL60 in series. That gives the same current limiting as the two CL60 to the twin 110/120Vac primaries.
If you have a low power transformer then the CL60 has too low a resistance to be an effective current limiter.
Try a series pair of 20ohm thermistors for a medium power 220/240Vac transformer.
A 50ohm to 80ohm thermistor would suit a lower power 220/240Vac transformer for effective current limiting during start up.
Last edited:
I have a bunch of 10 ohm 10A NTC thermistors. The CL60 is 10R 5A. I plan to use four or five of these in series with a 120VAC 500VA transformer that has dual 115VAC primaries. I am looking for a rather slow start up over the first couple of seconds, removing the thermistors after 5-10 seconds.
I can't find a good way to calculate how the thermistors' resistance will drop over time (e.g. over the first ten seconds of so) in my application, although I can model the behavior with a pure resistance and know that the thermistor resistance will drop as it heats up. This is one of the reasons why I want to jury rig a "soft start" type of delay. In that way I can monitor the AC current with an analog panel meter and then decide how many thermistors to use and when to bypass them.
Where did you get them? I've been trying to find something other than CL60's as well with a higher current rating, but also 10-20ohm cold resistance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.