One can use NEGATIVE temperature coefficient (NTC)
thermistors to help "soft start" a power supply. At initial
turn-on their resistance is high and then drops as they
heat up. Thus they limit corrent more at the very start,
and have lower resistance when warmed up.
The trick is picking one with the correct characteristics for your
application, i.e., enough current limit at turn-on, minimal
resistance when warmed up and a reasonable time
constant.
But it's the simplest "soft start" you can do in terms of component
count.
thermistors to help "soft start" a power supply. At initial
turn-on their resistance is high and then drops as they
heat up. Thus they limit corrent more at the very start,
and have lower resistance when warmed up.
The trick is picking one with the correct characteristics for your
application, i.e., enough current limit at turn-on, minimal
resistance when warmed up and a reasonable time
constant.
But it's the simplest "soft start" you can do in terms of component
count.
Take a look at http://www.rti-corp.com/Electronics/surggard.htm
They have a small selection guide on that page to.
They have a small selection guide on that page to.
I have built soft-starts for my home theatre receivers using the circuit below. This is a modified version of one of the soft start circuits found on Rod Elliot's ESP pages. I modified it to use the existing 12V standby power supply and logic-level trigger signal in the receiver. I use a single ordinary 25 ohm, 10 W ceramic coffin resistor, which doesn't even get warm in use. I used R1=47K in my circuits, giving approx 0.8 second start time.
This is a more complicated circuit than many others, but it has important features great for toroids... it will dropout very quickly on a power blip, and the soft-start time is just as long following that as any other instance. An NTC thermistor, or the simple soft-starts that use a slowly charging capacitor on the secondary to power a relay, will not have either of those properties. That means that if your power "hiccups", your toroid will thud and draw a surge current just as if the soft-start wasn't there. The active circuit will however soft-start the toroid in this case, just as always.
This is a more complicated circuit than many others, but it has important features great for toroids... it will dropout very quickly on a power blip, and the soft-start time is just as long following that as any other instance. An NTC thermistor, or the simple soft-starts that use a slowly charging capacitor on the secondary to power a relay, will not have either of those properties. That means that if your power "hiccups", your toroid will thud and draw a surge current just as if the soft-start wasn't there. The active circuit will however soft-start the toroid in this case, just as always.
Attachments
Like macboy I built a softstart using ESP project 39, figure 2 to power my krell clone that has two large 700 VA xformers. Like you are thinking I used a thermister and a delay board.
People will tell you that you could just put a CL-60 thermister in line with the primary of your transformer and it would work great for years to come. They will say that Pass Labs does it this way. So it will work but note that it will not softstart if the amp has been on for a long time (thermister is hot) and then turned off and on again rapidly. A rare case to be sure but it does happen.
On the other hand a delay circ will always work, switching out either a resistor or thermister after a certain delay will keep that element cool and ready for another softstart.
thread 1
Here are some links to the way I did it-
My soft start circ.
My startup wiring diagram
People will tell you that you could just put a CL-60 thermister in line with the primary of your transformer and it would work great for years to come. They will say that Pass Labs does it this way. So it will work but note that it will not softstart if the amp has been on for a long time (thermister is hot) and then turned off and on again rapidly. A rare case to be sure but it does happen.
On the other hand a delay circ will always work, switching out either a resistor or thermister after a certain delay will keep that element cool and ready for another softstart.
thread 1
Here are some links to the way I did it-
My soft start circ.
My startup wiring diagram
Thermistor soft starts are fine for constant loads like class-a amps, but the highly dynamic current draw nature of class-b amps precludes their effectiveness. They will limit the inrush but also spoil your dynamic headroom and burst power availability, and also increase distortion.
Thanks for the replies.
I'm building some active speakers with a variation of MaxHawk's soft switch. This will avoid the inrush upon "hiccup" because the amps will remain off. I think I'll take the simple route and put in an NTC thermistor. By the time I get up to turn that puppy back on, the thermistor should have cooled and be ready to work again.
Cheers.
I'm building some active speakers with a variation of MaxHawk's soft switch. This will avoid the inrush upon "hiccup" because the amps will remain off. I think I'll take the simple route and put in an NTC thermistor. By the time I get up to turn that puppy back on, the thermistor should have cooled and be ready to work again.
Cheers.
The thermister worked great on my Aksa 100s. I had twin 400va toroids and went with a 10 amp, 1.8ohm piece, but I would have to double check to be sure. I bought several different types and figured out wich one took about 1-2 seconds for the voltage rails to come up. It's been over a year and I haven't had any problems. The thermister doesn't get hot either. Negligable voltage drop across it during normal operation also.
-Mike
-Mike
Hey, this may be an insane idea, but I thought I'd throw it out there anyway.
Comments? The goal is to have "hard wires" once the soft start procedure is complete (enough). And to have a circuit simple enough for me to understand!
Immediate flaw -- there will be a very small period of time (during the switchover) where there is no current supplied to the transformer. Hopefully capacitors down the line would absorb that. Or maybe use a "make before break" relay.
Obviously values for the components are somewhat critical.
BTW, I got the idea from my motorcycle's junction box... the headlight-on-only-after-start circuit.
Wes
Comments? The goal is to have "hard wires" once the soft start procedure is complete (enough). And to have a circuit simple enough for me to understand!
Immediate flaw -- there will be a very small period of time (during the switchover) where there is no current supplied to the transformer. Hopefully capacitors down the line would absorb that. Or maybe use a "make before break" relay.
Obviously values for the components are somewhat critical.
BTW, I got the idea from my motorcycle's junction box... the headlight-on-only-after-start circuit.
Wes
Attachments
Hi Wes,
Normally in a soft start circuit, the contacts only short out the thermistor or resistor. The AC is applied to the resistive element all the time with the contacts across the resistive element. No interrupton of current then. We would also use a fixed resistor, not a thermistor. That's entirely up to you.
-Chris
Normally in a soft start circuit, the contacts only short out the thermistor or resistor. The AC is applied to the resistive element all the time with the contacts across the resistive element. No interrupton of current then. We would also use a fixed resistor, not a thermistor. That's entirely up to you.
-Chris
Off-Topic
Hey Wes,
Nice to see you lurking around here. I'm "CapJohn" from the Ninja250.net forums.
I'm doing the layout for my power amp now. When that's done, I'll work on my soft-start board. Like Richieboy, I find the thermistor method too crude. My goal for soft-start is not only as soft-start but to reduce AC wiring (i.e. no AC wires to multiple devices and from back to front of case). Schematics done, no layout yet.
Hey Wes,
Nice to see you lurking around here. I'm "CapJohn" from the Ninja250.net forums.
I'm doing the layout for my power amp now. When that's done, I'll work on my soft-start board. Like Richieboy, I find the thermistor method too crude. My goal for soft-start is not only as soft-start but to reduce AC wiring (i.e. no AC wires to multiple devices and from back to front of case). Schematics done, no layout yet.
Chris --
Oh ho, of course! Short out the thermistor with the relay! See attached revision.
The NTC thermistor is critical here, as it allows a smooth voltage rise in the transformer; then when the voltage is high enough, the relay kicks in. Once the relay kicks in, the thermistor is out of the picture; the relay stays on by virtue of it being on. Very appealing in a self-excited sort of way.
I think this circuit would work if tuned properly; one concern I have is if the transformer's output voltage isn't linear during core saturation, the circuit will turn on prematurely.
CapJohn --
Hey, good to see a familiar face! Now you know why I haven't been posting much... to busy building my GainClone! I thought the amp part was going to be tricky; turns out the hard part is the freakin' chassis!! If I tap another freakin' crooked hole, or break another $4 drill bit, I think I'm gonna scream.
My (untested) dislike to the thermistor startup is about the same as Richie's -- I'm concerned that there will be a delay turning on full voltage when the amp has been idle, and that it will add extra noise or choo-choo-ing in the system.
My *like* for the thermistor startup is the symplicity, and zen-like nature: it's on when it needs to be on.
These are the two issues (aside from cost and part count) I tried to address in my circuit.
Wes
Oh ho, of course! Short out the thermistor with the relay! See attached revision.
The NTC thermistor is critical here, as it allows a smooth voltage rise in the transformer; then when the voltage is high enough, the relay kicks in. Once the relay kicks in, the thermistor is out of the picture; the relay stays on by virtue of it being on. Very appealing in a self-excited sort of way.
I think this circuit would work if tuned properly; one concern I have is if the transformer's output voltage isn't linear during core saturation, the circuit will turn on prematurely.
CapJohn --
Hey, good to see a familiar face! Now you know why I haven't been posting much... to busy building my GainClone! I thought the amp part was going to be tricky; turns out the hard part is the freakin' chassis!! If I tap another freakin' crooked hole, or break another $4 drill bit, I think I'm gonna scream.
My (untested) dislike to the thermistor startup is about the same as Richie's -- I'm concerned that there will be a delay turning on full voltage when the amp has been idle, and that it will add extra noise or choo-choo-ing in the system.
My *like* for the thermistor startup is the symplicity, and zen-like nature: it's on when it needs to be on.
These are the two issues (aside from cost and part count) I tried to address in my circuit.
Wes
Attachments
Now to simplifiy the circuit more and protect from other problems ....
The relay runs off the secondary DC supply. Resistor as required to limit the coil current. The best part is that if you have a short, the secondary voltage may drop enough to drop the relay out. The resistor then acts as a fuse and opens. Of course the relay can be opened with a protection circuit as well.
-Chris
The relay runs off the secondary DC supply. Resistor as required to limit the coil current. The best part is that if you have a short, the secondary voltage may drop enough to drop the relay out. The resistor then acts as a fuse and opens. Of course the relay can be opened with a protection circuit as well.
-Chris
That schematick loks strange n wrong polarityChris --
Oh ho, of course! Short out the thermistor with the relay! See attached revision.
The NTC thermistor is critical here, as it allows a smooth voltage rise in the transformer; then when the voltage is high enough, the relay kicks in. Once the relay kicks in, the thermistor is out of the picture; the relay stays on by virtue of it being on. Very appealing in a self-excited sort of way.
I think this circuit would work if tuned properly; one concern I have is if the transformer's output voltage isn't linear during core saturation, the circuit will turn on prematurely.
CapJohn --
Hey, good to see a familiar face! Now you know why I haven't been posting much... to busy building my GainClone! I thought the amp part was going to be tricky; turns out the hard part is the freakin' chassis!! If I tap another freakin' crooked hole, or break another $4 drill bit, I think I'm gonna scream.
My (untested) dislike to the thermistor startup is about the same as Richie's -- I'm concerned that there will be a delay turning on full voltage when the amp has been idle, and that it will add extra noise or choo-choo-ing in the system.
My *like* for the thermistor startup is the symplicity, and zen-like nature: it's on when it needs to be on.
These are the two issues (aside from cost and part count) I tried to address in my circuit.
Wes
Just flip the switch slowly...
Hi zycomatique,
They are both types of inrush current limiting. A soft start circuit can be a timed relay across a resistor used to reduce the peak charging current. This is shorted out once the relay times out and after the high rush is over. A (negative co-efficient) thermister has a higher resistance cold than hot. The inrush and normal running current causes the temperture to increase, sometimes quite a bit, and that causes the resistance to drop. It will always have higher resistance than a relay type. NTC thermistors are better suited to tube amplifiers that have a higher steady state current as opposed to most AB transistor amplifiers that can draw very little current until you start delivering a lot of power (crank it!). Those are much better suited to a soft start circuit.
Hi amplidude,
Your response wasn't very helpful at all.
-Chris
They are both types of inrush current limiting. A soft start circuit can be a timed relay across a resistor used to reduce the peak charging current. This is shorted out once the relay times out and after the high rush is over. A (negative co-efficient) thermister has a higher resistance cold than hot. The inrush and normal running current causes the temperture to increase, sometimes quite a bit, and that causes the resistance to drop. It will always have higher resistance than a relay type. NTC thermistors are better suited to tube amplifiers that have a higher steady state current as opposed to most AB transistor amplifiers that can draw very little current until you start delivering a lot of power (crank it!). Those are much better suited to a soft start circuit.
Hi amplidude,
Your response wasn't very helpful at all.
-Chris
Hello, someday I removed the switch and replaced it by a solid wire and heard better sound especially in bass. So I understood that on the 230 Vac path the less resistance was the better for sound so your answer wich says that soft start has less resistance than thermistor Is an argument for better sound too.
Thank you
Thank you
Hi zycomatique,
No for the power switch, you are not getting any improvement at all. You just think you are. Switch contact resistance is less than 0.1 ohm.
I guess in a pure, measured sense in a tube amplifier the soft start might be better. In the real world a properly sized NTC thermistor would be fine and you wouldn't hear it at all. In a solid state amplifier, excepting maybe a classic class A amplifier, an NTC thermistor is the wrong device to use - so not recommended.
Try not to read reports without measured results. All kinds of false information and silly ideas flourish that way. People tend to repeat things without any thought, and if you think about replacing a power switch with wire (shorting it out), you will quickly realize it cannot be true. That's if you understand even the basics of electronics. Same for power cords and all kinds of fads popular in the high end community. These people will defend their ideas with their dying breath even faced with clear evidence they are wrong. So save your money and think carefully about suggestions that come your way. ... and please, put the power switch back in.
-Chris
No for the power switch, you are not getting any improvement at all. You just think you are. Switch contact resistance is less than 0.1 ohm.
I guess in a pure, measured sense in a tube amplifier the soft start might be better. In the real world a properly sized NTC thermistor would be fine and you wouldn't hear it at all. In a solid state amplifier, excepting maybe a classic class A amplifier, an NTC thermistor is the wrong device to use - so not recommended.
Try not to read reports without measured results. All kinds of false information and silly ideas flourish that way. People tend to repeat things without any thought, and if you think about replacing a power switch with wire (shorting it out), you will quickly realize it cannot be true. That's if you understand even the basics of electronics. Same for power cords and all kinds of fads popular in the high end community. These people will defend their ideas with their dying breath even faced with clear evidence they are wrong. So save your money and think carefully about suggestions that come your way. ... and please, put the power switch back in.
-Chris