I guess if all you were worried about was the in rush relating to filter capacitor charging it doesn't matter whether it is on the primary side or secondary side given the proportional relationship between Is and Ip. But if you are also concerned by the in-rush associated with magnetising the transformer then the resistance needs to be on the primary side, no? If both then managing them both together from the primary side is most convenient.
Last edited:
convenient, yes.
But you (and many others) are alleging that slow charging is OK, if the NTC is in the primary circuit.
You have not produced evidence to show that the NTC manufacturers showing the NTC in the capacitor circuit are being overly cautious, or that they are wrong.
I have all along said the slow charging should copy the method that all the manufacturers show in their appnotes.
Do the comparisons (before, or after the transformer) in the sim with the models you have.
But you (and many others) are alleging that slow charging is OK, if the NTC is in the primary circuit.
You have not produced evidence to show that the NTC manufacturers showing the NTC in the capacitor circuit are being overly cautious, or that they are wrong.
I have all along said the slow charging should copy the method that all the manufacturers show in their appnotes.
Do the comparisons (before, or after the transformer) in the sim with the models you have.
To slow the cap charging one needs to restrict current to them. This can be done at the secondary or, given the proportional nature of primary current and secondary current, at the primary. One need only ensure that an appropriate amount of resistance is in play long enough to achieve the stated goal.
But there are many ways to skin this cat. You could go to all sorts of extremes, for example individually restricting current to each cap.
But there are many ways to skin this cat. You could go to all sorts of extremes, for example individually restricting current to each cap.
No.
A transformer has proportionality between the voltages, not the currents.
Primary T / Secondary T = Primary Voltage / Secondary open circuit Voltage
Andrew, the point I am making is that there is a defined relationship between Is and Ip (defined by the turns ratio) just as there is for Vs and Vp. Restrict current into the primary and there is an impact to current from the secondary. Perhaps I should have said "given the defined relationship between Is and Ip". If my shorthand choice of words was poor then I apologise. But the point remains that one can restrict current flow to the filter capacitance by placing the resistance on either side of the transformer. I'm sure we are boring Gudmund silly...
What about PTC?
Pardon me if I steer the debate to another direction, but I believe I am still on topic.
What about the use of a simple PCT (or parallel of thereof) in series to the primary circuit?
The solution is documented in some manufacture´s notes like:
http://en.tdk.eu/blob/173720/download/5/thermistos-inrushcurrentlimiters-pb.pdf Proper components are readily availabe i.e.
PTC Epcos B59840C120A70, 6Ω ±25%, 17.5 Dia. x 5mm.
http://docs-europe.electrocomponents.com/webdocs/13c0/0900766b813c0c19.pdf
With 5 to 10 of this components in parallel you can protect your typical 500-800 W transformer setup.
There is the advantage that if you are not bound to it by psychological/legal/receptacle constraints you can do away with the primary fuse (a correctly rated fuse will never blow anyway).
So the circuit acts both like a fuse and and a current limiter and it is self protecting. Those components can typically withstand their full rated voltage across and have "open" failure mode, like common resistors.
The price to pay again is a little power loss on the component itself. I you want you can short it out with a relay or a triac after a while, but then , of course, you lose the self protection.
Comments?
Pardon me if I steer the debate to another direction, but I believe I am still on topic.
What about the use of a simple PCT (or parallel of thereof) in series to the primary circuit?
The solution is documented in some manufacture´s notes like:
http://en.tdk.eu/blob/173720/download/5/thermistos-inrushcurrentlimiters-pb.pdf Proper components are readily availabe i.e.
PTC Epcos B59840C120A70, 6Ω ±25%, 17.5 Dia. x 5mm.
http://docs-europe.electrocomponents.com/webdocs/13c0/0900766b813c0c19.pdf
With 5 to 10 of this components in parallel you can protect your typical 500-800 W transformer setup.
There is the advantage that if you are not bound to it by psychological/legal/receptacle constraints you can do away with the primary fuse (a correctly rated fuse will never blow anyway).
So the circuit acts both like a fuse and and a current limiter and it is self protecting. Those components can typically withstand their full rated voltage across and have "open" failure mode, like common resistors.
The price to pay again is a little power loss on the component itself. I you want you can short it out with a relay or a triac after a while, but then , of course, you lose the self protection.
Comments?
PTC is similar to using a Mains Bulb Tester.
It goes into full limiting, i.e. almost no voltage at the primary, when the current is just enough to cause the filament to warm up and resistance to go higher as more heat is developed.
It goes into full limiting, i.e. almost no voltage at the primary, when the current is just enough to cause the filament to warm up and resistance to go higher as more heat is developed.
It has been instructive to read what happens at startup of large transformers.
As I have read the thread, it should work with some serial resistors on the primary side of the transformer, that after about 200ms is bypassed by a relay. This is solved with a small delay circuits.
Since I have a double delay circuit, I will try and bypass the NTC resistance at the transformer's secondary side between 6 and 10 seconds later.
A pair of large ampere meters in series with the NTC resistor on the secondary side of the tranformer must be able to read the actual ampere current during the startup phase.
regards Gudmund
Yes, the bulb behaves like PTC and very much so.
Any "normal" wire resistor has a Positive Temperature Coefficient. What differentiates the commercial PTC is their highly non-linear R vs T characteristics, that is made for that protection circuit purpose.
What I wonder is why I have seen so few use of them in audio Power Supplies circuits. I other applications (i.e. motor drives ) it is quite common.
A PTC thermistor most likely is not a good choice for used as a current limiting device in a soft starter, or an inrush controller/limiter, although that is not to say one cannot design to use a certain PTC in a certain application successfully.
In a power amp, the idle current is usually significant enough to keep the PTC hot and its resistance too high to have the bulk capacitors fully charged up to normal operating voltage. Then when the relay pulls in to short out the PTC it in fact generates a inrush transient to charge up the caps that is right against our objective.
I have recently built a power amp that at first powering up the 60W/120V test bulbs on the supply rails lit brightly and dropped most of the supply voltage. It took me a while to realize there was nothing wrong with the circuit as the amp would start to idle at 350mA when the rails reach +/- 8V. I'd certainly not use a PTC in the inrush circuit in this amp.
In a power amp, the idle current is usually significant enough to keep the PTC hot and its resistance too high to have the bulk capacitors fully charged up to normal operating voltage. Then when the relay pulls in to short out the PTC it in fact generates a inrush transient to charge up the caps that is right against our objective.
I have recently built a power amp that at first powering up the 60W/120V test bulbs on the supply rails lit brightly and dropped most of the supply voltage. It took me a while to realize there was nothing wrong with the circuit as the amp would start to idle at 350mA when the rails reach +/- 8V. I'd certainly not use a PTC in the inrush circuit in this amp.
I use NTC thermistors in soft starters. They are small in size and many are rated for mains circuit use. These properties are Helpful for PCB layout and assembly to meet safety standars with little challenge usually. However NTCs usually have quite finite energy handling capabilities, they are usually rated for certain uF of capacity under either 120 or 240V mains. In audio power amps, especially DIYed ones having banks of zillion Farad capacitors, one would want to calculate the energy storage and watch out not to blow the NTCs up. I use multiple of them in series to increase energy handling in a few of my inrush control circuit with good result.
I had always expected that with a typical rectifier + capacitance on the secondary then the heavy primary current pulses would be mainly thru the diodes with the capacitors to handle turn-off and non linearity.
So the circuit could be considered either as capacitors with diode protection or as turned-on diodes with capacitance to smooth.
That is also what Cordell implies, as I read it, and my comment was based on this idea of a different perspective.
But now I think about it the answer is not entirely obvious and Rod Elliot, for one, contradicts that view.
Was your comment based on Rod's article or have you done some work on this yourself?
Yes, of course, the innovation is to use the same circuit to also control the DC.
This was my assumption too, but only tentative, that's why I posted the idea for comment.
I will run a few simulations.
Best wishes
David
ESP came after DIYaudio.
It appears he read our comments and produced his webpage.
The capacitors have a finite passing current.
If this current is exceeded, then the voltage drop goes too high and the bypass diodes turn on.
The "turned on" diodes are a bypass for the protection of the capacitors.
During transformer start up, or during a abuse incident, it is very likely the capacitors will have excessive Vdrop and it is during these VERY SHORT periods that the protection diodes turn on.
A 300VA 115:30+30Vac transformer run at full power will draw an rms current of nearly 3Arms. The peak current taking account of the capacitor input filter that most of us use may be as high as 5times the rms current.
That would be a maximum current of ~15Apk. That will really stress the ripple capacity of any electrolytic.
You Members running big power on 115Vac supplies have a real problem in not overloading the DC blocking capacitors.
Just as well our audio amplifiers tick over with average power levels around -20dB ref full power. That is what prevents the bypass diodes turning on.
It appears he read our comments and produced his webpage.
The capacitors have a finite passing current.
If this current is exceeded, then the voltage drop goes too high and the bypass diodes turn on.
The "turned on" diodes are a bypass for the protection of the capacitors.
During transformer start up, or during a abuse incident, it is very likely the capacitors will have excessive Vdrop and it is during these VERY SHORT periods that the protection diodes turn on.
A 300VA 115:30+30Vac transformer run at full power will draw an rms current of nearly 3Arms. The peak current taking account of the capacitor input filter that most of us use may be as high as 5times the rms current.
That would be a maximum current of ~15Apk. That will really stress the ripple capacity of any electrolytic.
You Members running big power on 115Vac supplies have a real problem in not overloading the DC blocking capacitors.
Just as well our audio amplifiers tick over with average power levels around -20dB ref full power. That is what prevents the bypass diodes turning on.
if you are reffering to the turns ratio between primary and secondary,
yes, there is such a relationship...it is the square of this ratio...
in adittion, the dc resistance of the secondary winding is reflected to the
primary and added in series with the primary dc resistance.
together they are the ones that limit surge currents in addistion to any resistor in series at turn on....
also, the ac voltage wave point where the primary circuit is closed also has a big impact,
the biggest surge is when the traffo was turned on at the zero crossing point....
placing a resistance in series with either the primary or secondary winding
is a sure way to limit turn on currents....
after all this is why we use soft starting circuits...
No, the idle current it is NOT enough if you use the correct component.
If read the literature and the example I have attached, you will see that.
It is a matter of choosing the correct component, exactly as with NTC.
The fact that your bulb didn't work correctly says nothing about of a properly dimensioned PTC.
Not saying PTC is better than NTC, but it has it's own advantages and disadvantages.
As I said it is a standard technique in other fields, where relevant currents are involved.
yes, and these NTC's are found in ATX psu's, ranging in values from 4 to 12 ohms cold, they become less than 1 ohm when current passes and as they heat up...
- Status
- Not open for further replies.