-
Hey everybody, I'm studying in detail how SMPS power supplies work and I've got quite a lot of stuff figured out already. However, I hit a wall
with a specific aspect of the inrush input protection when using an NTC thermistor.
I'm trying to "fix" a cheap 24V/15A power supply to make it more reliable. So far the project has been very succesful (replaced capacitors with good quality ones, fuse with the correct rating one, rebuilt a high voltage pcb track that didn't have enough clearance, etc). Now one of the last things on my to-do list is checking if the thermistor is correctly rated for this power supply as I suspect the manufacturer just chose the most commonly available part and stuck it in there instead of doing the proper research.
The thermistor currently in use is a "NTC 5D-15" from a chinese manufacturer. The data for this thermistor is 5R, 6A max steady current, 0.132R approx max resistance at max current, 20mW/C° power dissipation coefficient, 76 time constant.
Here's my research so far:
I calculated how many Joules my system will absorb at startup. The SMPS has two 250V, 680uF capacitors in series (that should result in 340uF) and I've got a result of 23.7J with a Vpeak of 373 (derived from 240Vrms + 10%).
That's one of the parameters required for choosing a NTC thermistor, so far so good.
The second parameter is the resistance. The formula I found is Vpeak / Max inrush current (A) and this is where I'm having trouble: how can I calculate this max inrush current? Is that some arbitrary amount I need to choose? Is that based on a reasonable amount that's low enough not to trip the circuit breaker?
I don't think I have the proper equipment to measure it, unfortunately, so I need to calculate it somehow (my oscilloscope is pretty much a toy and I would never use that at mains voltage levels).
Thank you so much for reading this!
TL:DR;
The SMPS has an output rating of 15A at 24V, powered using 240V. Two 680uF caps in series and the NTC thermistor is a 5D-15, with 5 Ohms and 6A max current. I want to check if this is the correct value and I need to understand the Max inrush current (A). How is it calculated? -
At instant of turn-on, caps are at zero and worst-case line is at peak 340V. How much voltage drops across the NTC? How much current does that draw?
There may be some finagling about whether actual-peak current matters, or RMS over a full cycle or quasi-peak or something like that. You'd have to read the relevant standards to see how it's measured.
Tim -
Thank you for your reply, Tim.
I found this online calculator, hopefully I'm on the right track: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html
By playing with it a little bit it seems like a 5 Ohm NTC will allow for something like 70A through the instant the supply turns on. It takes about 5ms (3RC) to get to a much more comfortable 3.something A.
So, if the above is correct, that means 5R is not the ideal choice, as I read here (http://sound.westhost.com/articles/inrush.htm):QuoteMost manufacturers will use a negative temperature coefficient (NTC) thermistor to limit the inrush current, but it's not at all uncommon for them to get the value horribly wrong. One that I recently came across used a 4 ohm thermistor - completely useless, and I was able to measure 80A inrush peaks easily.
Now, what would be a better value? A 10R NTC would give me 37A: much better already. The 20R version brings it down to 18A, but it's less efficient afterwards as it will cause a higher voltage drop. There's an intermediate choice at 14R, too.
So, given the data above, should I go with the highest resistance that still allows me to be within specs? (for example, 20R?) Or should I choose a 10R which is more efficient when on a steady state?
I guess my question really is: how much inrush current should I be comfortable with, given the choices? -
Now, what would be a better value? A 10R NTC would give me 37A: much better already. The 20R version brings it down to 18A, but it's less efficient afterwards as it will cause a higher voltage drop. There's an intermediate choice at 14R, too.
There is no ideal choice: You can choose either a high inrush current + low steady state losses or lower inrush current + higher steady state losses. To get a high efficency many manufacturers use a rather low value for the NTC. You often see 20-40A inrush current even for a small SMPS.
So, given the data above, should I go with the highest resistance that still allows me to be within specs? (for example, 20R?) Or should I choose a 10R which is more efficient when on a steady state?
I guess my question really is: how much inrush current should I be comfortable with, given the choices?
In a good quality SMPS you often see a relay bypassing the NTC after a short time. That allows a high value NTC for a good soft start but eliminates the steady state losses. -
Thanks for your answer, bktemp.
I read about a few different ways of achieving a better efficiency (in the sound.westhost.com article I linked). Given the SMPS is built on a simple single sided board and there's actually plenty of space for modifications, I wonder if I could implement at least a simple bypass system.