Thermistor Location

[awhiteguy]

From searching, I have seen that alot of people use thermistors for inrush protection.

Where is the best location?
1. On each secondary rail between the bridge and caps
2. On the primary side of the transformer

I see that it is done both ways. Someone suggested using the CL-60 device made by NTC. The datasheet for this device shows the thermistor in location 1. The datasheet also shows a Cx, maximum capacitor size. For the CL-60 it is 5000uF at 120vac, but my rails are going to be 70vdc.

How do I know how many of the thermistors I need to parallel at my given voltage? I will be using 40000uF caps at +- 70v rails and a 750va toroid. I would guess like 3 or 4 in parallel for each rail, but I don't like guessing. Thanks,

awhiteguy

[dlharmon]

I wouldn't use thermistors for that. I would put some large power resistors between the bridge and the caps, and short them with a relay after a delay of 5 seconds or so. At the same time another relay could be thrown connecting the caps to the load. I wired something like this up in my power supply with 2x 80000uF 35V because it was cheaper and better than thermistors.

Darrell Harmon

[peranders]

Originally posted by awhiteguy
From searching, I have seen that alot of people use thermistors for inrush protection.

How do I know how many of the thermistors I need to parallel at my given voltage? I will be using 40000uF caps at +- 70v rails and a 750va toroid. I would guess like 3 or 4 in parallel for each rail, but I don't like guessing.

NTC, forget it! They are well suited for switched power supplies in order not to kill the caps and the rectifier bridge.

Use as suggested a resitor 100 ohm/5-10 W along with a relay which are delayed on (0.5-2 sec will do).

Your big problem isn't the 40000µF, it's the toroid. Connect this big mama without any load at all. You will still get inrush currents.

I use a 11-pin plug-in relay with a timer module and 100 ohms/10W. The time is 1 s. The transformer is 500 VA with a max inrush current of 77 A at 230 VAC. With 100 ohms the max current will be 3 A.

[dlharmon]

I said to put the resistor between the bridge and the caps, but it really should be in the ac line. You could use a triac instead of the relay if you put it there.


Darrell Harmon

[awhiteguy]

I was under the impression that it was the capacitors that caused the inrush, but you guys are saying that the inrush is due to the transformer. That would mean that the power resistors must be on the primary side of the transformer

I use a 11-pin plug-in relay with a timer module and 100 ohms/10W

Peranders, when you say "module" are you refering to a packaged device or a circuit board you made yourself?

You could use a triac instead of the relay

What is the function of a triac? Anyone know of any schematics I could look at when a triac is used in a PS?

awhiteguy

[peranders]

Originally posted by dlharmon
I said to put the resistor between the bridge and the caps, but it really should be in the ac line. You could use a triac instead of the relay if you put it there.

Yes, I thought you ment that (didn't read carefully). It is the transformer which creates the currents all alone (more or less) if the caps aren't huge.

[Peter Daniel]

Sorry to disapoint you guys but resistor won't work in all cases. I tried fixed resistor with Aleph and depending on the value, there was always voltage drop on a resistor, never allowing the caps to be fully charged. In that case when the relay engages you will still get inrush current. I settled on two thermistors connected in series, because they actually worked the best (on primary).

[dlharmon]

He has a lot of capacitance (+-70V 40000uF), and it seems it would take a lot of thermistors. My ps also has huge amounts of capacitance. The reason for the relay to connect the load is so the caps can charge without the load present.

Darrell Harmon

[daatkins]

Put them on the primary or between the sec and caps!!

Don't make the mistake I made and put the thermistors between the transformer secondary and the bridge rectifier. One thermistor had a slightly lower resistance characteristic so initially more current took that route, so the thermistor warmed up and lowered it's resistance so more current took that route.... And so on. The effect was that one thermistor and one pair of rectifier diodes in the bridge was passing all the charging current while the other half of the circuit did nothing.

Nice one,
David.

[Circlotron]

Put a thermistor on the input side of the tranny, and a relay across it, timed to shut after several seconds when the thermistor gets hot and lowers it's resistance. You could use a higher than normal resistance thermistor for an even softer start, knowing that it will be bypassed shortly and not overheat. Best of both worlds here.

GP.