My PSU design incorporates separate filament and B+ transformers and line switches. I have heard that a current inrush limiter on the B+ side is strongly recommended. My questions are: what kind of a device should I use, and where in the PSU should it be placed?
Thanks in advance for your help.
Dave
Thanks in advance for your help.
Dave
Are you using SS or tube rectification?
They are usually placed either on the mains side of the mains transformer, or the center tap of the transformer secondary (assuming FWCT rectification). Sometime folks use them in series on the B+ after the rectifier as well.
Typically CL-90, CL140, CL150 or thereabouts. These devices have a high resistance when they are cold and a much lower R when they heat up, so you can apply a little ohms law to figure out what size you need where.
http://www.ge-mcs.com/download/temperature/920_325a.pdf
They are usually placed either on the mains side of the mains transformer, or the center tap of the transformer secondary (assuming FWCT rectification). Sometime folks use them in series on the B+ after the rectifier as well.
Typically CL-90, CL140, CL150 or thereabouts. These devices have a high resistance when they are cold and a much lower R when they heat up, so you can apply a little ohms law to figure out what size you need where.
http://www.ge-mcs.com/download/temperature/920_325a.pdf
Last edited:
It is worth noting that placing the NTC thermister on the mains side alleviates inherent transformer inrush current, as well as initial SS diode or pre-heated tube rectifier surge current pulses. This can assist in using a mains side fuse (and possibly secondary side fuses depending on location) that is more closely matched to maximum operating rms levels, rather than having an extra margin included for the in-rush (and hence a lower level of protection).
More design effort is required with thermisters, as they get hot, and have a design range that needs assessment. And they take significant time to 'reset'.
Note that although a pre-heated tube diode inrush may be reduced by a thermistor, it doesn't per se change the peak repetitive current level through the diode - and a tube diode typically has a number of 'peak' specs that have to be cross-checked.
The thermister is not as effective with reducing heater inrush, due to heater thermal time constant - although valve heaters are designed for inrush, and seldom fail from heater related issues, it is always better imho to reduce stress where practical.
Tim
More design effort is required with thermisters, as they get hot, and have a design range that needs assessment. And they take significant time to 'reset'.
Note that although a pre-heated tube diode inrush may be reduced by a thermistor, it doesn't per se change the peak repetitive current level through the diode - and a tube diode typically has a number of 'peak' specs that have to be cross-checked.
The thermister is not as effective with reducing heater inrush, due to heater thermal time constant - although valve heaters are designed for inrush, and seldom fail from heater related issues, it is always better imho to reduce stress where practical.
Tim
I use the CL-90. I get them at Digikey. They do the job... I put the thermistor on the primary in series with the mains fuse.
One thing about thermistors like the CL-90 in general is that during normal operation, the device is fairly hot. Hence, you need to allow it to cool before it'll be effective for inrush limiting again. It'll limit the inrush when you first turn on the equipment, but will not limit any inrush on voltage glitches or if you quickly power cycle the equipment. One way to fix this is to short out the thermistor with a relay. I use a simple RC timer and a bipolar transistor to drive a relay shorting the thermistor about 0.1~0.2 seconds after turn-on.
~Tom
One thing about thermistors like the CL-90 in general is that during normal operation, the device is fairly hot. Hence, you need to allow it to cool before it'll be effective for inrush limiting again. It'll limit the inrush when you first turn on the equipment, but will not limit any inrush on voltage glitches or if you quickly power cycle the equipment. One way to fix this is to short out the thermistor with a relay. I use a simple RC timer and a bipolar transistor to drive a relay shorting the thermistor about 0.1~0.2 seconds after turn-on.
~Tom
- Status
- Not open for further replies.