Hi Bonsai,
there are two different conditions at work here.
1.) in rush current into a de-energised toroid
2.) charging a capacitor input filter.
1.) the inrush current is best dealt with by a current limiter in the transformer primary circuit. This can usually be done with either Thermistor or by a power resistor. The resistor solution must have a time delayed bypass of about 200mS to 500mS. The Thermistor also benefits from the time delayed bypass.
2.) the capacitor input filter is better solved by fitting a current limiter in the secondary of the transformer. This again can be resistor or thermistor.
Depending on the value chosen then a bypass can be used or for very low values of limiting resistor can be omitted. The delay used here can be from 1000mS to 10S.
Do not try to do both duties with one circuit. Do not compromise the primary circuit by increasing the time delay to slow the capacitor charging rate.
Now back to in rush current. I do not believe inductance and phase are an issue at first start up. Once the transformer is working i.e. energised, then we have a completely different senario. Phase may be a problem during this second period, but again I believe that the transformer load appears as a slightly reactive load not as a pure inductance. The 90 degree phase angle never happens in practice and we never even get close to 90degrees.
there are two different conditions at work here.
1.) in rush current into a de-energised toroid
2.) charging a capacitor input filter.
1.) the inrush current is best dealt with by a current limiter in the transformer primary circuit. This can usually be done with either Thermistor or by a power resistor. The resistor solution must have a time delayed bypass of about 200mS to 500mS. The Thermistor also benefits from the time delayed bypass.
2.) the capacitor input filter is better solved by fitting a current limiter in the secondary of the transformer. This again can be resistor or thermistor.
Depending on the value chosen then a bypass can be used or for very low values of limiting resistor can be omitted. The delay used here can be from 1000mS to 10S.
Do not try to do both duties with one circuit. Do not compromise the primary circuit by increasing the time delay to slow the capacitor charging rate.
Now back to in rush current. I do not believe inductance and phase are an issue at first start up. Once the transformer is working i.e. energised, then we have a completely different senario. Phase may be a problem during this second period, but again I believe that the transformer load appears as a slightly reactive load not as a pure inductance. The 90 degree phase angle never happens in practice and we never even get close to 90degrees.
Hi Andrew,
That's what I was about to say.
At startup there is no current flow. You must apply an emf across the transformer to get things moving first. You could use a triac as the switch in series with the resistor as the resistor would make the circuit look primarily resistive. Once the resistor is shorted out (you could use a triac here too) the load would then be primarily reactive.
Hi Bonsai,
-Chris
That's what I was about to say.
At startup there is no current flow. You must apply an emf across the transformer to get things moving first. You could use a triac as the switch in series with the resistor as the resistor would make the circuit look primarily resistive. Once the resistor is shorted out (you could use a triac here too) the load would then be primarily reactive.
Hi Bonsai,
That would be a very basic step in the design. There is no excuse these days for this to occur. It comes down to using the correct part and designing the circuit properly. The app. notes for the device should point the way clearly enough.
-Chris
Andrew, they're on Photobucket. I can see them, and Chris can see them.
Maybe they will work for you this way.
http://i56.photobucket.com/albums/g196/dkleitsch/Inrush2.gif
http://i56.photobucket.com/albums/g196/dkleitsch/Inrush.gif
Chris, I'm not sure why they did it this way, although the amp is rated at 6.6KW. McIntosh just uses four inrush limiters in parallel on their big amps on the primary, and shorts them out with a relay (same basic idea at BGW). I had a 10 ohm inrush limiter blow up once on a 1.2KW amp, the surge was just too much for the part. I switched to a 2R5 type and have never had problems after that. The 2R5 part is only rated at 6A, so if the relay dropped out (from the output stage dumping) it would go too, but that is a secondary failure, and not a cause (I'm willing to live with it, BGW actually uses a fuse to protect theirs).
Crest uses a pair of the same SG328 in series (with a shunt relay) on the CA12 and 8002 (both about 3KW total), and no secondary side resistors. The SG328 is rated at 1 ohm cold and 30A maximum. The CA6 and CA9 Crest amps use just one.
http://www.rtie.com/ntc/sgspec.htm
Maybe they will work for you this way.
http://i56.photobucket.com/albums/g196/dkleitsch/Inrush2.gif
http://i56.photobucket.com/albums/g196/dkleitsch/Inrush.gif
Chris, I'm not sure why they did it this way, although the amp is rated at 6.6KW. McIntosh just uses four inrush limiters in parallel on their big amps on the primary, and shorts them out with a relay (same basic idea at BGW). I had a 10 ohm inrush limiter blow up once on a 1.2KW amp, the surge was just too much for the part. I switched to a 2R5 type and have never had problems after that. The 2R5 part is only rated at 6A, so if the relay dropped out (from the output stage dumping) it would go too, but that is a secondary failure, and not a cause (I'm willing to live with it, BGW actually uses a fuse to protect theirs).
Crest uses a pair of the same SG328 in series (with a shunt relay) on the CA12 and 8002 (both about 3KW total), and no secondary side resistors. The SG328 is rated at 1 ohm cold and 30A maximum. The CA6 and CA9 Crest amps use just one.
http://www.rtie.com/ntc/sgspec.htm
Hi Andrew,
I guess they are spreading out the dissipation a bit, doing limiting in more than one location.
Hi richie00boy,
Hi Andy,
-Chris
I guess they are spreading out the dissipation a bit, doing limiting in more than one location.
Hi richie00boy,
My guess would be to cut the dissipation in half. I do this with LED lighting on AC some times.
Hi Andy,
I don't think it would be as much as you would think. Once the relay shorts them out is issue is gone as well. The huge current pulse caused when there is no inrush limiting would affect everything far worse as well.
-Chris
Hi Andy,
-Chris
True, but this has nothing to do with inrush current.
Well ......... what ever you install as the losser element, it should be removed from the circuit once it's job is done. Otherwise any parts represent resistance in the line. Not good.
Nothing should be more complicated or simpler than it needs to be. You still should be using a relay or triac to short these parts out (removing them from the circuit).
-Chris
Update on Inrush Limiting on the Monster Toroid.
I was doing final tests on the method I was using in which I was plugging in the trafo first and then throwing a switch to the secondary circuit. The trafo got into a state where it could not overcome the inrush current it was drawing. Each time I would plug it in, my fast blow fuses would go "pop".
So I added a 300 ohm resistance in series with the primary and placed a switch across it. It has started smoothly and consistently from then on. Ideally, the resistance would be rated for 50w or so. I did not have an optimal value on hand, so I soldered 3 100ohm 3w ones in series. Now if I were to forget to throw the primary switch, the center resister should unsolder itself. It is free to fall.
The wattage of the three resistors combined is close enough to the value needed that it gives them time to heat gradually enough and not blow. This would be a good place for the 60w light bulb trick. I may do it since I just remembered that I have some extra light sockets.
I was doing final tests on the method I was using in which I was plugging in the trafo first and then throwing a switch to the secondary circuit. The trafo got into a state where it could not overcome the inrush current it was drawing. Each time I would plug it in, my fast blow fuses would go "pop".
So I added a 300 ohm resistance in series with the primary and placed a switch across it. It has started smoothly and consistently from then on. Ideally, the resistance would be rated for 50w or so. I did not have an optimal value on hand, so I soldered 3 100ohm 3w ones in series. Now if I were to forget to throw the primary switch, the center resister should unsolder itself. It is free to fall.
The wattage of the three resistors combined is close enough to the value needed that it gives them time to heat gradually enough and not blow. This would be a good place for the 60w light bulb trick. I may do it since I just remembered that I have some extra light sockets.
Hi,
you use rather high value resistors for this duty, but it works.
The Mains light bulb with a metal filament has a positive temperature coefficient which is the opposite to what is required.
A carbon filament bulb would be much better but it would be easier to find Thermistors, which will do better yet.
you use rather high value resistors for this duty, but it works.
The Mains light bulb with a metal filament has a positive temperature coefficient which is the opposite to what is required.
A carbon filament bulb would be much better but it would be easier to find Thermistors, which will do better yet.
Hi,
I have some 10ohm 10w and some .5ohm 20w. The former is susceptible to quickly burning out while the latter is too low in resistance value for my needs. It could be different if the resistor had to provide current to charge filter caps. Now, though, the secondary has its own switch which I engage after the few seconds the transformer takes to stabilize.
I have some 6ohm cold, NTC thermistors, one of which was in the circuit during earlier trials. It crumbled after one particularly bad start-up attempt. I feel more comfortable with the high value resistance way, based on my experiences.
I have some 10ohm 10w and some .5ohm 20w. The former is susceptible to quickly burning out while the latter is too low in resistance value for my needs. It could be different if the resistor had to provide current to charge filter caps. Now, though, the secondary has its own switch which I engage after the few seconds the transformer takes to stabilize.
I have some 6ohm cold, NTC thermistors, one of which was in the circuit during earlier trials. It crumbled after one particularly bad start-up attempt. I feel more comfortable with the high value resistance way, based on my experiences.
Apologies for a late return to this thread.
Andrew, I have combined both transformer in rush limiting and filter cap pre-charge and it works very well. So I do not see the point of handling these two issues with separate circuits - it leads to additional cost and complexity when in fact its not required.
Andrew, I have combined both transformer in rush limiting and filter cap pre-charge and it works very well. So I do not see the point of handling these two issues with separate circuits - it leads to additional cost and complexity when in fact its not required.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.