ble0t,
That site was the most helpful and quite informative. I ordered a 10R 300W resistor and a timed relay rated for 20A on the switch. It's internal circuitry takes care of the timed relay activation, and one uses a simple resistor to set the delay. It's driven from 3 to 15V. Part number is CC1175-ND. You can read the datasheet here. I plan on using a small tiny transformer, a bridge rectifier, filter cap, and a linear voltage regulator. I'm confident this will do the trick for me, supress the inrush and the current spike from the discharged power caps. Anything I'm missing?
Thanks.
That site was the most helpful and quite informative. I ordered a 10R 300W resistor and a timed relay rated for 20A on the switch. It's internal circuitry takes care of the timed relay activation, and one uses a simple resistor to set the delay. It's driven from 3 to 15V. Part number is CC1175-ND. You can read the datasheet here. I plan on using a small tiny transformer, a bridge rectifier, filter cap, and a linear voltage regulator. I'm confident this will do the trick for me, supress the inrush and the current spike from the discharged power caps. Anything I'm missing?
Thanks.
2kVA is a large beast, you need to have some sort of inrush protection...
Here's the circuit I use for my 1KVA units, you must change the 330nf cap to other value 220 or 470 - trial & error (sorry I don't know it) for it to work on 110Vac/60hz. Very simple, works like a charm, the 2x470uf caps can be increased to 3x470uf for more delay at startup, I'd also change to 15W resistors in your case.
http://mitglied.lycos.de/Promitheus/delay_circuit_for_toroids.htm
Here's the circuit I use for my 1KVA units, you must change the 330nf cap to other value 220 or 470 - trial & error (sorry I don't know it) for it to work on 110Vac/60hz. Very simple, works like a charm, the 2x470uf caps can be increased to 3x470uf for more delay at startup, I'd also change to 15W resistors in your case.
http://mitglied.lycos.de/Promitheus/delay_circuit_for_toroids.htm
lucpes,
I purchased one 10R 300W resistor. I'm going to use this with the timed relay I purchased. Do I need more resistance? I'd assume 20R is enough, because high resistance will keep fuses from blowing. I have seen tons of delay circuits, but digikey sells these timed relays. Just connect a timing resistor and DC and you are good to go. The one I have is rated 20A at the switch. A simple power supply will do, I'll use a small transformer, bridge, filter cap, and regulator to drive the relay.
I purchased one 10R 300W resistor. I'm going to use this with the timed relay I purchased. Do I need more resistance? I'd assume 20R is enough, because high resistance will keep fuses from blowing. I have seen tons of delay circuits, but digikey sells these timed relays. Just connect a timing resistor and DC and you are good to go. The one I have is rated 20A at the switch. A simple power supply will do, I'll use a small transformer, bridge, filter cap, and regulator to drive the relay.
ble0t,
Yeah, his article was the best in terms of theory that I really needed to know. I'll let you all know how it works out. The resistor was $16 at digikey. It's kind of expensive, but if you need it them you need it. No one said an electronics hobby was cheap.
Thanks everyone for all your help.
Yeah, his article was the best in terms of theory that I really needed to know. I'll let you all know how it works out. The resistor was $16 at digikey. It's kind of expensive, but if you need it them you need it. No one said an electronics hobby was cheap.
Thanks everyone for all your help.Well, I just wanted to tell you guys NOT to use an inrush thermister on a 2kVA toroid, it does nothing but spark, burn, and explode with red hot leads. My next option is a solid state timed relay but I haven't gotten it to operate very well. With no load, it does absolutely nothing. It has an idle current of 10mA, so a neon lamp that used 2mA was always lit. I then tried my portable TV that draws around 550mA. It worked the first time, but after that it did not, Then I tried a craftsman flourescent work light that was rated for 13A and it still instantly turned on. Do I have to have everything connected up the way it should and then the relay will operate the way it should? I'd expect a $40 relay to do it's job. If it fails, I'm going to have to buoild myself a delay circuit. Anyone have experience with solkid state relays?
Thanks.
My next option is a solid state timed relay but I haven't gotten it to operate very well. With no load, it does absolutely nothing. It has an idle current of 10mA, so a neon lamp that used 2mA was always lit. I then tried my portable TV that draws around 550mA. It worked the first time, but after that it did not, Then I tried a craftsman flourescent work light that was rated for 13A and it still instantly turned on. Do I have to have everything connected up the way it should and then the relay will operate the way it should? I'd expect a $40 relay to do it's job. If it fails, I'm going to have to buoild myself a delay circuit. Anyone have experience with solkid state relays?Thanks.
You might be better off using a resistor specifically designed for inrush applications. We had a drive that had a blown inrush resistor. A WW substitute didn't last long. We had to have one made, to the original specifications, before we got it right. These special types were carbon - composition, like the extremely old resistors you used to see in early radio sets.
Similar problems occur with brake resistors in drives, ordinary WW types don't last long.
Hope this helps
Similar problems occur with brake resistors in drives, ordinary WW types don't last long.
Hope this helps
For those high power applications I think that Triac-based non-dissipative phase ramp startup is much more convenient. This method ensures a fixed startup peak current and allows for gadgets such as overvoltage protection or secondary side controlled shutdown. It also prevents transformer saturation at startup and generates a smooth voltage ramp across the main filter capacitors
The control circuit may be accomplished with a cheap LM339 quad comparator and some transistors or with a simple PIC12C508 microcontroller. That kind of control circuit may be powered from mains [not for testing!!] through a ballast resistor, a zener and a capacitor due to its very low current requirements [<2mA]. In order to drive high power triacs or thyristors without drawing excessive current, 10% duty cycle gate pulses are advidsed and a small ferrite step-down pulse transformer is also useful
I'm currently testing a controlled-rectifier [full bridge with 2 diodes and 2 thyristors] prototype that provides soft start to smoothly charge a big 2.200uF capacitor from 230V AC for SMPS applications. It also provides shutdown so the SMPS is able to instantaneously remove power to itself in case of failure or overvoltage
The control circuit may be accomplished with a cheap LM339 quad comparator and some transistors or with a simple PIC12C508 microcontroller. That kind of control circuit may be powered from mains [not for testing!!] through a ballast resistor, a zener and a capacitor due to its very low current requirements [<2mA]. In order to drive high power triacs or thyristors without drawing excessive current, 10% duty cycle gate pulses are advidsed and a small ferrite step-down pulse transformer is also useful
I'm currently testing a controlled-rectifier [full bridge with 2 diodes and 2 thyristors] prototype that provides soft start to smoothly charge a big 2.200uF capacitor from 230V AC for SMPS applications. It also provides shutdown so the SMPS is able to instantaneously remove power to itself in case of failure or overvoltage
emuman100 said:Well, I just wanted to tell you guys NOT to use an inrush thermister on a 2kVA toroid, it does nothing but spark, burn, and explode with red hot leads.
Thanks.
Everyone on this forum seems to promote the use of thermistors
include Pass. The real question to ask is... what differs from
your implementation of it vs. the others that have had good
success?
thylantyr,
They can't handle the high current inrush. This solid state timed relay operates great with a 10R 10W resistor. Thing is, my power supply doesn't work right. The positive side outputs 90V while the negative side outputs 108V (as I calculated). I did some damage to the diode bridge used on the positive side, I think that is why. I connected just the negative side to the amp and it blew the output caps on the positive side and the breaker. You know why?
Thanks.
They can't handle the high current inrush. This solid state timed relay operates great with a 10R 10W resistor. Thing is, my power supply doesn't work right. The positive side outputs 90V while the negative side outputs 108V (as I calculated). I did some damage to the diode bridge used on the positive side, I think that is why. I connected just the negative side to the amp and it blew the output caps on the positive side and the breaker. You know why?
Thanks.
Inrush limit
I encounterd a similar problem in the desing of a sub amp for a premium speakr. It also had a 2 KVA toroid. the problem with the big toroid is that if it is shut off at the wrong part of the cyccle the core is saturated and it looks line a dead short to the AC line.
What I did was to insert a large resistor 20 Ohms I think in series with a relay that closed when the supply reached most of its voltage. A 48V relay on a 60V rail, for example, will close when the supply has come up enough and the transformer core is no longer saturated. It should be suppressed with a .1 cap snubber to protect the relay contacts.
The delay comes from charging the power supply. And its simple.
Solid state relays are not well suited to this application. Your sample may have an internal snubber thats running the light with the relay off. And they can also generate noise.
I have also been interested in SCR phase controlled regulators but again noise is a concern.
-Demian
I encounterd a similar problem in the desing of a sub amp for a premium speakr. It also had a 2 KVA toroid. the problem with the big toroid is that if it is shut off at the wrong part of the cyccle the core is saturated and it looks line a dead short to the AC line.
What I did was to insert a large resistor 20 Ohms I think in series with a relay that closed when the supply reached most of its voltage. A 48V relay on a 60V rail, for example, will close when the supply has come up enough and the transformer core is no longer saturated. It should be suppressed with a .1 cap snubber to protect the relay contacts.
The delay comes from charging the power supply. And its simple.
Solid state relays are not well suited to this application. Your sample may have an internal snubber thats running the light with the relay off. And they can also generate noise.
I have also been interested in SCR phase controlled regulators but again noise is a concern.
-Demian
Hi
I once had probs with inrushcurrent from a 400V 3-phase 250kVA transformer (load was the 12kV/20Adc anode of a transmitting tube).
The inrush-current was in kA ranges.
Tried to use 50+ pieces (per phase) of metalcased 300W resistors in paralell (per phase) mounted on huge heatsinks.
Did only last some weeks before they popped 1 after n other.
The problem got cured after I realized that what I need are resistors
with high HEATCAPACITY. In this respect ceramic is the best choice.
I finally used only glasscoated resistors wich had a thick and short ceramic body (long,thin tubetypes cannot cope nearly as good , even when theyr mass would be the same).
It is vital to understand "heatcapacity" of resistors in applications like
inrushcurrent-limiting. A short, full ceramic body, glass coated resistor
will outperform a bank of metalcased resistors with over 20 times higher (continious) power-rating.
From this I would advice to tackle the inrush-current of 2kVA toroids by simply using the cascaded turn-on delay of 2-3 relays shortcutting a string of 2-3 sturdy built 10-25W ceramic resistors (for 120V around 2ohm each should be ok).
I once had probs with inrushcurrent from a 400V 3-phase 250kVA transformer (load was the 12kV/20Adc anode of a transmitting tube).
The inrush-current was in kA ranges.
Tried to use 50+ pieces (per phase) of metalcased 300W resistors in paralell (per phase) mounted on huge heatsinks.
Did only last some weeks before they popped 1 after n other.
The problem got cured after I realized that what I need are resistors
with high HEATCAPACITY. In this respect ceramic is the best choice.
I finally used only glasscoated resistors wich had a thick and short ceramic body (long,thin tubetypes cannot cope nearly as good , even when theyr mass would be the same).
It is vital to understand "heatcapacity" of resistors in applications like
inrushcurrent-limiting. A short, full ceramic body, glass coated resistor
will outperform a bank of metalcased resistors with over 20 times higher (continious) power-rating.
From this I would advice to tackle the inrush-current of 2kVA toroids by simply using the cascaded turn-on delay of 2-3 relays shortcutting a string of 2-3 sturdy built 10-25W ceramic resistors (for 120V around 2ohm each should be ok).
I also use a 2KV torroid on in my amp (special low noise variery) that has significant in rush currents.
I ended up using a microcontroller to drive a 16A relay. Initially, I have a c. 15 ohm 40W resistor in series with th e primary, and then after about 1.5 secs, the relay comes in.
Also, dont forget that as you r filter caps charge up, th e main s current is quite significant. I did an LTspice analysis and for 44mF per rail, I got someting like 400A pk.
If you ar e not into micro's, the Elektor circuit put up by Lucpes a few posts back is quite a popular way to go.
I ended up using a microcontroller to drive a 16A relay. Initially, I have a c. 15 ohm 40W resistor in series with th e primary, and then after about 1.5 secs, the relay comes in.
Also, dont forget that as you r filter caps charge up, th e main s current is quite significant. I did an LTspice analysis and for 44mF per rail, I got someting like 400A pk.
If you ar e not into micro's, the Elektor circuit put up by Lucpes a few posts back is quite a popular way to go.
A 555 timer can get you a 1.5 second delay. The sensing of the voltage on the filter caps is the best way to know if they are charged and the current demand has dropped. While a microcontroller isn't expensive using one for strictly a timer doesn't make sense.
I will admit my first amp used a microcoded bit slice processor to operate an on-off switch. it was fun but made no sense, and was much more expensive than the amp itself. The subsequent production units replaced all that with a nice looking switch lifted from the electrical business. Good looking, nice feel 30A and $5.00 at the time. I think they are cheaper 30 years later. Decora
I will admit my first amp used a microcoded bit slice processor to operate an on-off switch. it was fun but made no sense, and was much more expensive than the amp itself. The subsequent production units replaced all that with a nice looking switch lifted from the electrical business. Good looking, nice feel 30A and $5.00 at the time. I think they are cheaper 30 years later. Decora
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.