I actually had to open the amp I built in 1988 to have a look at the soft start, I get the feeling my brain developed a soft start of its own. I only remembered it use some trick circuit but forgot the details.
It puts a 3uF motor start capacitor and a 47R 5W WW resistor in series with the primary for as long as it takes to charge the cap bank to 80%, then they get bridged with a relay. The relay is powered from the main caps via some Z-diodes, so if something is wrong the start cap will never get bridged and power stays limited. The start cap cannot burn out (unlike resistors), the 47R is just to prevent contact welding when the cap gets shorted.
I do not even use speaker relays, the rails come up so slowly (20s) you hear no thump at all (please don’t say DC fault now).
This method only works for low bias, or the rails will not come up far enough to engage the relay.
It puts a 3uF motor start capacitor and a 47R 5W WW resistor in series with the primary for as long as it takes to charge the cap bank to 80%, then they get bridged with a relay. The relay is powered from the main caps via some Z-diodes, so if something is wrong the start cap will never get bridged and power stays limited. The start cap cannot burn out (unlike resistors), the 47R is just to prevent contact welding when the cap gets shorted.
I do not even use speaker relays, the rails come up so slowly (20s) you hear no thump at all (please don’t say DC fault now).
This method only works for low bias, or the rails will not come up far enough to engage the relay.
soft starting the primary is all you will ever need.....
1. soft starting was invented to avoid your lamps dimming at turn on, but since incandescent lamps are almost extinct, and led's do not dim as much as incandescents, no need to worry too much...
2. soft starting was invented to avoid switch contacts from melting or getting welded oa turn on or turn of, the obvious solution is to use switches of sufficient interruption ampere ratings..
3. soft starting also lets you use smaller ampere rated fuses than otherwise would melt without it...
so for 520kufd filtering i will go for soft starting because of the above reasons and for nothing else...
1. soft starting was invented to avoid your lamps dimming at turn on, but since incandescent lamps are almost extinct, and led's do not dim as much as incandescents, no need to worry too much...
2. soft starting was invented to avoid switch contacts from melting or getting welded oa turn on or turn of, the obvious solution is to use switches of sufficient interruption ampere ratings..
3. soft starting also lets you use smaller ampere rated fuses than otherwise would melt without it...
so for 520kufd filtering i will go for soft starting because of the above reasons and for nothing else...
brilliant, will try this also, a capacitor is a good way to limit current without the heat incurred in a resistor....
The current is limited precisely in the primary winding in order to reduce the inrush current of the transformer caused by the hysteresis loop in the magnetic circuit. No one knows in what position of the hysresis loop the transformer was turned off before. But I would install along with a thermistor a circuit breaker with a D characteristic.
In Audio Research amplifiers, inrush currents in each capacitor circuit are typically limited by RL filters. They are probably trying to extend the life of the capacitors, thus reducing the ripple of the charging currents and distributing the charging current more evenly through each capacitor.
The magnetizing current is creating saturation in the core because it might double the value when switched on. Then it is constant even with open secondary.
A saturated transformer creates big spikes in the primary, only ohmic resistance of the winding limits it at up to 10x normal. So a primary resistor can limit it.
The magnetizing current is 90° out of phase, the maximum occurs when the mains is 0V and the energy has to come from the (limited) flux in the core. it is dependent on the core data, number of windings and mains voltage usually 70-80% of Isat.
The secondary current is pulsing in phase when the mains is high and can provide higher current without saturation.
But diodes and Caps might be stressed, that is why also in the secondary NTCs are used sometimes.
Here the limiting factor is the heat created by ohmic resistance of the winding.
A saturated transformer creates big spikes in the primary, only ohmic resistance of the winding limits it at up to 10x normal. So a primary resistor can limit it.
The magnetizing current is 90° out of phase, the maximum occurs when the mains is 0V and the energy has to come from the (limited) flux in the core. it is dependent on the core data, number of windings and mains voltage usually 70-80% of Isat.
The secondary current is pulsing in phase when the mains is high and can provide higher current without saturation.
But diodes and Caps might be stressed, that is why also in the secondary NTCs are used sometimes.
Here the limiting factor is the heat created by ohmic resistance of the winding.
Last edited:
Usually audio amplifiers don't have a big enough transformer for the inrush current caused by the iron itself to be a real issue. Even with the transformers in some of the biggest PA amps, it wouldn't cause an issue. A bit of a thud maybe, but nothing to harmful.
With really big transformers I've measured that inrush current to be over 3 kA, but that was a bit of an unusual (and enormous) transformer.
The problem audio amps have is that we have a toroidal transformer with thick secondaries (so it can take big peak currents) and up to 60 mF of capacitance.
With really big transformers I've measured that inrush current to be over 3 kA, but that was a bit of an unusual (and enormous) transformer.
The problem audio amps have is that we have a toroidal transformer with thick secondaries (so it can take big peak currents) and up to 60 mF of capacitance.
Last edited:
Then it will be clipping all the time.
Unless you *only* play sinewaves that is.
I have a Yamaha amp with 900Wrms and it has a 1800VA transformer, so I recommend this relation for PA use.
magnetizing current in the primary is due to number of primary turns, magnetic path length and flux density...
what will make starting current high is when 1: when the primary voltage is much higher than designed for, 2: when secondary load currents are much much more than what the traffo core was designed for...
in transient terms, the huge amounts of filter capacitors in the secondary psu will cause a transient current very very high however momentary....the thing that limits those currents are the dc resistance of the traffo primary and secondary windings...
what will make starting current high is when 1: when the primary voltage is much higher than designed for, 2: when secondary load currents are much much more than what the traffo core was designed for...
in transient terms, the huge amounts of filter capacitors in the secondary psu will cause a transient current very very high however momentary....the thing that limits those currents are the dc resistance of the traffo primary and secondary windings...
The starting currents of the transformers reach values that significantly exceed the operating currents. So, for a toroidal transformer with a rated power of 5 kVA, the inrush current pulse reaches 1000 ... 2000 A. Inrush currents can trigger current protection devices (for example, circuit breakers). The following ways to reduce inrush currents can be proposed.
Connecting the transformer to the mains at the moment when the mains voltage is at its maximum value (that is, at the moment φ = π / 2). This method is the most effective, but it requires the use of special switching devices.
The inclusion of an active resistance (resistor) in series with the primary winding of the transformer. The disadvantage of this method is the heating of such a resistor, as well as the associated decrease in efficiency.
It is much more efficient to use a thermistor - i.e. resistor with negative temperature coefficient of resistance. A thermistor has significantly less resistance when hot than when it is cold, so the heat loss is also significantly less than when using a conventional resistor. So, for example, for use in conjunction with a 3 kVA transformer, a thermistor of the SCK-2R515 type can be recommended, having a cold resistance of 2.5 Ohms and designed for a rated current of 15 A.
Recently, the so-called single-phase inrush current limiters of the ESB and ESBH series for rated currents of 10 A and 16 A have appeared on the market. Their principle of operation is based on the connection in series with the load of a current-limiting resistor (usually 5 Ohm), and this resistor is closed by relay contacts with some delay (adjustable from 20 to 50 ms).
The circuit breakers (circuit breakers) used to connect the transformer to the mains must have tripping characteristics "D" (IEC / IEC 898 standard) and "K" (DIN VDE 0660 standard). Automatic machines with such characteristics are designed specifically for active-inductive loads (electric motors, transformers), characterized by a high multiple of the rated current value (that is, the ratio of the starting current to the rated value). For machines with characteristic "D" the multiplicity is about 15, and for machines with characteristic "K" - about 10. If there are no circuit breakers with the indicated letters at hand, and the transformer needs to be connected urgently, you can also use devices with the letters B, C ( the most common), but then they must be taken with a 2-3-fold current margin. It should be remembered that in this case, the circuit breaker will operate at a current 2-3 times higher than the rated current, that is, the protective function of the breaker will deteriorate.
In any case, the problem of inrush current is the problem of the hardware designer, not the transformer manufacturer, since the transformer manufacturer cannot influence the value of this parameter in any way. Depending on the specific situation, the design engineer decides which method to reduce the starting current to choose. For transformers less than 1 kVA, there is usually no need to deal with inrush currents. It should be noted that in some cases, the relatively large internal resistance of the supply network can reduce the starting current to an acceptable value.
Connecting the transformer to the mains at the moment when the mains voltage is at its maximum value (that is, at the moment φ = π / 2). This method is the most effective, but it requires the use of special switching devices.
The inclusion of an active resistance (resistor) in series with the primary winding of the transformer. The disadvantage of this method is the heating of such a resistor, as well as the associated decrease in efficiency.
It is much more efficient to use a thermistor - i.e. resistor with negative temperature coefficient of resistance. A thermistor has significantly less resistance when hot than when it is cold, so the heat loss is also significantly less than when using a conventional resistor. So, for example, for use in conjunction with a 3 kVA transformer, a thermistor of the SCK-2R515 type can be recommended, having a cold resistance of 2.5 Ohms and designed for a rated current of 15 A.
Recently, the so-called single-phase inrush current limiters of the ESB and ESBH series for rated currents of 10 A and 16 A have appeared on the market. Their principle of operation is based on the connection in series with the load of a current-limiting resistor (usually 5 Ohm), and this resistor is closed by relay contacts with some delay (adjustable from 20 to 50 ms).
The circuit breakers (circuit breakers) used to connect the transformer to the mains must have tripping characteristics "D" (IEC / IEC 898 standard) and "K" (DIN VDE 0660 standard). Automatic machines with such characteristics are designed specifically for active-inductive loads (electric motors, transformers), characterized by a high multiple of the rated current value (that is, the ratio of the starting current to the rated value). For machines with characteristic "D" the multiplicity is about 15, and for machines with characteristic "K" - about 10. If there are no circuit breakers with the indicated letters at hand, and the transformer needs to be connected urgently, you can also use devices with the letters B, C ( the most common), but then they must be taken with a 2-3-fold current margin. It should be remembered that in this case, the circuit breaker will operate at a current 2-3 times higher than the rated current, that is, the protective function of the breaker will deteriorate.
In any case, the problem of inrush current is the problem of the hardware designer, not the transformer manufacturer, since the transformer manufacturer cannot influence the value of this parameter in any way. Depending on the specific situation, the design engineer decides which method to reduce the starting current to choose. For transformers less than 1 kVA, there is usually no need to deal with inrush currents. It should be noted that in some cases, the relatively large internal resistance of the supply network can reduce the starting current to an acceptable value.
Last edited:
The starting current of transformers is completely different, depending on the power of the transformer and its design. And even with an absolutely correctly designed transformer, it can be very significant, it depends on the time of switching on to the network and the residual magnetization of the magnetic circuit due to hiseresis during the previous disconnection.
I suggest a sanity check when reading this thread.
For starters, cakyol started by identifying the use of a 2.5kVA toroid for a (stereo ?) amp with 500Wrms /channel rating. For a diy designed amp, that is a worry for starters, especially as the topic of inrush and how to mitigate it is being managed by starting a forum thread. This is not imho a simple task to design for, or meant for a typical domestic application.
As there is no information on the mains AC connection, as to what upstream protection is in place, and what prospective peak current could be drawn, then trying to handwave a particular type of CB as being required is way too simple for advice.
This is a diy forum, and people will propose using NTC's and relay contacts etc., and make quick estimates as to what should be suitable, but I recommend not following this form of simplistic design approach, especially as the current levels are getting beyond domestic application.
For starters, cakyol started by identifying the use of a 2.5kVA toroid for a (stereo ?) amp with 500Wrms /channel rating. For a diy designed amp, that is a worry for starters, especially as the topic of inrush and how to mitigate it is being managed by starting a forum thread. This is not imho a simple task to design for, or meant for a typical domestic application.
As there is no information on the mains AC connection, as to what upstream protection is in place, and what prospective peak current could be drawn, then trying to handwave a particular type of CB as being required is way too simple for advice.
This is a diy forum, and people will propose using NTC's and relay contacts etc., and make quick estimates as to what should be suitable, but I recommend not following this form of simplistic design approach, especially as the current levels are getting beyond domestic application.
No, unfortunately you are wrong. Starting magnetizing current inrush can be relatively small only if transformer winding is designed to work with a relatively low induction (<1T for a 1.5-1.8 T max rated core material). But this is rare case. So current inrush always is large. It only depends how large it will be. It is easy to check with an AC ammeter capable to show a peak value.
Last edited:
It is so for set mode, but not for transient. Energizing is much more close to mode, for example, if we apply 1 Hz AC voltage instead of 50/60 Hz AC voltage to the primary.
Last edited:
This (stereo) amp has a 2.5 kVA toroidal and 520,000 uF of capacitance. Its lines are at +/- 95 volts DC. It is extremely unlikely it will clip at 70 % capacity. But its power on inrush will be a beast. In any case, I have decided to put a 10 ohm 50 watt chassis mount resistor which will be shorted out by a 25 amp relay after about 3-4 seconds.
If the relay does not engage, I may be in trouble
But that is very rare
If the relay does not engage, I may be in trouble
But that is very rare
Last edited:
At a certain point more filter caps don't really make much difference. The MC2 MC1250 is a much bigger amplifier than you are talking about, and it used about 30,000 uF of capacitance per rail. I forget what the rails were, but I want to say they were something like 115V DC rails. I've never heard complaints about that amp having a "wimpy" power supply, although they do suffer from filter cap failures since they run the caps on the bleeding edge of their voltage rating.
The point I'm making, is you are talking about having almost 10 times the capacitance of any normal amplifier in that power range... just something to think about. It's okay to go a bit overkill, but at a certain point you're just wasting money and making a bigger explosion if something ever fails. With that much capacitance and those voltage rails, it's going to be an explosion to say the least.
The point I'm making, is you are talking about having almost 10 times the capacitance of any normal amplifier in that power range... just something to think about. It's okay to go a bit overkill, but at a certain point you're just wasting money and making a bigger explosion if something ever fails. With that much capacitance and those voltage rails, it's going to be an explosion to say the least.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.