I'm building a LM4870 audiosector amp.
I've got the amplifier itself up and running with a temporary power supply. My problem is that the power supply I've got consist of a 500VA transformer (dual 20v secondaries) rectified seperately. With two 10000uF caps on each rail...
The inrush current is killing me.
I've got an Idea on limiting the current with a RC time delay activating relays to bypass current limiting resistors.
I'm just wondering if anyone knows of a similar or better design to look at/compare.
I must go now unfortunately but will check back tomorrow to see if anyone could help me.
Thank you to anyone who could.
PS. I have been searching for something but could not find the right circuit.
I've got the amplifier itself up and running with a temporary power supply. My problem is that the power supply I've got consist of a 500VA transformer (dual 20v secondaries) rectified seperately. With two 10000uF caps on each rail...
The inrush current is killing me.
I've got an Idea on limiting the current with a RC time delay activating relays to bypass current limiting resistors.
I'm just wondering if anyone knows of a similar or better design to look at/compare.
I must go now unfortunately but will check back tomorrow to see if anyone could help me.
Thank you to anyone who could.
PS. I have been searching for something but could not find the right circuit.
Hi,
I had a similar problem with a 1KVa transformer.
Relays will work switching in resistors OR..
you can use 'inrush limiters' which are beefy NTC thermistors, when they heat up they lower in resistance. Therefore your transformer can only draw full current once theyre hot.
They are cheap and really simple. Make sure they can handle your current requirements (you can parallel them up) and put them in a sensible place (they do heat up a fair bit).
I know RS have them under 'inrush limiters' because thats where i got mine.
Craig
I had a similar problem with a 1KVa transformer.
Relays will work switching in resistors OR..
you can use 'inrush limiters' which are beefy NTC thermistors, when they heat up they lower in resistance. Therefore your transformer can only draw full current once theyre hot.
They are cheap and really simple. Make sure they can handle your current requirements (you can parallel them up) and put them in a sensible place (they do heat up a fair bit).
I know RS have them under 'inrush limiters' because thats where i got mine.
Craig
I developed a soft start circuit for an amp I was working on.
The schematic is here; http://www.diyaudio.com/forums/showthread.php?postid=784687#post784687 - post 224.
The PCB layout is here; http://www.diyaudio.com/forums/showthread.php?postid=783781#post783781 - posts 217 & 218.
I like relay contacts because they offer the ultimate in on resistance.
Use it if you want to.
Cheers
The schematic is here; http://www.diyaudio.com/forums/showthread.php?postid=784687#post784687 - post 224.
The PCB layout is here; http://www.diyaudio.com/forums/showthread.php?postid=783781#post783781 - posts 217 & 218.
I like relay contacts because they offer the ultimate in on resistance.
Use it if you want to.
Cheers
I'm using 20mm fuses, slo-blow on mains input
I've seen the circuit on ESP but want something simple.
And I don't know which NTCs to choose, even after looking at some datasheets.
I've seen the circuit on ESP but want something simple.
And I don't know which NTCs to choose, even after looking at some datasheets.
Aswell. Ntcs work by warming up. If you put the amp off for a couple of seconds, the caps discharge, but their still hot and thus not going to stop the inrush.
Relays will work everytime.
If nobody has one I'll work on my design and try to post it.
Relays will work everytime.
If nobody has one I'll work on my design and try to post it.
Here we have a typical trade off:
1 Inrush current
2 We must have fuses and they should also protect, meaning they should be chosen after the load.
The question is how you solve it. 😕
There are a couple of threads, one of them is here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=55799&highlight=
1 Inrush current
2 We must have fuses and they should also protect, meaning they should be chosen after the load.
The question is how you solve it. 😕
There are a couple of threads, one of them is here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=55799&highlight=
Wynand said:
http://home.swipnet.se/~w-50719/hifi/sst01
http://www.hypex.nl/docs/softstart.pdf
http://sound.westhost.com/project39.htm
http://lcaudio.com/index.php?page=12 in the middel of the page
For those who are facing the problem of inrush current, here is a link to the solution http://www.diyaudio.com/forums/showthread.php?s=&threadid=6654&highlight= which in addition protects speaker thumps, detects DC to cut off speakers and has thermal protection if you wish to add.
Cheers,
Cheers,
And the simplest relay soft start. The relay "waits" until sufficient charge has built across the main power supply capacitors. It then closes shorting out the primary series resistor.
This circuit must be used in conjuction with a speaker delay circuit that closes after the soft start relay. Otherwise the amp may be drawing too much current at turn-on. If this happens the soft start relay will never close and the primary connected resistor burns out. This could be also be viewed as cheap insurance should something be wrong with an amp module.
Cheers
This circuit must be used in conjuction with a speaker delay circuit that closes after the soft start relay. Otherwise the amp may be drawing too much current at turn-on. If this happens the soft start relay will never close and the primary connected resistor burns out. This could be also be viewed as cheap insurance should something be wrong with an amp module.
Cheers
Attachments
The trouble with NTC thermistors is that they don't get hot enough to reach a low resistance unless your amp is pulling a significant amount of current.
They don't work on class A/B or class B amplifiers. If you've got a class A amp with a high quiescent current , then NTC thermistors are the cheap and easy option.
Edit:
As stated earlier though, if you turn your amp off and then straight back on again, the NTC will still be hot and you lose the inrush limit. You'd have to wait at least 5 minutes for the NTC to cool before switching back on. A relay is still the best option all round.
For class A/B and class B amps you really need a relay switched power resistor. Here's a nice simple circuit that does the trick:
At power on, the 7.5 ohm 10W resistor is in the transformer primary and limits the maximum current that can flow. As soon as the filter caps start to charge, the 1000uF cap across the relay starts to charge via the 1.1k resistor. Once the relay sees enough voltage ~19V, it pulls in, switching the resistor out of circuit. The whole process takes just about a second, which is plenty enough time for everything to settle. My NAP135 clones will blow 3 Amp mains fuses at every switch on without these. I've also added a second relay using a similar circuit fed from the other supply rail to the speaker outputs to get round the switch on thump. My amps don't have the Naim protection circuits, but the amps are well protected by the regulators anyway. I used a 2200uF cap here to provide a longer delay.
If your supply rails are at a different voltage, then you'll need to use a different value for the 1.1k resistor (the relay needs roughly 40 mA) and then recalculate the cap value to maintain the time constant.
They don't work on class A/B or class B amplifiers. If you've got a class A amp with a high quiescent current , then NTC thermistors are the cheap and easy option.
Edit:
As stated earlier though, if you turn your amp off and then straight back on again, the NTC will still be hot and you lose the inrush limit. You'd have to wait at least 5 minutes for the NTC to cool before switching back on. A relay is still the best option all round.
For class A/B and class B amps you really need a relay switched power resistor. Here's a nice simple circuit that does the trick:
An externally hosted image should be here but it was not working when we last tested it.
At power on, the 7.5 ohm 10W resistor is in the transformer primary and limits the maximum current that can flow. As soon as the filter caps start to charge, the 1000uF cap across the relay starts to charge via the 1.1k resistor. Once the relay sees enough voltage ~19V, it pulls in, switching the resistor out of circuit. The whole process takes just about a second, which is plenty enough time for everything to settle. My NAP135 clones will blow 3 Amp mains fuses at every switch on without these. I've also added a second relay using a similar circuit fed from the other supply rail to the speaker outputs to get round the switch on thump. My amps don't have the Naim protection circuits, but the amps are well protected by the regulators anyway. I used a 2200uF cap here to provide a longer delay.
If your supply rails are at a different voltage, then you'll need to use a different value for the 1.1k resistor (the relay needs roughly 40 mA) and then recalculate the cap value to maintain the time constant.
Hi,
there is some advantage in running the soft start limiter as a series pair of fixed resistor and NTC thermistor.
The resistor provides the restart protection but at a reduced level (hopefully the caps will still hold most of their charge).
The thermistor provides a variable effect, maximum at switch on keeping the start current low then slowly heating and allowing current to be maintained as the transformer impedance rises.
After a time delay the limiting pair are shorted out by relay.
The relay can be purely on a time out basis or as Quasi & Gfinlayson said on voltage detection at PSU.
If the voltage detection circuit is adopted then the relay and it's feed resistor have a tendency to run hot due to excess voltage across the pair. You have to choose the series R to allow switch on at a low voltage limit but at full voltage and/or mains supply maximum the current rises above the switch on needs. A circuit to limit current through these would be better.
there is some advantage in running the soft start limiter as a series pair of fixed resistor and NTC thermistor.
The resistor provides the restart protection but at a reduced level (hopefully the caps will still hold most of their charge).
The thermistor provides a variable effect, maximum at switch on keeping the start current low then slowly heating and allowing current to be maintained as the transformer impedance rises.
After a time delay the limiting pair are shorted out by relay.
The relay can be purely on a time out basis or as Quasi & Gfinlayson said on voltage detection at PSU.
If the voltage detection circuit is adopted then the relay and it's feed resistor have a tendency to run hot due to excess voltage across the pair. You have to choose the series R to allow switch on at a low voltage limit but at full voltage and/or mains supply maximum the current rises above the switch on needs. A circuit to limit current through these would be better.
This is the design I've made so far (before coming to diyaudio).
The advantage is that you don't need another x-former or windings. The reason that theres two relays is because I had two 12v coil 6A relays with one set of contacts only.
The disadvantage.... please tell me.
I'm far from finishing the design so any comments are MOST welcome.
The advantage is that you don't need another x-former or windings. The reason that theres two relays is because I had two 12v coil 6A relays with one set of contacts only.
The disadvantage.... please tell me.
I'm far from finishing the design so any comments are MOST welcome.
Hi Wynand
Your circuit seems to be intended to be connected to the transformers secondary winding, and seems to limit the charge currents to the capacitors.
All soft start circuits that I have seen limit the current flow to the transformer primary i.e, the mains side. This is done to limit both the inrush of the transformer as well as the DC supply. Transformers can have huge inrush currents (particularly toroids) because they are effectively a short circuit until the magnetic field is established.
Still you are limiting some of the inrush and that could be enough to stop your fuses blowing.
Cheers
Your circuit seems to be intended to be connected to the transformers secondary winding, and seems to limit the charge currents to the capacitors.
All soft start circuits that I have seen limit the current flow to the transformer primary i.e, the mains side. This is done to limit both the inrush of the transformer as well as the DC supply. Transformers can have huge inrush currents (particularly toroids) because they are effectively a short circuit until the magnetic field is established.
Still you are limiting some of the inrush and that could be enough to stop your fuses blowing.
Cheers
Hi Wy,
that is difficult to read!
the limiting resistors are in the secondary feed to the caps.
This reduces the peak current charging the caps, there is nothing wrong with this philosophy, it is shown in the data sheets for NTCs.
However it does NOTHING to help with current surge in a large toroid when first switching on.
To achieve soft start to the toroid you need the resistor limiters in the primary of the transformer.
BTW the 50r resistors in the secondary circuit are far too big. All five resistors could be put in parallel giving 2r and limiting the charging current to the caps. Or four Rs in series // giving 10r for a lower charge current.
How long is your relay RC time constant? What will the running current be through the relay coil? Will it run hot?
If you move your limiter to the primary side then 50r is about right for 240Vac. 110Vac needs a lower value, try about 20r to 30r.
Edit,
20mF is nothing special. That is not the problem, If you had >56mF then charging will be more of a problem.
Quasi,
you type quicker than me, but can manage with fewer words.
that is difficult to read!
the limiting resistors are in the secondary feed to the caps.
This reduces the peak current charging the caps, there is nothing wrong with this philosophy, it is shown in the data sheets for NTCs.
However it does NOTHING to help with current surge in a large toroid when first switching on.
To achieve soft start to the toroid you need the resistor limiters in the primary of the transformer.
BTW the 50r resistors in the secondary circuit are far too big. All five resistors could be put in parallel giving 2r and limiting the charging current to the caps. Or four Rs in series // giving 10r for a lower charge current.
How long is your relay RC time constant? What will the running current be through the relay coil? Will it run hot?
If you move your limiter to the primary side then 50r is about right for 240Vac. 110Vac needs a lower value, try about 20r to 30r.
Edit,
20mF is nothing special. That is not the problem, If you had >56mF then charging will be more of a problem.
Quasi,
you type quicker than me, but can manage with fewer words.
- Status
- Not open for further replies.