don't try to combine Soft Start for transformer startup and Slow Charge for the capacitance after the rectifier.
Soft Start is fitted in the Primary circuit of the transformer. delay typically <<500milli-seconds.
Slow Charge is fitted in the Secondary circuit of the transformer. delay typically >>1second.
Soft Start is fitted in the Primary circuit of the transformer. delay typically <<500milli-seconds.
Slow Charge is fitted in the Secondary circuit of the transformer. delay typically >>1second.
Last edited:
Hi Dazed2,
You shouldn't need anything but the softstart which is nothing more than a in-rush current limiting Thermistor. If you are blowing the fuse then your thermistor isn't the correct value. For the power supply I am currently working on the mains fuse is 10 amps so with 115v mains I can get away with 10 ohms for the thermistor. To make sure it does heat up too fast I use 4 thermistors, 2 in parallel, in series with another parallel pair. That keeps the initial resistance at 10 ohms and splits the load. Then I have a NE555 timer pull in a relay that bypasses the thermistors so there is no current limiting and the thermistor aren't working continuously. The relay is capable of 30 amps on the mains and the relay coil can be pulled in with 12 volts and a few milliamps of current the NE555 can sink.
Have a good one,
John
You shouldn't need anything but the softstart which is nothing more than a in-rush current limiting Thermistor. If you are blowing the fuse then your thermistor isn't the correct value. For the power supply I am currently working on the mains fuse is 10 amps so with 115v mains I can get away with 10 ohms for the thermistor. To make sure it does heat up too fast I use 4 thermistors, 2 in parallel, in series with another parallel pair. That keeps the initial resistance at 10 ohms and splits the load. Then I have a NE555 timer pull in a relay that bypasses the thermistors so there is no current limiting and the thermistor aren't working continuously. The relay is capable of 30 amps on the mains and the relay coil can be pulled in with 12 volts and a few milliamps of current the NE555 can sink.
Have a good one,
John
I agree with Zen, don't parallel NTC Thermistors.
A 10A fuse on 115Vac mains allows for 1150VA of load. That is high, is your load really that high?
A 10A fuse on 115Vac mains allows for 1150VA of load. That is high, is your load really that high?
As AndrewT and Zen said before and on other threads as well.
Soft starting a transformer and charging a cap bank is very different. The two are not the same.
If you try and slow start a transformer with a large cap bank behind it, it won't work.
The two masters here correct me if I'm wrong but slow starting the trafo will mean the trafo will be restricted on both voltage and current and thus never get the caps charged enough that when the relay switches over inrush will still be very high. The purpose of the soft start is just to start a large trafo. Then when the trafo is at full voltage the cap bank can slowly charge up via a slow charge circuit.
I tried what you were thinking too..... it doesn't work! I was trying to be lazy too.... I had 3 10ohm 10A thermistors in series in a soft start.... it blows a 10A fuse when it switches over, if it doesn't blow the thermistors first. I had the timer anywhere from 0.5 sec to 10secs.
With that much capacitance, you must have BOTH a soft start to start the transformer AND a slow charge to charge the cap bank. ESPECIALLY the amp you build is a class A amp. ie constant current draw at startup. Listen to AndrewT
Soft starting a transformer and charging a cap bank is very different. The two are not the same.
If you try and slow start a transformer with a large cap bank behind it, it won't work.
The two masters here correct me if I'm wrong but slow starting the trafo will mean the trafo will be restricted on both voltage and current and thus never get the caps charged enough that when the relay switches over inrush will still be very high. The purpose of the soft start is just to start a large trafo. Then when the trafo is at full voltage the cap bank can slowly charge up via a slow charge circuit.
I tried what you were thinking too..... it doesn't work! I was trying to be lazy too.... I had 3 10ohm 10A thermistors in series in a soft start.... it blows a 10A fuse when it switches over, if it doesn't blow the thermistors first. I had the timer anywhere from 0.5 sec to 10secs.
With that much capacitance, you must have BOTH a soft start to start the transformer AND a slow charge to charge the cap bank. ESPECIALLY the amp you build is a class A amp. ie constant current draw at startup. Listen to AndrewT
Last edited:
examples :
1. 800VA donut , 220Vac mains , 15R NTC , four 18Vac secondaries (2 used per F4 channel ) , one cap bank per channel having 4 x 33mF , 3A16 slow or 4A fast fuse
2. M2 - 2x300VA donuts , common 15R NTC for both , 220Vac mains , each donut feeding 4 x 33mF , common fuse 3A16 slow
no need for any slow start
in case that I choose relay based soft start , I would use at least 2x10R NTC in series ( up to 4x15R) and set at least 3sec relay time
1. 800VA donut , 220Vac mains , 15R NTC , four 18Vac secondaries (2 used per F4 channel ) , one cap bank per channel having 4 x 33mF , 3A16 slow or 4A fast fuse
2. M2 - 2x300VA donuts , common 15R NTC for both , 220Vac mains , each donut feeding 4 x 33mF , common fuse 3A16 slow
no need for any slow start
in case that I choose relay based soft start , I would use at least 2x10R NTC in series ( up to 4x15R) and set at least 3sec relay time
I have 40Vac secondaries, and 282mF per rail, or 564mF total. The amp is biased at 2.8A or 5.6A at 52Vdc.
You try and do that with just a soft start! 😀
johnhenryharris is looking to run 48Vac secondaries and 544mF in caps and 2.4kVa trafo... good luck trying to start that up with just a soft start.
I was using a 1kva trans with .120 farads with just a softstart. The SA/1 idle bias is 5 amps total and 3 to 3.5 more amps at full power, or about 835 va., about 7 amps from the mains.
Paralleling NTCs has been done without problems, 1 NTC with greater amperage rating might be better but is harder to get and more expensive.
Dazed2 has the new part specs correct. BTW there is a 3 millihenry toriod inductor between the first two capacitors and the next 2 capacitors of each bank.
Once I receive the caps and transformers I will test it extensively and see how things hold up. Hopefully no smoke.
The older parts will shift over to the new
Paralleling NTCs has been done without problems, 1 NTC with greater amperage rating might be better but is harder to get and more expensive.
Dazed2 has the new part specs correct. BTW there is a 3 millihenry toriod inductor between the first two capacitors and the next 2 capacitors of each bank.
Once I receive the caps and transformers I will test it extensively and see how things hold up. Hopefully no smoke.
The older parts will shift over to the new
another way to put it.
W =½ CV^2
Zen 's power supply is 23.5Vdc with 264mF caps. About 70J of energy to charge
at 52Vdc with 564mFcaps were talking 762J of energy storage! 10x more
johnhenryharris you have an even more insane PSU
48Vac or about 62Vdc with 544mF caps equals to 1046J !
And all that is before having a constant current draw that is the amp.
A soft start will have a lot of trouble charging that!
W =½ CV^2
Zen 's power supply is 23.5Vdc with 264mF caps. About 70J of energy to charge
at 52Vdc with 564mFcaps were talking 762J of energy storage! 10x more
johnhenryharris you have an even more insane PSU
48Vac or about 62Vdc with 544mF caps equals to 1046J !
And all that is before having a constant current draw that is the amp.
A soft start will have a lot of trouble charging that!
dazed2 wrote: Soft starting a transformer and charging a cap bank is very different. The two are not the same.
If you try and slow start a transformer with a large cap bank behind it, it won't work.
This information was interesting and new for me. I have used slow start for many years without any problems probably due to the fact that condenser bank have not been big enough, as I understand.
Have any practical idea to a slow start for the condenser bank been brought? I hope for what componets to use and a drawing how to incorprate this in the amplifiers power unit.
Eivind Stillingen
If you try and slow start a transformer with a large cap bank behind it, it won't work.
This information was interesting and new for me. I have used slow start for many years without any problems probably due to the fact that condenser bank have not been big enough, as I understand.
Have any practical idea to a slow start for the condenser bank been brought? I hope for what componets to use and a drawing how to incorprate this in the amplifiers power unit.
Eivind Stillingen
Last edited:
Hi Eivind,
I will draw out the softstart Schematic, annotate it with part references, that I have used and scan it and see if I can post it here. Probably get to it tomorrow.
John
I will draw out the softstart Schematic, annotate it with part references, that I have used and scan it and see if I can post it here. Probably get to it tomorrow.
John
Johnhenryharris
Thanks, I think that many other members are looking forward to take a look sølution.
Eivind Stillingen
Thanks, I think that many other members are looking forward to take a look sølution.
Eivind Stillingen
Hi Dazed2,
Just noticed on your post, I have 48vdc at the caps or 36vac prior to rectification.
So not quite so insane but pretty hefty, about like the Pass Aleph 1.2.
John
Just noticed on your post, I have 48vdc at the caps or 36vac prior to rectification.
So not quite so insane but pretty hefty, about like the Pass Aleph 1.2.
John
dunno - I'm seeing that as xformer is source for cap bank
wherever you put a pipe on that chain , it will choke it
it's certainly completely different issue how amp circuit is behaving with achieved voltage rise ....
wherever you put a pipe on that chain , it will choke it
it's certainly completely different issue how amp circuit is behaving with achieved voltage rise ....
dunno - I'm seeing that as xformer is source for cap bank
wherever you put a pipe on that chain , it will choke it
it's certainly completely different issue how amp circuit is behaving with achieved voltage rise ....
This is where I will defer to AndrewT explain it. 😕
From what I understand, if you choke the transformer and it doesn't get to its full operating voltage it may not achieve the necessary magnetic flux, then when the relay switches over you still get a pretty big inrush current. So you then have to deal with both the transformer inrush plus the incomplete charge of the caps.
Again, I may be wrong here, but I did try it with just a soft start, and it almost always blows a 10A fuse on a 120V mains! sometimes it even blows a series of 4 inrush thermistors!
Oh sorry, I thought it was AC voltage.
Anyways at 48vdc that's still 626J ! thats a very large amount!
The 15ohm NTC in both examples IS THE SOFT START.
Slow Charge is a different issue.
is confusing and ambiguous.
What do you mean?
Last edited:
yup , NTC in primary is soft start
what I meant to say - no need for any other slow start , in secondary circuit
if NTC soft start is supplemented with bypass mechanismus (time delayed relay) , all one needs is to determine exact time needed to cap bank reach operating voltage - speaking of case of enormous cap bank size
sure , one also need to fiddle with exact NTC chain value
possibilities are numerous (stepped starting etc. ) but there is no need to brake that function to two different places - to have it on both primary and secondary side
what I meant to say - no need for any other slow start , in secondary circuit
if NTC soft start is supplemented with bypass mechanismus (time delayed relay) , all one needs is to determine exact time needed to cap bank reach operating voltage - speaking of case of enormous cap bank size
sure , one also need to fiddle with exact NTC chain value
possibilities are numerous (stepped starting etc. ) but there is no need to brake that function to two different places - to have it on both primary and secondary side
Last edited:
Slow charging of capacitors is a recognised method of extending their life and maintaining their specified parameters.
Fast charging of capacitors can damage the capacitors and damage the circuit feeding the capacitors.
A bank of ten 15mF 63V capacitors on each pole of a dual polarity supply puts a lot of instantaneous load on the preceding bridge rectifier. This 150mF could be the supply to a very big ClassA amplifier, or to feed some industrial equipment.
The lower the esr of the capacitor the more one needs a slow charge circuit.
The higher the charge voltage the more one needs a slow charge circuit.
Take a +-58Vdc from the bridge rectifier feeding that +-150mF cap bank.
Assume the cable resistances are 10milli-ohms, the rectifier dynamic resistance is 20milli-ohms, the secondary resistance is 50milli-ohms and the esr of each capacitor is 18milli-ohms.
The 58Vdc sees a total resistance of 10+20+50+1.8 =81.8mill-ohms.
The instantaneous charging current will be ~58/0.0808 >700Apk
Tell us again that slow charge is not required.
NTC manufacturers describe how to use their NTC to slow charge that cap bank.
Remove 18 of those capacitors, leaving just +-15mF to be charged.
Do the arithmetic. Include the effect of primary resistance and the effect of Mains Source resistance.
Fast charging of capacitors can damage the capacitors and damage the circuit feeding the capacitors.
A bank of ten 15mF 63V capacitors on each pole of a dual polarity supply puts a lot of instantaneous load on the preceding bridge rectifier. This 150mF could be the supply to a very big ClassA amplifier, or to feed some industrial equipment.
The lower the esr of the capacitor the more one needs a slow charge circuit.
The higher the charge voltage the more one needs a slow charge circuit.
Take a +-58Vdc from the bridge rectifier feeding that +-150mF cap bank.
Assume the cable resistances are 10milli-ohms, the rectifier dynamic resistance is 20milli-ohms, the secondary resistance is 50milli-ohms and the esr of each capacitor is 18milli-ohms.
The 58Vdc sees a total resistance of 10+20+50+1.8 =81.8mill-ohms.
The instantaneous charging current will be ~58/0.0808 >700Apk
Tell us again that slow charge is not required.
NTC manufacturers describe how to use their NTC to slow charge that cap bank.
Remove 18 of those capacitors, leaving just +-15mF to be charged.
Do the arithmetic. Include the effect of primary resistance and the effect of Mains Source resistance.
Last edited:
Hi All,
Attached is the softstart circuit I use to start my big amps. The CL40s are 5 ohms and 6 amps. In parallel and series it is still 5 ohms and capable of more current. You could eliminate the 12v transformer and the NE555 circuit and just use a switch to pull in the relays after a slow count of 5 or 10 seconds. I found another softstart supplier on Ebay in Hong Kong or Shenzeng that has every thing on one small PCB, I purchase one but haven't tested it yet. The parts alone was worth the price.
John
Attached is the softstart circuit I use to start my big amps. The CL40s are 5 ohms and 6 amps. In parallel and series it is still 5 ohms and capable of more current. You could eliminate the 12v transformer and the NE555 circuit and just use a switch to pull in the relays after a slow count of 5 or 10 seconds. I found another softstart supplier on Ebay in Hong Kong or Shenzeng that has every thing on one small PCB, I purchase one but haven't tested it yet. The parts alone was worth the price.
John
