Is there any way to find out the power on surge current of a torodial transformer?
I have a1.5KVA with 23,000uf per rail and sometimes when I power it up the breaker trips.
Also next month I plan to build a power amp that will use two 1.5KVA toroids, will I have to use a soft start circuit?
Thanks in advance
I have a1.5KVA with 23,000uf per rail and sometimes when I power it up the breaker trips.
Also next month I plan to build a power amp that will use two 1.5KVA toroids, will I have to use a soft start circuit?
Thanks in advance
Measure the DC resistance of the primary. Take top voltage 160/310 volt divided by the resistance. This gives the absolute maximum and is in fact not far from the real value! = huge current!
Your problem isn't the 23000 uF, it's the toroid.
You MUST have a soft start and I recoemmend that you have 20-100 ohms (10-50 watts) in series for 1 second and shortened by a relay. It's important that you have fast reset of the circuit otherwise a short glitch on the mains power can make your fuse blow.
Your whole problem is about building up magnetic energy in the toroid. If you switch on the toroid at top voltage yopu will saturate the core and this lasts for a while.
Your problem isn't the 23000 uF, it's the toroid.
You MUST have a soft start and I recoemmend that you have 20-100 ohms (10-50 watts) in series for 1 second and shortened by a relay. It's important that you have fast reset of the circuit otherwise a short glitch on the mains power can make your fuse blow.
Your whole problem is about building up magnetic energy in the toroid. If you switch on the toroid at top voltage yopu will saturate the core and this lasts for a while.
Crazy current
Hey peranders,
I tried your suggestion, the resistance of the primary is 0.2 ohms.
So 160/0.2 = 800 amps that is some serious current.
So each time I turn on the transformer I pull 800 amps from the mains.
Are you sure this is right?
Hey peranders,
I tried your suggestion, the resistance of the primary is 0.2 ohms.
So 160/0.2 = 800 amps that is some serious current.
So each time I turn on the transformer I pull 800 amps from the mains.
Are you sure this is right?
No, not exactly but one time of ten according to my experience. But you will get huge currents statisticly rather often so you will need a soft start anyhow.
I guess your start current will be close the short circuit current in your wiring, not 800 A maybe, but much!
I got 77 A with a 500 VA toroid at 230 volt and no load at all. This was measured with a current probe (hall sensor) and oscilloscope.
The primary resistance was 2.5-3 ohms, don't remember the exact value.
I guess your start current will be close the short circuit current in your wiring, not 800 A maybe, but much!
I got 77 A with a 500 VA toroid at 230 volt and no load at all. This was measured with a current probe (hall sensor) and oscilloscope.
The primary resistance was 2.5-3 ohms, don't remember the exact value.
lawbadman said:
peranders said:
In an article a few years back on soft start circuits in Audio Amateur magazine an American author was making measurements on a 120V toroid with a hall-effect probe and could routinely overrange it. The probe was rated at 100 Amps full scale.
The currents can be quite large at swich on but only for the first 4 or 5 cycles and decrease exponentially.
A soft-start circuit is highly recommended. Several designs have already been discussed in this forum.
James
IIRC, turn on surge is minimum when you turn the amp on right at the moment where the mains is at max voltage (exact opposite to what you would think.... it's got to do with setting up the magnetizing current in the transformer core) so with a bit of thought you could make an electronic switch that turns on power right at the peak of the mains AC wave. That way, the soft start current limitter would have a consistent and relatively easy job. And yes, turn on surge with a large toroidal transformer can commonly be in the range of 100 Amps or more.
> the resistance of the primary is 0.2 ohms. So 160/0.2 = 800 amps that is some serious current. So each time I turn on the transformer I pull 800 amps from the mains. Are you sure this is right?
Roughly right.
You should add the resistance of the wires in the wall behind the wall-outlet, all the way back to the generator. However as a first approximation, there will normally be less than 2% voltage drop at 20 Amps (in US residential work). 2% of 120V is 2.4V, 2.4V/20A= 0.12Ω. If you are close to the fusebox it may be less.
So you have from 0.2 to 0.3Ω, 840A to 560A first-cycle surge current.
> I have a 1.5KVA... will use two 1.5KVA
These are really too big for US residential use with a capacitor input supply. Use smaller (higher resistance) transformers, or rig some soft-start scheme.
Roughly right.
You should add the resistance of the wires in the wall behind the wall-outlet, all the way back to the generator. However as a first approximation, there will normally be less than 2% voltage drop at 20 Amps (in US residential work). 2% of 120V is 2.4V, 2.4V/20A= 0.12Ω. If you are close to the fusebox it may be less.
So you have from 0.2 to 0.3Ω, 840A to 560A first-cycle surge current.
> I have a 1.5KVA... will use two 1.5KVA
These are really too big for US residential use with a capacitor input supply. Use smaller (higher resistance) transformers, or rig some soft-start scheme.
Soft Start
In the current issue (Q3/03) of Texas Instruments "Analog Applications Journal" there is a feature on Soft Start and LDO Regulators. The principals applied here can be ramped for your application:
http://focus.ti.com/lit/an/slyt047a/slyt047a.pdf
In the current issue (Q3/03) of Texas Instruments "Analog Applications Journal" there is a feature on Soft Start and LDO Regulators. The principals applied here can be ramped for your application:
http://focus.ti.com/lit/an/slyt047a/slyt047a.pdf
Things that go "bump" in the night.
What happens during normal running is the AC voltage swings from zero to 339v peak (240vac here) and the core magnetising flux (lagging 90 degrees) goes from *negative peak* to zero *at the same time*. Then as the voltage swings down from 339v to zero the mag flux goes from zero to max. Everybody is happy.
Buuuuut... if you turn on the tranny at the zero voltage crossing point of the AC, as it rises normally toward 339v - the peak voltage - the magnetising flux goes from **zero** (not peak negative value as before) to the safe working maximum at the same time. Then... as the voltage continues from 339v to zero the magnetising flux just keeps on going up and up until the core saturates, and then it goes *way* up.
Takes quite a few cycles to settle down.
My suggestion? Use an NTC thermistor in series with the primary, and short it with a relay after a few seconds. Thermistors are really rugged, 100% safe for this kind of use, and as they heat up they reduce resistance, making life nicer for the relay. Also, if a momentary power interruption makes the relay contacts open, the thermistor will be cold and therefore high resistance, making for a really yummy restart.
That is exactly right, especially if there is nothing connected to the secondary. The dc caps will pull a fair bit but nothing like the tranny.AudioFreak said:
What happens during normal running is the AC voltage swings from zero to 339v peak (240vac here) and the core magnetising flux (lagging 90 degrees) goes from *negative peak* to zero *at the same time*. Then as the voltage swings down from 339v to zero the mag flux goes from zero to max. Everybody is happy.
Buuuuut... if you turn on the tranny at the zero voltage crossing point of the AC, as it rises normally toward 339v - the peak voltage - the magnetising flux goes from **zero** (not peak negative value as before) to the safe working maximum at the same time. Then... as the voltage continues from 339v to zero the magnetising flux just keeps on going up and up until the core saturates, and then it goes *way* up.
Takes quite a few cycles to settle down.My suggestion? Use an NTC thermistor in series with the primary, and short it with a relay after a few seconds. Thermistors are really rugged, 100% safe for this kind of use, and as they heat up they reduce resistance, making life nicer for the relay. Also, if a momentary power interruption makes the relay contacts open, the thermistor will be cold and therefore high resistance, making for a really yummy restart.
As far as I remember the moment when you switch if OFF is also determining the rush in current next time you swich it on.
If you switch OFF at a moment of large flux in a particular direction, the core remains oriented. If next time you switch it on at a point on the mains wave that would require a reverse flux, there is a HUGE current, as the transformer L is non-existent and basically you are looking at a low-resistance piece of wire.
Because this is random, it explains why sometimes the breaker trips, sometimes not.
Circlotron, does that make sense to you?
Jan Didden
If you switch OFF at a moment of large flux in a particular direction, the core remains oriented. If next time you switch it on at a point on the mains wave that would require a reverse flux, there is a HUGE current, as the transformer L is non-existent and basically you are looking at a low-resistance piece of wire.
Because this is random, it explains why sometimes the breaker trips, sometimes not.
Circlotron, does that make sense to you?
Jan Didden
No, no, no!janneman said:
The thing that conrols this is _when_ you switch on in a half period, nothing else. The size of the inrush current is just a random thing.
Peranders: Observe, how can the inrush current be so high? After all, you switch a voltage across a coil, so current gradually builts up > no inrush current. Yet, reality shows us otherwise. An it's not the capacitors, because try just the xformer - it also occasionally blows the fuse.
If there is a better explanation than I gave, I haven't seen it yet. You?
Jan Didden
If there is a better explanation than I gave, I haven't seen it yet. You?
Jan Didden
OK, here's the deal: The inrush current is max, indeed, if you switch the mains on at the zero crossing. BUT, if there is residual flux (from the last time you switched off) the inrush will even be greater. Because the excellent magnetic properties of a toroid, they have very high magnetic remanence and thus stay magnetically oriented from the time of switch off.
You can (if you are courageous enough) even calculate it...
As is seen, at the first zero-crossing of the voltage (red) the flux (green) is maximum.
Check the text below fig 1.
Jan Didden
You can (if you are courageous enough) even calculate it...
As is seen, at the first zero-crossing of the voltage (red) the flux (green) is maximum.
Check the text below fig 1.
Jan Didden
Attachments
transformer surge
Far too many ciruit preambles on an old industry problem!
I designed a circuit some 20 yrs ago for use in the music industry and for welding equipment which works a treat on any power transformer with sensitive windings and high CV product. The point everyone misses is that the fusing of toroids and power transformers becomes a far more reliable proposition to actual circuit RMS load. It will reset on a serious "brown out" or sudden interruption.
Using a relay, thyristor by switching out a power resistor (the cheapest method) the dropout speed is fast enough to avoid the danger of the flux reversal collision. Power up can be in seconds and will reset as fast as a duff plug. original circuit in in microsoft paintplus program.
As circuit is older than 20 years and already in published equipment it cannot be patented
Interested users be patient as I have to dig it out of old redundant 1.44 disks. Any interest ?
Far too many ciruit preambles on an old industry problem!
I designed a circuit some 20 yrs ago for use in the music industry and for welding equipment which works a treat on any power transformer with sensitive windings and high CV product. The point everyone misses is that the fusing of toroids and power transformers becomes a far more reliable proposition to actual circuit RMS load. It will reset on a serious "brown out" or sudden interruption.
Using a relay, thyristor by switching out a power resistor (the cheapest method) the dropout speed is fast enough to avoid the danger of the flux reversal collision. Power up can be in seconds and will reset as fast as a duff plug. original circuit in in microsoft paintplus program.
As circuit is older than 20 years and already in published equipment it cannot be patented
Interested users be patient as I have to dig it out of old redundant 1.44 disks. Any interest ?
I see what you mean. I expect you could prevent this by open circuiting the transformer at switchoff by using a triac. It will turn off "naturally" at the AC *current* (and therefore magnetising flux) zero-crossing), which is just a little before the next peak of the voltage waveform, which would mean the core has completely reset from the previous magnetic polarity and has not started to go the other way. Also,janneman said:
you would get a *noiseless* power down, no random =crack= as you shut the power off. 
That feature alone might be worth playing with?BTW, you would also want to have a conventional switch as well, to make sure it is 100% certain shut off for electrical safety reasons.
P.S. Why are simple things so darned complicated???
The thing is that you will saturate the core if you are unlucky (random thing) and to only thing which limits the current is the R since the L is gone. The toroids have very little leakage inductance, good and bad. Compare EI-cores.janneman said:
Circlotron said:
I see what you mean. I expect you could prevent this by open circuiting the transformer at switchoff by using a triac. It will turn off "naturally" at the AC *current* (and therefore magnetising flux) zero-crossing), which is just a little before the next peak of the voltage waveform, which would mean the core has completely reset from the previous magnetic polarity and has not started to go the other way. Also,
BTW, you would also want to have a conventional switch as well, to make sure it is 100% certain shut off for electrical safety reasons.
P.S. Why are simple things so darned complicated???
Correct. Ideal shut-off point, to avoid chance of too high inrush current next time you switch on, would be a current zero crossing. Which is the case 90% of the time anyway. That 90% of the time the xformer is effectively disconnected from the amp. Except for that 1mS or so out of every 10 that the diodes conduct. But that's another story.
PS: Because there is always another skin after you peel the skin from an onion....
Jan Didden
PPS: Here the full link to the article. Note the influence of the residual flux:
http://www.qte.com/Main Pages/Technical Papers Page/Inrush Page/qte_tech_papers_inrush.htm#
janneman said:
If you don't believe me: Take a transformer WITHOUT anything connected. Then test the inrush current! It's the transformer itself! A huge smoothing battery can make the inrush current longer but the max current is the transformer itself responsible for.
peranders said:
If you don't believe me: Take a transformer WITHOUT anything connected. Then test the inrush current! It's the transformer itself! A huge smoothing battery can make the inrush current longer but the max current is the transformer itself responsible for.
Why do you repeat what I said (post 13)?
Did you read the link I provided in post 18? What's your opinion on that?
Jan Didden
- Status
- Not open for further replies.