inrush current

Possibility: original fuse was 1A slow-blow. It died because of "normal" fatigue. Or of some temporary overload that cleared.

Then it was replaced with a 1A fast-blow. This seems consistent with your description of the problem (especially as you leave out the very important facts of slow-blow vs fast-blow).

Slow-blow fuses for capacitor-input DC filters are the norm, not the exception.

Look carefully with the factory fuse or if not available the markings saying what sort of fuse is supposed to go there.

Also, even though the average primary current may be 20mA, what actually matters to the fuse is something closer to the RMS current, and the current waveform is often wildly non-sinusoidal. A factor of 50 to 100 between average and RMS seems implausable, though with no load (capacitor charging only on the very tip of waveform) it may actually be the ratio.

Tim.

Apart from the slow/fast blow advice given elsewhere, check that the mounting bolt etc for the toroidal isn't causing a 'shorted turn' by touching the case on both sides.....if the case is metal then any bolt through the middle of the toroid must only connect with the case on one side.

Phil.

Seems very unlikely inrush current increased. I use torroids in my amps and find they do use lots of inrush current (particularly with a stiff power supply) I fuse them with slow blow fuses and had to install two power switches - one for each pair of amps, or my clock and modem reset themselves when I turn on the stereo.

Torroids have better regulation as a rule - so the current required to charge the supply caps is reflected in the inrush current to a greater degree than with E-I core transformers.

Does it require a slow blow fuse?

Sounds like you already eliminated the filter caps as a direct source of the problem?

If you're using the correct fuse type and it still blows, and its been working all along prior to this, check the amp. Ideally, look at the current with a scope.. You may have a problem in the bias supply - coming up unbalanced or very high current due to a bad cap.

A scope on the output terminals will show it coming on in an unbalanced condition (and a loud speaker thump) - but depending on the design, that could be normal.

--

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+
Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

I have been thinking along these lines. But because we tried slow blows, the only thing I can imagine is that the original fuse is extra slow blow, but I don't know if those things even exist. I'll ask him if the transformer has any info written on it about special kinds of fuses.

Perhaps, but there were no filtercaps connected. Going with the "extra slow blow" theory, perhaps this transformer has some kind of very deformed current waveform, which the original fuse was designed for.

He doesn't have a scope, but I do. If we can't find the problem, we could bring it here, and put it on my scope. But I don't think I'm gonna measure the primary current with it, because I don't know enough about putting scopes on the mains.

We/he tried with the tranny in the free. It was just lying there on the table. And, the metal plate to hold it in place in the (wooden) case, has a rubber sealing of course, so there cannot be any shorts.

I have contructed a four channel power amp with a 600 VA torodial myself. I use two special NTC's in series with the primary coil to limit inrush current. Is that no solution for you?

As I said, the fuse also blows with nothing connected to the transformer. It's tjust the tranny itself which is causing it to blow, not even filter caps.

I thought speaker thumps were the result of the amp's inability to cope with low voltages. A trick in poweramp design is employing a constant current source to avoid that, if I'm not mistaken.

That would work, but sort of defeats the purpose of a super stiff supply. I have four 100 watt amps with banks of computer grade caps feeding photo flash low ESR caps in the amps. Four transformers, one per channel.

I could use some NTC thermistors with time delay relays to short them after the supplies charged - just never got around to it. The two switch technique works well enough.

It would be an excuse to add some more pilot lights and if I used a DPDT relays, I could shunt the caps to pairs of light bulbs to show the caps charge and discharge . . . "but that way insanity lies," and it wouldn't do anything for the sound.

I still can't conceive of a way that a transformer can suddenly develop a problem with inrush current without increasing the excitation current. 20 milliamps sounds high for a torroid, but still reasonable. Do you have another torroid to compare it to? Leave it powered up and check the heat? Shorted turn perhaps?

I have a small 120 VA supply with a torroidal transformer on my workbench. The inrush current is well over an amp and excitation current, with just the filter caps is ~ point four milliamps.

My supply has a mil spec 1 amp magnetic circuit breaker on the 120 input and it will pop that about half the time from a cold start - it has a variac input so if I turn it up from zero the circuit breaker holds. I'm sure a 1 amp slow blow would work since this magnetic breaker is very fast and holds only to 1.2 amps. 20 volt no load output with a 10,000 mfd cap.

Constant current in the diff amp helps if that's the cause, but bias supplies and mismatched output transistors are another cause. With a one amp fuse in a subwoofer system I figure we're talking about a relatively inexpensive amp in something like a computer system.

But you eliminated the amp already so that only leaves the transformer. Right?

--

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+
Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

There is in fact a commercially available chip that switches out an inrush limiting resistor/NTC after a delay.

Sounds like a shorted woofer coil. Sometimes they won't show a short until you 'push' on the cone, or until it moves when you apply signal to it. Scott

I rather think not, as if you read the whole thread, you will see that the OP has carefully stated, on more than one occasion, that the fuse is blowing with absolutely nothing connected, except the transformer primary ...

Arfa

Depending upon the manufacturing method and materials used the inrush current can be as much as 50 times the normal primary current for 1/2 cycle averaging 15 times for about 2 cycles.

See this paper

These NTC's get a very low resistance once they are warm. In believe it was in the order of 0.02 Ohms or something. The only thing I don't understand is, that they don't feel warm, even when the amp has been on for a few minutes. And, when the entire casing of the amp has warmed up, therefore the NTC's as well, and I turn it off and back on immediatly, there is nothing that indicates a high inrush current. Even my bench power supply has a higher inrush current, because sometimes one of my computer resets when I turn it on (and sometimes even when I turn it off...)

This reasoning has me doubting a bit if the resistance really does get very low. Also because perhaps in most devices, a 600VA transformer normally has a higher minimal/standy current than what I'm using it for (the amp in idle only pulls about 25 watts of the mains).

I'll see if I can measure the resitance when they're cold, and when the amp has been warmed up.

Until someone who doesn't know how to operate it comes along...

I don't have a torroidal to test with, but I'll ask if he does. I'll ask if he can measure the idle current, and check for heat on the "broken" one.

0.4 mA? That is 0.0004 A. My DM can't even measure that, because you have to use the 10A unfused connection, which is only accurate to the mA.

BTW, I got a reply to my question I asked him if the transformer said anything about what kind of fuse it required, other than the fact it needs to be slow blow. It doesn't say anything. But, he also mentioned that Amplimo (transformer manufacturer) requires fuses with a high I²t (I*i*t for those who cannot see the squared sign). And, also according to

there are different kind of slow blows. He's asking the local electronics shop for advice, perhaps they indeed have a fuse for him with an extra high I²t.

Do you also know it's typenumber? I think that would be a great solution.

Thanks, I'll have a look at it.

The 25 watts I determined my poweramp uses, was determined with a special device, capable of keeping the angle between current and voltage into account, so I guess that's acurate. But, it's not the transformer alone which does that, it's the quiescent current of four amps.

My electronics is mainly focussed on audio, and the small magnetic field of torroids is a good advantage for that. And, they are easily mountable.

I guess there is a normal type of slow blow, because I hardly ever see/hear anybody talk about the I²t requirements of fuses.

I have results. The NTC's are about 9 Ohms total when cold. And they don't decrease by any usable amount when warm (well, as far as they get warm)... On to find a better solution...

About this. Ever since the beginning, there is a slight 50 Hz hum in the output, almost inaudible, you have to put your head against the speaker to hear it. There are no harmonics, it's just 50 Hz. Could this be caused by these NTC's, that the supply ripple is getting to high? How much ripple can an amp with good common mode rejection ratio handle?

The filterbanks are 4x4700 µF, positive ripple is about 300 mV, negative 260. Transformer is 25-0-25, so DC is about +35 and -35. That ripple does seem quite high. The difference in ripple can be explained because the positive rail is used to drive 4 relays.

The hum is not present when the amp is powered from my bench supply.

The NTC thermistors I've salvaged from a whole range of scrap PC monitors have ranged from a few hundred Ohms to as high as 12k at room temperature, the running resistance is usually 5 Ohms or less.

Its among 20 or so Gb of data sheets on my other PC, but next time I'm there I'll see if I can remember which folder to look in!

These were also inrush limiters? Monitors also use these things in their degausing system, if I'm not mistaken. I can imagine their values needing to be a lot higher, to allow the alternating current to diminish to zero. Yet, you'd think they'd use PTC resitors for that (to let the current start high, end low when the resistor is hot), so perhaps I'm wrong here :)

But that is quite a difference. Mine don't differ at all when in use or not.

The 9-10 Ohms explains why there is "5.0" written on both of the NTC's BTW :). I wonder if they mean that's their room temperature resistance, or running-hot resistance.