PDA

View Full Version : OT-Fuse question?



wierdscience
04-14-2004, 08:38 PM
I got a strange thing happening at work,I built a welding positioner out of a bunch of junk we had laying around the shop,its for a new regular job and I only needed a small but still variable speed range.I used a dc motor of 1/4hp and 4 amps @90vdc,the control is noting but a variac and a bridge,the variac has a quarter turn fuse on the front rated at 8amps 125volts.I wired everything up and it runs perfect,I set a limit pin so the variac dial can't go past 100vac,there is very little load on the motor and nothing is overheating,but when I first start it up in the morning,it runs about 30 seconds and blows a fuse,I can replace it and run all day long without a hitch,but always in the morning it runs and blows that fuse.I even tried a different pack of fuses and the same thing occurs.This one has me stumped.If there was a big resistance in the system I could see it,but there isn't one.It also doesn't matter what speed the dial is set for.

Evan
04-14-2004, 08:44 PM
Try a slo-blow fuse. It could be warming up the bearings or something.

ibewgypsie
04-14-2004, 09:20 PM
Slo-Blo, time delay, I think they are the same thing..

PSD KEN
04-14-2004, 10:29 PM
Any large electolytic filter cap would be suspect. If leaky, would draw much current til' it "forms"

J Tiers
04-14-2004, 11:25 PM
Get out your clamp-on meter.

If the variac had a partly shorted turn that opened when it warms up, that could do it.

I assume "any" position includes all the way off, leaving just the variac on.

I can see a turn-on transient blowing the fuse, but a 30 sec delay is pretty odd. There has to be something drawing an overload for just a little while.

The usual fuse rating is on a curve, so a 30 sec blow time is a fairly significant overload, something between 135% of rating and 200% of rating according to fuse data.

The 135% is 1 hour max blow time, 200% is 5 sec min blow time for a time-delay fuse.
<A HREF="http://www.littelfuse.com/data/Data_Sheets/313_315.pdf" TARGET=_blank>
http://www.littelfuse.com/data/Data_Sheets/313_315.pdf (http://www.littelfuse.com/data/Data_Sheets/313_315.pdf</A>)[/url]

If its a fast-blow, then the time is a little different, 200% of rating blows in 5 sec MAX.
http://www.littelfuse.com/data/Data_Sheets/312-318.pdf


So it looks like something is making that draw about 12 to 15A for a while at turn-on, and you are going to need your clamp-on to find it.

If you can't find an overload, I'd be surprised, unless the fuse holder is corroded and gets hot.

[This message has been edited by J Tiers (edited 04-14-2004).]

wierdscience
04-14-2004, 11:29 PM
Evan,thats what these were,slow blows with the little indicator pin that pops out when burnt.

Thought about the bearing warm up,they run for an hour straight with no change.This rig has a 460:1 Gear reduction,all gears in 10wt,I could probibly run it with a 500lb part in the chuck and get by with a 1/8hp motor,its loaded really light.

I got no caps in the system so their out as a possibility.

I'm wondering if maybe the brush arm in the variac might be heating,but if thats the case it should blow all the time thou.

Inrush?Nah couldn't be.I'm driving all this through a foot switch,on the ac side of course to save the switch,maybe it is inrush but a bridge shouldn't be a high resistance right?

wierdscience
04-14-2004, 11:36 PM
Jtiers,Didn't think of that,I may just try a different variac tomorrow,Boss came up with a good idea,He said why don't we drop in an old lathe quick change box off something small like a SB or Atlas and widen out the range?Sounds like a plan to me.He even volunteered to throw down for a actual PM motor control.

Evan
04-15-2004, 02:50 AM
Just what RPM is that motor rated at for 1/4 HP? A dc motor current rating at rated HP and RPM is nowhere near the same as the stall current. DC motors draw max current at stall. If that motor is a fairly high RPM motor and is being loaded down quite a bit RPM-wise then it will be drawing a LOT more current than nameplate. If you have an ampmeter stick it in the circuit. Betcha the motor isn't suitable for the load. Also, you are running it on pulsating 120hz DC. This will screw with the current readings, especially with a digital meter.

wierdscience
04-15-2004, 09:31 AM
1800rpm,I digital tached the shaft under load,reads 1791,no big load at all.Also,no heat whatsoever,I can run it at 30volts for an hour or more and the motor doesn't even get warm.I thought about current rise,but if that were the case it should blow the fuse all the time,not just on first startup.

mikem
04-15-2004, 10:23 AM
My guess is that the variac is the problem. We use one at work for ramping up the voltage from 50% to full for diagnosing electronic circuits for partial shorts and voltage regulation problems. Sometimes, when I shut off the load or unplug the variac, the fuse will blow. The magnetic field of the auto transformer in the variac will charge and discharge when AC is applied and removed. Maybe the wiper in the variac is making and breaking contact and causing a big transient current flow. Thanks--Mike.

J. R. Williams
04-15-2004, 10:37 AM
Change the switch to the output side of the variable transformer as switching the input will cause inrush problems.

JRW

Paul Alciatore
04-15-2004, 12:29 PM
If you want to isolate the problem you can disconnect the motor at night before leaving and when you turn it on the next AM it will be just the variac in the circuit. Disconnecting one wire should be enough. If it pops the fuse that way, it can't be the motor. You could try cleaning the brushes and exposed turns on the variac.

If it stays on, leave the motor disconnected for a few minutes and then connect it. It would be best to do this while the power is still on - CAREFULLY of course. If it pops then I would suspect the motor or something connected to it. Sometimes a tight mechanical assembly can "stick" when left idle for a while (overnight) and after it frees up it will run until another long idle period occurs.

Paul A.

J Tiers
04-15-2004, 01:35 PM
If it blows in 30 sec, it is NOT an inrush problem. That inrush is over in milleseconds.

The indicator type fuse may be different from the standard type 312 or 313 that I linked to.

But almost all "variac" or powerstat units I have ever seen use the standard type 3AG fuse, or possibly the 3AB (ceramic sand-filled) which is the same size.

The 3AB does have a 5 sec min and 60 sec max blow time at 200% of rating, so it is a little more tightly controlled.
http://www.littelfuse.com/data/Data_Sheets/326-325.pdf

docsteve66
04-15-2004, 06:12 PM
Weird: measuring current to motor tells little. Measure amperage going TO the variac. A fuse makes an unreliable amp meter.
Your most likely problem is excess current TO variac or heat at the fuse.
Two things I can think of:
first, the fuse holder may be defective, the spring having lost its temper or having a dirty contact. If that be the case, the blown fuse should not be blown in the middle, just melted at one end due to heat.

Second: check the brush on the variac and be sure it is not "bridging" two wire (which would give the equivalent of a shorted turn and blow a fuse.

Neither of those defects would explain why the system runs the rest of the day, so I am mystified about the "repeatable intermittent" fault.

wierdscience
04-15-2004, 10:08 PM
Well I found it,it was the variac,swapped it for another and everything is fine.Funny thing is I never once while building this did I turn it over,I did when I took the cover off and thats when I saw the word "junk" written in marks-a-lot http://bbs.homeshopmachinist.net//biggrin.gif

I popped the cover off for curiosity and from about 110-200 degrees of arc on the windings the copper was discolored(looked like it got real hot)so I stripped it,chopped off the cord and threw the rest in the pile.Normally,if it doesn't work and can't be fixed I reduce the objects size in the shop press and discard BEFORE it hits the shelf,Guess this one slipped in.Thanks for the ideas.I will post pictures when I finish modifying to the final version.

[This message has been edited by wierdscience (edited 04-15-2004).]

J Tiers
04-15-2004, 11:21 PM
Well, you sure proved it was junk! Good. Figured it had to be something about the variac.



<font face="Verdana, Arial" size="2">Originally posted by docsteve66:


Second: check the brush on the variac and be sure it is not "bridging" two wire (which would give the equivalent of a shorted turn and blow a fuse.

</font>

Point of information: pretty much all the variacs I have seen do have brush overlap. It doesn't seem to cause a problem, any more than the overlap on a DC motor commutator.

Even the ones that are beveled a bit still bridge over at least two contacts in some positions.

I think some brushes may be made to have better coductivity in one direction than the other. This can be done in various ways when making the brush from granular/powdered materials.

wierdscience
04-15-2004, 11:33 PM
I was tuaght that a variac/rehostat brush was supposed to be kind of a triangular affair so the contact surface wasn't so wide as to bridge more than two loops at once,this from a Licolon welder service man.He said that the brushes wearing flat was the cause of more than one complaint about vague voltage controls.

darryl
04-15-2004, 11:39 PM
Interesting about the variac. I've oftem wondered why there isn't a shorted turn with the brush, since it does bridge more than one winding. In a transformer, which is basically what this is, a shorted turn will draw lots of current. Shouldn't the brush be burning up? Anybody have a definitive answer on this? Maybe that variac is bad because of something related to this. Polish up the surface where the brush wipes, and it's good as new?

Evan
04-16-2004, 03:50 AM
Simple. Shorting a turn makes no difference. If you short a couple of turns on the primary of a regular transformer nothing bad happens, it just appears to have less turns. With an auto transformer the "primary" winding is the turns from the contact point and hot. The shorted turns are effectively part of the primary. The secondary is all the turns from the tap to neutral, none of them are shorted.


Think of a regular transformer with a centertapped input/output for 110/220 conversion. Short one end of the primary to centertap and apply 110. Smoke? No. Not the same as shorting the turns on the secondary of a regular transformer.

[This message has been edited by Evan (edited 04-16-2004).]

J Tiers
04-16-2004, 10:07 AM
<font face="Verdana, Arial" size="2">Originally posted by Evan:
Simple. Shorting a turn makes no difference. If you short a couple of turns on the primary of a regular transformer nothing bad happens, it just appears to have less turns.

Think of a regular transformer with a centertapped input/output for 110/220 conversion. Short one end of the primary to centertap and apply 110. Smoke? No. Not the same as shorting the turns on the secondary of a regular transformer.

[This message has been edited by Evan (edited 04-16-2004).]</font>

All I can say to this is "try it".......make my day.....

Ain't so.

Smoke? HELL YES!

ANY shorted turn on a transformer with magnetic alternating flux in the core will have current flow in proportion to the volts per turn, the number of series turns shorted across, and the resistance of the "shorted" turn. Makes no difference if you "call it" primary or secondary. If shorted, its a secondary.

The resistance of the variac brush is high enough, and the volts per turn low enough, that the "short" current is low. Same reason carbon brushes worked on motors when vanDepoele tried them.

If the variac brush NEVER "shorted" turns, the output voltage of the variac would go to zero periodically when the brush was "between" turns. It does not.

Evan
04-16-2004, 12:01 PM
JT,

Let explain this a little better. You have a primary that is centered tapped. You connect one end of the winding to neutral. You connect the center tap and the other end together and to hot. Those two connections are at exactly the same potential so there can be no current flow in that portion of the winding from the source. Current induced in that closed loop by core fields is 180 degrees out of phase with the core field so the only dissipation that occurs is due to hysteresis losses, very small. It is the same reason why a transformer with no load on the output but connected to power draws very little current.

As to the resistance of the variac brush, it better be in the milli-ohm range or it would burn with any significant load.

Paul Alciatore
04-16-2004, 12:36 PM
Re: Shorted Turns

Yes, a shorted turn in a variac or any transformer will have a heavy current. And yes, most variacs are constructed to have contact on two or more turns at the same time. If they weren't, they would have many, many places where you get no output and the language would fly.

However, if you look at most variacs, you will see wire that is very heavy gauge for the current rating of the device. In addition, the resistance is low and the voltage is low (determined primarily by the contact resistance, not the wire) so the power formula P = I*V or P = I*I*R gives a low value becaues the voltage or resistance (they are related) is so low. Very little power is generated there and since heat is a form of power, very little heat. So the winding does not burn up. No smoke.

I think that you see problems when the sliding contact wears down because all that conductive dust is left in the spaces between the coils and eventually forms a short circuit with relatively high resistance. This does generate considerable heat which causes the damage. Variac life can be significantly prolonged by periodic cleaning and brush replacement.

Paul A.

J Tiers
04-16-2004, 02:14 PM
It is worth understanding this stuff, since it applies to motor windings etc. So I will go through it again...


<font face="Verdana, Arial" size="2">Originally posted by Evan:
JT,
Let explain this a little better. You have a primary that is centered tapped. You connect one end of the winding to neutral. You connect the center tap and the other end together and to hot.
</font>

A center-tapped winding is one in which a wire is wrapped around a core a number of turns. Say 220 turns total. The wire has a connection made at the middle of the winding (in this case after turn 110), called the center tap (CT). So there are three leads. Wire 1, CT and wire 2.

At that point you have two equal windings, on each side of center tap (definition of center tap). They are wound in the same direction around the core.

Since your example is applying a voltage to the CT and "grounding" wire 1 to neutral, the CT is defined as "in phase" for our discussion.

Wire 2 will be "out of phase" with wire 1, but "in phase" with the incoming voltage to CT. Otherwise it isn't a normal CT, and isn't like a variac.

the reason:
Any winding develops an induced voltage which opposes the incoming voltage (AC). So there is a voltage from wire 1 to CT, opposing the applied voltage.

Since the winding from CT to wire 2 is wound the same direction and same number of turns, there will also be an equal voltage from CT to wire 2 , of same polarity as from wire 1 to CT.

You have 110 volts AC applied from neutral to CT.
Since there are the same number of turns from CT to the other end of the winding (wire 2), and the windings are on the same core, etc, etc, you MUST have a potential of 110VAC from CT to wire 2.

If you didn't short wire 2, at the instant when there is +110V on the CT relative to neutral, there will be a voltage of +220V on wire 2 relative to neutral (sum of the two voltages). Wire 2 will always follow at 2X the voltage of the CT with respect to neutral.

This applies to 110/220 motors also, where identical windings are put in series or parallel depending on voltage.

If you get them backwards, you get a heavy current draw.


<font face="Verdana, Arial" size="2">
Those two connections are at exactly the same potential so there can be no current flow in that portion of the winding from the source.
</font>

There can be, and if there were a load from neutral to wire 2 there would be. That discussion will confuse the issue, and I don't want to go through it here.


<font face="Verdana, Arial" size="2">
Current induced in that closed loop by core fields is 180 degrees out of phase with the core field so the only dissipation that occurs is due to hysteresis losses, very small.
</font>

No, there is a potential from CT to wire 2, and if those points are shorted, the only opposition to the flow of current is via the resistance of the wire, and any parasitic or "stray" inductance. There will be a lot of current.

As we found above, the induced voltage would add to the CT voltage with respect to neutral if not shorted.


<font face="Verdana, Arial" size="2">
It is the same reason why a transformer with no load on the output but connected to power draws very little current.
</font>

No, that is because of the "magnetizing inductance" of the primary. That is due to the many turns around the magnetic core.

The induction *between* windings in a transformer is termed "mutual inductance", which is a measure of the coupling between them.

The "mutual inductance" from primary to secondary is effectively "in parallel with" the magnetizing inductance, so transferred current it is not affected by the magnetizing inductance. In this example the effective secondary is the winding from CT to wire 2.

The center-tapped transformer you describe is essentially the same as a variac with the brush set at half rotation. If you look at the variacs behaviour as per the above, you will find that it has to agree with what I said.


<font face="Verdana, Arial" size="2">
As to the resistance of the variac brush, it better be in the milli-ohm range or it would burn with any significant load.
</font>

Brushes can be made with differing resistance across the face vs through the thickness, by layering materials, etc.

I don't have the numbers for variacs, but I would assume they do this. The brush can't be allowed to get very hot, but it will heat somewhat.

The "through" resistance would need to be low, but the "across" resistance could be higher without harming the brush.

The brush resistance is one of the main limits on the current rating of a variac. At low settings, the low voltage output current may be far above the input current. The brush must be protected from such high currents.


[This message has been edited by J Tiers (edited 04-16-2004).]

darryl
04-16-2004, 03:06 PM
Interesting discussion. Ok, it seems probable that the brush is specially made to minimize conduction across the contact area, but have high conduction up into the brush. Makes sense, I had no idea that a brush could me made that way. Also if the contact face of the brush is tapered, that width will grow as the brush wears, increasing the current flow in the brush, and increasing the heating in the brush. Also makes sense. If the contact area of the windings is not clean, that may decrease the brush heating due strictly to it's contact with the wires, but will increase the heat when a load current is drawn. I have to suggest then that as long as the variac can be powered without overheating, meaning no shorted windings, has a cleaned contact surface on the wiring, and the brush can be re-tapered to original width, then it can be restored to normal operation. I make the assumption that it is mechanically operable, and that all other connections, switches, fuses and holders, have proper low resistance contacts.

Evan
04-16-2004, 04:13 PM
JT,

"Magnetizing current lags applied voltage by 90°, while core loss is in phase with the applied voltage (Figure 6b). [i]VP and VS are shown 180° out of phase.[/] IH is very small in comparison with Im, and Im is nearly equal to IE. No-load current, IE, is also referred to as exciting current."

http://www.tpub.com/content/doe/h1011v4/css/h1011v4_57.htm


Vp and Vs correspond to the primary and secondary windings. The example I gave is not the same as an autotransformer as it is not connected in the same way. The induced current in the shorted primary section is 180 out of phase with the supply current so, not including losses, no current flows in the shorted winding. If you took the output tap of an autotransformer and connected it to hot the you would have the same configuration.

However, in the autotransformer in this configuration you could fry the windings below the tap if you set it too low as the magnetic efficiency would fall dramatically as the number of unshorted windings fell and the exciting current climbs. Of course, there is no reason to do this but it explains why windings shorted by the tap don't fry.

J Tiers
04-16-2004, 05:48 PM
[QUOTE]Originally posted by Evan:
The example I gave is not the same as an autotransformer as it is not connected in the same way. The induced current in the shorted primary section is 180 out of phase with the supply current so, not including losses, no current flows in the shorted winding.........
[QUOTE]

I understand you to be suggesting that a shorted portion of the primary will NOT have a large current flow in it.

I am suggesting that you don't quite have a handle on the physics involved.

Rather than drawing this discussion out, try it....


Take a dual primary transformer, and short one primary. You can hook the shorted point to the line or not, won't matter. Put your clamp-on ammeter on the wires of the shorted section (or other ammeter as applicable, being careful to avoid touching live wires).

I would suggest bringing the other primary section up slowly on a variac, watching the current in the shorted section. Directly hooking across the line may produce unpleasant surprises.

That should provide a swift, clear illustration of what happens.

Paul Alciatore
04-16-2004, 06:52 PM
Re: Brushes with directional characteristics

Anything is possible and some manfacturers may have found a way to make such brushes. If so, they would have to have not only directional conductance characteristics but also diode like properties to avoid current flowing back to the other side of the brush from the common connection point for the wire. But then, this is AC and it can't have diode properties, can it?

But even if it was possible to somehow engineer something like this in today's world (and I doubt it), I know for a fact that variacs have operated for many years without the aid of such Sci-Fi devices. As I said before, a single shorted turn or two will have a heavy current but will not generate enough heat to do any damage unless the brush or the wire is damaged. The low resistance of the shorted turn will prevent any significant voltage from building up and it will not disipate any significant power.

Paul A.

Paul Alciatore
04-16-2004, 07:02 PM
<font face="Verdana, Arial" size="2">Originally posted by J Tiers:
.....

I am suggesting that you don't quite have a handle on the physics involved.

Rather than drawing this discussion out, try it....


Take a dual primary transformer, and short one primary. .........

That should provide a swift, clear illustration of what happens.

[/B]</font>

J,

This is not a fair test as ordinary transformers are not built to the same specs as variacs. Variacs will have much heavier wire for similar overall ratings.

Variacs work. They have for over 50 years. The brushes do short the turns. They really do. If they didn't, there would be spots with no output. Variacs keep working anyway.

Paul A.

Arcane
04-16-2004, 07:18 PM
Evan is right. What many are forgetting is that one turn (or even several) of a primary or secondary has zippola for resistance and impedance and a short consisting of the distance of a few turns won`t be noticeable. If there is a considerable distance involved, it`s possible that the output voltage will be higher or lower than design says, and the transformer will continue operating like that almost indefinitely. There does come a point when impedance does not limit the current enough to allow the transformer to continue operation and it burns out. Also think about this for a second... How else does a power transformer in a substation vary the voltage (can be once every 45 seconds) except by the tap changer physically changing the number of windings by moving a contact from tap to tap under full load and without ever having an open circuit. (An open circuit would destroy the tap changer in short order, not to mention every voltage sensitive piece of equipment on the feeders would be shuting off-think computers w/o UPS).

darryl
04-16-2004, 07:23 PM
Another factor with a variac which I didn't think of til now if the fact that the turns per volt is usually higher. This means that the voltage between windings is less, so less current will flow in a given resistance, in this case the brush. Also, the length of the turn is usually longer, due to the construction method of the variac, so the wire's resistance is higher, further minimizing the power lost in the brush.
A test to see the effect of one shorted winding is to pass one loop of wire through the core on any transformer, then hook up power to the transformer through an ammeter. When you short the loop, the current will rise, and the current through the loop will be significant. Whatever the increase in current draw, that times the primary voltage will be the power drawn by the shorted loop. Not exactly, but close enough for this exercise. Use your fingers to hold the ends of the loop together, and see if there's any heat. Carefully.

wierdscience
04-16-2004, 08:38 PM
Allow me to throw a wrench in the works,same man who told me about the brushes also told me why it cuases trouble.

If you look at the variac brush arm you will notice that spring tension is applied to the brush via a flat spring,when the point is worn off flat the tension decreases,if the machine is mounted in such way (as in the case of a welding machine rehostat)that it vibrates it can cuase the brush to make and break contact between two or more turns,the resulting arcs are what cause the problem(heat)that destroys the windings.Also as the brush jumps it tends to heat the spring cuasing it to lose even more tension exacerbating the problem.This I think explains the variac I was using,I noticed it did turn very easy compared the the one I replaced it with.

docsteve66
04-16-2004, 09:12 PM
I go with darryl here. I just inspected a variac. Most transformers have one turn per volt, variacs have more (my 1000 kva) does any way. A smaller one looks as though it has even more turns per volt. The brush does bridge two turns. I see no sparking nor signs of heating. The length of the wire per turn is very long. I suspect the "short circuit current is low due to high resistance of the wire. the unit is rated at 6 amps max in any part of the circuit, the wire size is larger than 14 gauge. the variac is made by general radio. The original brush (which is still in my variac) MAY be of special construction- I did not take it apart to measure, mainly cause I don't think the brush is special. BUT I HAVE used the center carbon from a BA-30 (C or D cell) to get a variac back in service. It acted normal and was in use for a year or so that I know of.

In a normal small power transformer, a single shorted turn, in low or high voltage winding will "burn out" the primary winding or burn in two pieces its self (open winding fault). MOst times it is the primary (with much larger wire than the high voltage secondary wire size) that releases the smoke.

The comparison of variac to motors is flawed. Consult section two of crofts, and they discuss pitting of brushes due to the shorting of commutator segments. Solution is to shift brushes to a neutral point. I have had to do this on large plating generator to get acceptable brush life on a generator with multi ple brush, the brush face measured about 2" thick and 6 inches wide, two such brushes per segment. Brush life was in days because the commutator segment shorted.

Now for some useful dope: I needed (wanted) to heat a bearing, no electric bearing heater, Took an old microwave transformer (driving the magnetron), removed the secondary and put the top piece back on (to be honest, the transformer was disassembled long ago because of another thread). The bearing was too small for the whole top piece to fit through so i used maybe 1/3 the top piece. Plugged it in and bearing was hot in a few seconds maybe a minute. Noisy (60 cycle "hum") but nothing exciting happened. Electric bearing heater go for hundred of dollars, this was just from junk. a used microwave in the 1500 watt range is cheap or free. despite the lack of a full sized top piece, the 16 gauge cord or the winding nor the frame heated- just the shorted single turn represented by the bearing got hot.

J Tiers
04-17-2004, 12:53 AM
I hate to belabor this, but with the amount of electrical questions that come up in the boards, it is worth getting it right and avoiding mis-information.



<font face="Verdana, Arial" size="2">Originally posted by Arcane:
Evan is right. What many are forgetting is that one turn (or even several) of a primary or secondary has zippola for resistance and impedance and a short consisting of the distance of a few turns won`t be noticeable.
</font>

Actually, "zippola for resistance" and a volt per turn (or half a volt, etc) leads to a high current in a wire perhaps intended for 5 amps long term when wrapped into insulation and buried in a multi-layer winding. A surprisingly small amount of heat loss leads to high temps over the insulation class limits. Burning and additional shorts or inter-winding contact can result, possibly connecting the primary to secondary, with obvious safety hazards.

18GA wire is about 8 milliohms per foot. With a voltage of 1 volt, a turn a foot long could allow up to about 125A current.....which will rapidly overheat the insulation. Most small transformers have less than one foot of wire per turn.

Even a quarter volt would give 30A in the wire of what may be a 1A or 2 A winding. Power dissipation is proportional to current squared, do the math....

In fact, you have identified the exact reason FOR the problem.....not the reason for "no problem".


<font face="Verdana, Arial" size="2">
Also think about this for a second... How else does a power transformer in a substation vary the voltage (can be once every 45 seconds) except by the tap changer physically changing the number of windings by moving a contact from tap to tap under full load and without ever having an open circuit. (</font>

Tap changers often are operated with a tapped series inductance that limits the "circulating current" without reducing the line voltage much. Tap changers often have about a 2 volt step. A couple volts drop is enough to flicker lights, but not enough to reset computers.

But a relatively small impedance will limit current flow with a 2 volt source to an amount that is tolerable for the short time. It is a transient condition, not lasting more than a few dozen milliseconds. The only consideration is to avoid damaging the contacts with a heavy current.

Those tap changers are substation equipment, on large transformers. They can afford the current limiters.


<font face="Verdana, Arial" size="2">
Re: Brushes with directional characteristics
Anything is possible and some manfacturers may have found a way to make such brushes. If so, they would have to have not only directional conductance characteristics but also diode like properties to avoid current flowing back to the other side of the brush from the common connection point for the wire. But then, this is AC and it can't have diode properties, can it?

</font>

No, I got the info from an old Stackpole Carbon book on brushes from in teh 50s or 60s. Details were not given, but since metal-loaded brushes are used for some purposes, I would presume that layers of loaded material interspersed with higher resistance material could be produced.
This would have higher resistance perpendicular to the layers than along them.

Differing materials are used to produce high and low resistance brushes, depending on the "volts per bar" of the motor for which they are designed, among other things. So there are options for doing just that, and have been for 100 years.

No "diode action" is involved, BTW.



<font face="Verdana, Arial" size="2">
If you look at the variac brush arm you will notice that spring tension is applied to the brush via a flat spring,when the point is worn off flat the tension decreases,if the machine is mounted in such way (as in the case of a welding machine rehostat)that it vibrates it can cuase the brush to make and break contact between two or more turns,the resulting arcs are what cause the problem(heat)that destroys the windings.Also as the brush jumps it tends to heat the spring cuasing it to lose even more tension exacerbating the problem.This I think explains the variac I was using,I noticed it did turn very easy compared the the one I replaced it with.
</font>

Yes, low pressure could cause a problem.

I know one manufacturer is pretty fussy about their replacement brushes, might be layered, might not.

I doubt that the volts per turn is enough to cause a problem from bridging turns. I have several variacs, some obviously have more than 120 turns, some have about that. All of the brushes bridge at least two turns, depending on position.

I brought up the layered deal simply because it was one possible way to avoid the issue that was claimed to be a problem.
However it is done in a variac, it ain't a problem, they work great.

Arcane
04-17-2004, 02:04 AM
I still stand by my claim that shorting out a few turns of wire isn`t going to be the problem you say it is. A transformer (and I am talking power transformers like you see hanging on the power poles) has several hundred turns on the primary. Think of shorting out a few turns as being the same as winding it with that many turns less. This will raise the secondary voltage by the ration of original turns divided by the resulting turns. Also, since the high side voltage varies considerably, this will give the same effect at the secondarys. Off load tap changers on some transformers do exactly this, they "short out" a portion of the windings to achieve a different ratio to correct the secondary voltage.
I also don`t get your 18 gauge wire with 8 milliohms/foot and 1 volt giving 125 amps. DC theory? It doesn`t apply here because it`s AC in a transformer and you also have not considered the resistance and impedance of the remaining primary winding. All that you have said is that with a resistance of 8 milliohms and 125 amps current flow, the voltage drop will be one volt, which would be true for DC in a simple circuit, but where does the 125 A come from in the first place? No transformer I know of would ever pull 125 amps on the high side thru 18 gauge wire in the primary windings and if it was on the low side where 125 amps is a reasonable load, you won`t find 18 gauge wire used there either. Primary load would be 2.67 amps roughly for 125 amps on the secondary. (14400V-240/120V) Shorting out a secondary turn has a more pronounced effect, but if it just eliminated one turn, it might operate like that for a long time, load being a deciding factor. Lightly loaded and it might last forever, because of the heat factor. Once again, ratios would be changed,
On load tap changers are designed to tap the windings up or down without in any way interrupting the load circuit. At some point in the transition from one tap to another, both taps must be joined together, and as the two taps will be at different voltages with respect to the end of the transformer winding, a circulating current will flow in the (bridge) circuit. This circulating current is limited to a safe value by either using a center tapped reactor or a preventitive autotransformer in the bridge circuit. The point is, this is a deliberate "short" and is a normal operating procedure.

J Tiers
04-17-2004, 11:49 AM
<font face="Verdana, Arial" size="2">Originally posted by Arcane:
I still stand by my claim that shorting out a few turns of wire isn`t going to be the problem you say it is. .
I also don`t get your 18 gauge wire with 8 milliohms/foot and 1 volt giving 125 amps. DC theory? It doesn`t apply here because it`s AC in a transformer and you also have not considered the resistance and impedance of the remaining primary winding.

All that you have said is that with a resistance of 8 milliohms and 125 amps current flow, the voltage drop will be one volt, which would be true for DC in a simple circuit, but where does the 125 A come from in the first place? No transformer I know of would ever pull 125 amps on the high side thru 18 gauge wire in the primary windings and if it was on the low side where 125 amps is a reasonable load, you won`t find 18 gauge wire used there either. Primary load would be 2.67 amps roughly for 125 amps on the secondary. (14400V-240/120V) Shorting out a secondary turn has a more pronounced effect, but if it just eliminated one turn, it might operate like that for a long time, load being a deciding factor. Lightly loaded and it might last forever, because of the heat factor. Once again, ratios would be changed,
On load tap changers are designed to tap the windings up or down without in any way interrupting the load circuit. At some point in the transition from one tap to another, both taps must be joined together, and as the two taps will be at different voltages with respect to the end of the transformer winding, a circulating current will flow in the (bridge) circuit. This circulating current is limited to a safe value by either using a center tapped reactor or a preventitive autotransformer in the bridge circuit. The point is, this is a deliberate "short" and is a normal operating procedure.</font>


So many misconceptions that I hardly know where to start.......

The current in a shorted turn comes from the physics of the transformer. A certain alternating flux exists in the core. Any closed conductive circuit around that flux will have a voltage induced in it, and a resulting current flow. If the circuit is open, it will have the voltage, but no current.

You cannot just "label" something as "primary winding, therefore irrelevant".

A secondary is ANY winding not powered from an outside source. Even the primary has an induced voltage in it in opposition to the applied voltage

Any unpowered wire wrapped around the core is a "secondary", whether it is connected at one point to the primary or not. A shorted turn is like that, connected at the point of the short, and unpowered (because shorted).

The 18Ga example is because I happen to remember the resistance of 18ga. And it is not uncommon in variacs to use a wire of that general size.

If there are 120 turns (not unreasonable for a medium power transformer at 120V), then there is 1 volt per turn, AC.

The turn of wire has some resistance, about .008 ohm for 1 foot in the case of 18 ga. There is also a stray inductance, which is generally pretty small, as it represents a loss that the designer minimizes.

The series impedance of the turn is therefore quite small, and the 1 volt per turn can produce a very large current. In this case it would be around 125A (before the copper heats and resistance goes up). What if other impedances limited it to 100A, is that much different? Nope...

THIS CURRENT ONLY FLOWS IN THE SINGLE TURN. The rest of the (un-shorted) primary is not carrying that current, it merely flows thru the short. If the short is opened, the 125A current would instantly stop.

The powered primary winding will have a current in it according to the turns ratio. With 120 to 1 turns ratio, it would be about 1A.

As far as the rest of the transformer is concerned, everything is normal aside from some extra current.

The problem is that the short is permanent....it simply heats up until additional damage is done to insulation, or the wire melts open.


The tap changer is totally different....it is a momentary short, which is known and designed for. The transformer winding cannot heat significantly in the small time available.

The reactors used to limit current protect the switch elements from overcurrent. The wires etc are fine in any case if they don't heat up or have so much current that magnetic forces do mechanical damage.

Motors, transformers, etc are generally OK with severe overloads if they don't persist so long that heat damages the device.

The exception is contacts and switches. They can be damaged by overcurrents, or by breaking larger currents than they are designed for.

So, yes it is true that if the shorted turn does not heat beyond limits, it can operate for a long time. But most small "dry" transformers such as we are discussing (or were) will be damaged by a shorted turn in a matter of minutes, because they are not oil-cooled like distribution units.

If you check the post on the bearing heater above, perfect example. So is the home-brew spot welder, one or two turns of heavy wire added to a large transformer. Big current because of low resistance.

So far as I know, physics hasn't been repealed yet.

darryl
04-17-2004, 05:58 PM
J Tiers has it right. Any closed conducting loop surrounding the core will be energized by the flux in the core. It doesn't matter if it's a primary winding, secondary, or whatever you call it. All the transformer knows is that the flux is being shorted out by a load, in this case a shorted loop or winding. The primary current will increase to restore the flux level, and the increased power goes into the short. We're assuming pure transformer action here, though. An interesting thing happens in a very small transformer, such as the so called plugpacks, or wall adapters. In these, the resistance of the primary is so high that even if the output is shorted, the thing can remain powered for a considerable time before having a meltdown. The primary winding changes from being an inductance to a resistance when it cannot maintain the flux in the core. If a few turns short out, the primary resistance becomes more dominant, and the power drawn by it doesn't rise much. Because there are so many turns per volt, the volts per turn are low, and the resistance of the wire so relatively high that the current flowing in the short is also low, and doesn't represent much power. It can go on working, though the output will be reduced. The larger the transformer, the more efficient is the transformation, and the more serious a short becomes.

[This message has been edited by darryl (edited 04-17-2004).]

Evan
04-17-2004, 09:54 PM
Me and my big mouth. I haven't seen anybody address the issue of core saturation. Transformers of all sizes are usually designed to operate just below core saturation (excepting saturated core voltage regulating transformers). Given that, then the flux strength cannot support a total wattage in any number of windings, including just one, above that the transformer is designed to deliver. Depending on the size of the primary windings this may well be within the capability of a single turn to withstand, even indefinitely. Note that core flux strength is NOT dependent upon load. The magnetic flux at zero load is the same as at full load.

docsteve66
04-17-2004, 10:33 PM
Evan: I suspect thats why the primary burns out. Saturation is by no means a definite point- the BH curve is linear for only a short distance. When you get into saturation (non linear region) harmonics are developed- be it 60 cycles or higher (audio). WHen driven into high portions of the bh curve, the windings act more like they are excited by DC and current flow is about what the dc Resistance of the wire would be.

ON the "shorted" turns, in open air, 30 gauge wire will carry 46 amps for .1 seconds or 4 amps for 10 seconds and then fuse. 20 gauge will carry 470 amps/.1 second or 47 amps for 10 seconds. even at one turn /volt and a 3 by 4 core winding you have 14 inches of wire, and at ten ohms per thousand feet the resistance is low enough to fuse the wires in short order.

A correction, please: I mistakenly said the 1000 watt variac has more than one turn per volt. I looked again- DID NOT COUNT all the turns but just for a ten volts range. I now think the turns per volt is one. Sorry for the mistake. My puzzlement is that one turn is shorted, high current should flow in the shorted loop. But this one is one i have owned over fifty years, and like most general radio variacs, is trouble free in long term service. I need an education!
Steve

darryl
04-17-2004, 11:50 PM
Having done a little research on variacs, I can now say that the brushes are indeed special, though not layered from side to side, as has been suggested. They are of a soft carbon compound, and do have a considerable resistance to them, otherwise they would burn up from the current. If they are replaced with ground-down motor brushes, that does cause a problem with brush and wire burning up. Also sparking. It will definitely not work to use a metallic contact, as then the types of currents we are talking about will try to flow in the shorted turn(s). If anyone would want to try making their own replacement brush, start with something like a vacuum cleaner brush, and not a starter motor brush. The answer is right there, the starter motor brush is designed to be highly conductive, and will create the type of short across windings that you don't want. Some of the literature I have read states a spec for a variac brush's power dissipation, one said five watts. Replacement brushes for variacs are not cheap, hence part of the reason why variacs are scrapped, even though they are essentially still functional. Don't toss them, you can always tap a winding to give a needed voltage, and just forget about the brush. I saw one brush set costing 250 bucks plus, and I stopped looking. One article stated that the winding had an over and under pattern, where every second turn was hidden below the 'active' turn. That must be a bitch to wind.
My own homemade variac has two sliding taps, one on each end of the core, so I have two separate outputs. One is for the lathe motor, the other for the toolpost motor. Rather than having a brush sliding on the wiring, I have made multiple contact points, so the contact is on only one point at a time. It's annoying to have all the dead spots, but I do get a positive contact when I stop at the right voltage. I suppost this shouldn't be called a variac, since technically, it's a multiple-tapped transformer, though I'm using only one winding, the primary.
Something else I found interesting is a reference to a 3-phase variac. Hmmm.