Vibrating Wire

I'm with Lyn, use DC, it's cheaper than getting a HF power supply.

Third vote for DC... or second best, increase that frequency by quite a bit.

Ultrasonic welder which is used in the paper industry making disposable diapers.

Dukane Electronics and better known Branson Ultrasonics , a subsidiary of Emerson Electronics.

I had design ultrasonics cutter using these makers. I have found at that time (over 15 years ago), that Bronson has quite a bit of experience in these, and as I recall the frequency what Bronson uses is what is a called a clipped sine wave running at 50 Hz.

Now to stop the vibration you will also affect it's performance. If that is what I understand from your post.

Ultra sonic welders require contact between the two materials and pressure.

My understanding is that this is a radiant heat process and the wire does not contact either surface.

Crap, I over wrote the post, But basically, I would need more information.

Hi Lyn,

Yes you are correct. It is a radiant heat. The wire does not contact the surface. When it does it burns the material, which is then scrap and ruins the wire as there is a carbon build up on it. It also makes a lot of smoke!

Paul.

I would also like to add my vote for DC, but suggest you not overdo it on the filtration of the DC bridge, try Yalasit's suggestion first. I suspect your welding filament is coupled magnetically to ferromagnetic parts of the machine, hence converting the AC to DC with even 10% ripple compared to AC will dampen the vibration enough to make your job sweet.

I would imagine that as the wire is red hot, adding more tension to it will only thin the wire and promote wire breakage. I would look to adding supports as close to the field of work as possible. That may attenuate the motion to the point where contect with the heated material is avoided. I would perhaps look to a DC power source as your solution.

More tension may not be feasible in this situation...or it might be feasible. Several options might not be feasible, but it doesn't hurt to suggest multiple options when brain storming.

By all means... I agree completely.

There is no such thing as a bad suggestion. If I gave the impression that I thought the suggestion of more tension was "bad", it was not my intent. I was just pointing out on of the downsides of more tension.

Can you move the transformer further away from the hot wire: the idea being to allow the wire's alternating current to act in an area where there is no magnetic field.

(You will need pretty thick wires.)

'.... allow the wire's alternating current to act in an area where there is no magnetic field....'

.

Not really an option.

The wire in question and the remainder of the path for current sets up the magnetic field.

Creating a longer path for the current by moving the transformer is more likely to increase the magnetic field, and it certainly won't eliminate it.

Maybe. I think it would depend on how the power wires are fed to the business end.

Another thought: I wonder if the transformer is mechanically coupled to the bits holding the hot wire.

GA on your post #6 BTW.

You are right, it is highly dependent on the configuration of the conductors leading to the wire.

My assumption was, that if you can improve the configuration, it most likely won't require moving the transformer.

I hadn't thought about the transformer being mechanically linked to the wire. Hopefully the people who designed this equipment did.

As others have said wires vibrated because of the frequency of the current. A DC current would eliminate that vibration caused in the wire itself. But not seeing the operation there could be other causes. Has the equipment performed as it should prior to this problem arising. If an automated process any of the motors could be causing it. Also improper tensioning of the wire. Improper spec on the clearance. Even other machines with out good vibration damping could cause it.

It is due to the thermal cycles of your hot wire plus an alternating magnetic field around the wire caused by the frequency of your main supply, you CANNOT increase the frequency, since you're using a variac and a step-down x-former.

You have to use a full-wave rectifier plus a capacitor of about 15000uF and of higher voltage than the maximum obtainable from your transformer.

Its pretty simple, you shouldn't have any problems figuring out the connections.

Another possibility (Tho I would go for DC myself) might be to mount one end with a spring - enough to keep a mild tension & stop vibration, but not enough to break it. Don't run current through the spring, connect power on resistance side. Also, a thermal insulator between resist wire and spring might be helpful.

While technically it might be a possibility, the probability of it being beneficial is likely exeedingly small.

I'm not so sure building in another degree of freedom via an un-damped spring would have much chance of reducing the vibration.

I'd give you good odds that the problem would get worse.

*All this is based on the assumption that the spring used has a spring constant sufficiently low to actually behave meaningfully as a spring (and not as if it were clamped into a very rigid block.

The great thing about living in a democracy. It might depend on whether you approached it from the attitude of making it work, or making it not work. Yes, some damping might be necessary, such as rubber mounting the cold end of the spring. This would become apparent after having a go.

I wasn't trying to be disparaging, though rereading, what I wrote could certainly come off that way.

You could certainly be correct, and I could be wrong. I'd still consider it a safe wager for me. Wouldn't be the first time hubris has cost me money.

I had a similar problem with a plastic sheet welder, the two pictures show the element hot and cold. The element is 0.006" x 3/16" x 900mm long and doesn't get red hot.

Tony

I used the DC method on a wire to cut glass. You heat and apply a bit of water from a applicator and it fractures along the heated line.

Using this method with a spiral wound curtain wire is used to soften plastic sheet for bending, the spring in the wire allows for the expansion.

Tony

There were many good advises, basically in each reply, which distinguish from each other in feasibility. We use from time to time a worksheet program like Excel or Numbers to create a decision table where we collect facts about each solution and give points for how they match a certain requirement. Points may be weighted to adapt to different scenario.

One thing which I haven't found in posts is used in our house: We used to feed quite high current (up to 250kA peak) thru round current bars of usually 50mm diameter and have to deal with high forces. To reduce these forces from the magnetic field we try to avoid sharp bends in the current path and try to keep the area which spanned by the current bars as small as possible.

We also try to avoid magnetic coupling by changing the orientation of the source of the magnetic field with respect to the receptor.

Applied to the current topic this means: Guide the feeding wires as long as possible in parallel or as twisted wires, before they go to the ends of the hot wire and maybe change the orientation of the transformer's stray field with respect to the hot wire.

Hi,

I will try to eloborate on the set up. The wire is under tension by a spring. This is insulated from the wire by a sliding block. This end of the wire is earthed to the machine frame. The other end of the wire is clamped to an insulated block. The power is fed in this end. The wire is 650mm long. The transformer is mounted in the electrical box, one side of the secondary is to the machine frame and the other side has a 6sq.mm cable to the welding wire. This cable is about 600mm long.

I think from reading all the comments the way to go is with a rectifier and capacitor to smooth the power to the wire.

I can't put any supports along the wire as this would cause cold spots in the wire and so will not weld at these points.

I have tried different spring tensions and this does change the amplitude of the vibrations slightly, but if the spring tension is too high it will break the wire.

I hope I have covered all the questions. Thanks for all your comments.

I wont have time this side of christmas to do the changes, so have a happy christmas and happy new year.

Once I have tried the mods I will let you all know the outcome.

Paul.

To give decent smoothing at 16A you'll need a BIG capacitor - about 0.15F (150,000μF) for 1 volt ripple.

I would also suggest that you add an inductor in series with the heating element to minimize/dampen the current surges.

This can/will also decrease if not eliminate some of the vibrations that are being caused by the AC harmonics.

Hi,

What size inductor? Would a few turns of the feed cable around a ferrite be ok?

Paul.

1. www.mag-inc.com/File%20Library/​Product%20Literature/...

Here is a link to help in calculating the inductor size for a DC application.

More info to follow on this as soon as I can do so.

http://www.rfcafe.com/references/electrical/inductance.htm

Here is a link to an excellent site for identifying proper inductor/reactor size.

Keep in mind that your application will require tuning in order to acheive optimum results.

It is possible to purchase a variable inductor but be sure to purchase one that is rated for the maximum voltage and maximum current values in your circuit.

Many of the "cracker-box" small AC & DC welders use a saturable core reactor/inductor to control the machine current output.

You could try wrapping the conductor (single) that attaches to the heating element around a ferrite 5-10 turns and see if there is an improvement.

I would not expect to see significant changes as the ferrite is more for small signal noise cancellation and not normally used for waveform stabilization.

Thanks for the additional details. This one in particular may suggest a route to a relatively simple solution:

'...This end of the wire is earthed to the machine frame....'

OK, so the return path is following some arbitrary path back through the machine frame...most likely not the path that creates the least magnetic field nor the path that creates a field oriented for least effect on the wire.

I would try supplying a low resistance path with rigid flat copper (possibly aluminum) conductor back to where the transformer is grounded. The path should

1. closely parallel (side on not face on) the conductor that supplies the wire on the other side, when possible.

2. Where it is not possible to closely parallel the supply path, position the conductor to either minimize the field (for instance splitting the return line to run on either side of the supply line) or to orient the field in a direction the flat wire will resist bending.

I would try supplying the low resistance return path, without disconnecting the original connection to the machine frame first...redundancy, safety, liability....

Also the new low resistance return line must be well supported, you don't want it flopping around.

1. Is this a new equipment? If it is, then discuss your problem with the OEM.

2. If it is an existing equipment that worked perfectly in the past, disable supply to

the heating element, run the equipment, observe if vibration still present, post

your feedback.

Although I believe that DC is the easiest answer. I'll propose another:

Support the wire its entire length using asbestos (probably not available) or ceramic. Even if the wire is only attached at the ends, if the backing is touching the wire, the vibrations should be greatly dampened.

Firstly, I have to admit that I have read most posts, but not all, apologies.

I don't believe that the transformer is the problem directly as a good transformer, especially Toroidal ones, emit very little. A shielded box will also improve that, connected to ground, as will moving it frárther away....

What I do believe is that there is a steady magnetic field somewhere near and the AC through the wire is causing the vibration (simple motor theory), in the magnetic field.

A good DC supply should help as mentioned already several times, though if the steady magnetic field remains, the wire will get bowed.....again simple motor theory.....

It may be that various steel parts of the machine have simply become magetized, all on their own, due to sitting in the earth's magnetic field and being vibrated. Thus causing the field.

Turning the whole machine through 180° may nullify the effects for a time, or even just 90° may fix it forever.........

Just a few thoughts.

Apologies to anyone who has already mentioned all of this.....

Have you verified that the problem is electrical and not machine vibration independent from the heating wire?

Yes, If I run the machine with the wire off there is no vibration at all. If the wire is turned on while it is running, as the wire heats up you can see it vibrate.

That test you have made implies to me quite strongly that magnetized parts near to the wire could be causing your problems.

With the machine fully switched off, passing a magnetic compass near to the wire may allow you to identify whether this is true or not as one or the other end of the needle will be strongly attracted.

A TV screen demagnetizer may help with removing the magnetism, though that is only a guess, I have only demagnetized certain magnetic tapes...not machines..

Best of luck.

Hi Paul,

There are two relevant physical phenomena not mentioned so far; they are the earth's magnetic field is part of the motor effect and mechanical resonance of the string. De-tuning the string by altering the tension , also by reducing the stiffness of the acoustic termination.

Modification of the earth's magnetic field can be a cheap and convenient approach, a test to observe this is to shield the wire using a mild steel channel say 10+10+10mm by 0.4 to 1.2mm sheet,angle iron could also suffice. 80 to 99% of string length would make a valid test. If the welder is easily moved try rotating it some 90 degrees about it's vertical axis.

The resonance problem is most conveniently tested by the use of a piece of material (wood OK) pressed against the wire. Sliding the wood along the wire will show resonance modification.

WARNING. Unless the Variac and transformer are massively current rated, the rectifier and capacitor will draw much more RMS (heating) current from the transformer than the wire's current (Schade O H "Analysis of rectifier operation). Fortunately the bridge rectifier, without the smoothing capacitor, will remove the power line frequency current and the 100/120 Hz second harmonic may well not resonate and is only about 20% of the un-rectified AC anyway

Paul, Is it necessary to use the variable and the step-down? Using both (if that is the case) is not terribly efficient.

You could simply rectify the 24VAC (4 diodes) and drive with that - that would be a 100Hz rectified DC which should not resonate with the wire, given that you now have resonation at 50Hz.

Then if necessary you could add some caps to reduce the ripple.

Q for all - can super caps or ultra caps be used in this situation?

What you say is possibly true and well worth testing out but the resonance at say 50Hz, will still be apparent when the supply is at 100 Hz DC, probably/possibly 50% down, assuming that the main resonance was at 50Hz. The cap would help smooth it out if used and lower or even fix the problems.

The problem is called "Harmonics". There are odd and even ones. 100 is an even one to 50 for example.

It is also possible that the wire may resonate even worse at 100Hz DC, as not enough testing has been possible to satisfy the "Gods" of resonance/Harmonics and the wire length as of now.....though I tend to agree with you that most probably a higher frequency is the way to go, at least at first....

A thicker wire (and a higher current) may also help to improve the situation......but may not be acceptable for other reasons.....

Great blog!!!

Andy - not sure I can agree with you on the harmonics. As it appears to be resonating at 50 Hz, as far as I know it could only resonate with odd harmonics. As the even harmonics (first at 100 Hz) will have equal amounts of 'up' and 'down' within a 50 Hz cycle, I can't see how the wire could resonate with a 100 Hz signal.

Unless it is not a mechanical resonance but a 50Hz thermal expansion / contraction, which would depend on temp, thermal mass of wire & radiation, this might also occur at 100Hz but should be somewhat less strong.

Stuart21, I'm pretty sure AndyGermany is right about this.

This may help.

You are assuming (maybe wrongly) that the wire has resonance at 50 Hz., that may or may not be true. The transformer is probably designed for max efficiency at 50 Hz. (most are deigned to work at a single frequency best), so other frequencies, even if available (they are not at this time anyway!), will probably be heavily attenuated.

Just because a wire is vibrated from an AC voltage applied across it at the frequency of the AC, is no proof of resonance at (in this example) 50 Hz...

Making assumptions in science are a sin worse than adultery!

If it were possible to have say a 24 volt VFD applied to the wire at the current required, you could vary the frequency and see exactly which frequency is resonant on the wire, that is when you get a sine wave starting at one end, goes through a complete cycle and ending at the other......

A careful study of Lissajous (spelling?) figures and the like, might increase your understanding of the problem and harmonics.

Just for clarity, 100 Hz is the next strongest even harmonic of 50 Hz. when increasing frequency, as is 25 Hz. when going down in frequency. There are also odd harmonics which should not be ignored either!!! 75 Hz is the next odd one going upwards in frequency for example.

See here:-

http://en.wikipedia.org/wiki/Harmonics_%28electrical_power%29

If P.Mason could tell us just how many "full and/or part cycles" he sees on the wire, we could even make an informative guess as to the basic resonant frequency of the wire. I checked all his posts, he has not mentioned that up to now I believe......

You wrote:-

Unless it is not a mechanical resonance but a 50Hz thermal expansion / contraction, which would depend on temp, thermal mass of wire & radiation, this might also occur at 100Hz but should be somewhat less strong.

I cannot believe this as the thermal inertia of a wire, would normally preclude this happening I feel.

Plus the effects of heating and cooling a wire are mostly noticeable in the length as far as I am aware.....cold = shorter, hot = longer. The wider/narrower effects will be noticeably more difficult to observe correctly. But due to thermal inertia, I do not think that they will be noticeable.

I hope this helps.

OP, P. Mason; Could you try the 4 diode setup shown in this link, & get back to us with how it goes?

Thanks.

http://en.wikipedia.org/wiki/File:Gratz.rectifier.en.svg

Diodes like this should do the job, http://www.diodes.com/datasheets/ds28010.pdf or these http://www.alibaba.com/product-gs/580205953/10A10_10_Amp_1000V_10A_1KV.html

Although current is 16 amp AC, each diode will be conducting half the cycle so 10 Amp should do the job. You could try doubling them if you get failure. (May need to match diodes) Also your power will be down abt 15% so you may need to tweak the variable transformer to get the same DC reading across resistance wire as previous AC reading.

Re: super/ultra caps - I don't think it's practical due to the low voltage rating (typically 3 or 4 volts) of readily-available ones.

'as the wire sometimes touches the product and burns it.'

To prevent direct touching, may I suggest use of very thin Teflon tape wrap, over the heating element. It may prevent burning of the PET.

'...may I suggest use of very thin Teflon tape wrap, over the heating element...'

.

You may, but I un-suggest it right away, to counter balance.

Note the OP stated the wire is red-hot.

When you were suggesting the Teflon tape, did you consider the melting point of PTFE? ...as compared to the temperature of a wire glowing red hot?

.

Putting Teflon on the heating element is ill advised. At best it will lead to a fair amount of time spent cleaning up a mess....but it could be much worse.

Or silicone? Not sure it can get red hot or not though.....

No, No, no.

The OP is welding thermoplastics together.

Silicone will not allow the plastics to bond properly.

Accepted, thanks.

Hi All,

EARTH'S MAGNETIC FIELD

EARTH'S MAGNETIC FIELD

Will someone please refer to my #41 where I pointed out that the 50 Hz force acting on the wire is due to the interaction of the EARTH'S (or a magnetised machine part's) magnetic field with the current's field.

Try magnetically shielding the wire;

or reorientating the machine;

or deflecting/ bending the earth's field with a large ferrous lump (as a test method).

If you reduce that 'motor effect' resonance etc all else becomes detail!

Good luck

'....or reorientating the machine....' .

. Why do you assume the machine was orientated already? .

Also, any suggestions on ways to implement the solition you recommend of 'magnetically shielding the wire? . :-)

Hi Tinac,

The earth's field runs north to south, my first post suggested rotating the machine about it's vertical axis. This should reduce the field and with it the force at 50Hz on the wire.

Since I can only guess the resources/junk near the machine I suggested folding up a sheet steel channel for magnetic shielding. Since this might be easy in some factories it would make a test easy. My suggestion of a length of angle iron is for a shield test with a common junk item (in many works). In both cases the shield should be close to surounding the wire, say 2mm clearance.

Don't you think a piece of ferromagnetic material in close proximity to the electromagnet (wire) will lead to a more severe interaction than if it were merely the Earth'magnetic field? .

.

. To shield something from a magnetic field, you typically have to create a inviting flux path around the object. This typically means surrounding it in something like mumetal or mild steel as you suggested.

.

. But shielding isn' t an option here, since the current in the wire generates the field. .

.

The other part isn't really serious. I was just razzing you about using 'orientated' instead of 'oriented' . .

.

. I'm sure it was 'occidental' (I'm sure I'll be sentenced to the 'punitentiary' for that remark).

Hi Tinac,

The 'magnetically soft' (high permeability) iron is an easier path for for the static (earth's or magnetised structure's) field. The lines of force pass through the 'soft iron' in preference to the air. so fewer lines pass through the air path. This results in less force to vibrate the wire.

And what about the field created by the current in the wire? why would it not interact with the iron?

GA

You do appear to understand the problems a lot better than some others here, thanks. I am on the same wavelength as you.

BUT, wouldn't that field have been there from day 1? So why was the problem not there from day 1 (assuming I have understood the problem fully!)

I think the the Earth's Magnetic Field (hitherto referred to as "EMF"), has poorly magnetized some machine parts after years of usage/vibration. That is causing the problems I believe. Which is why they weren't apparent from day 1.....

What is the length, of your flat red hot wire? You are using radiant heat to weld PET. Any idea as to the temperature, of the flat red hot wire? What is the minimum gap to be maintained, between flat red hot wire & the PET? What are the dimensions of PET? Is it feasible, to increase the level of radiant heat & gap between the flat red hot wire & the PET, to overcome contact between them, in order to avoid burning.

"as the wire sometimes touches the product and burns it"

P., where does the wire touch the product? In the centre of its (wires) length?

Yes, in the center of the wire.

In answer to moving the wire away from the PET, it can be done but the wire then needs to be hotter and therefore burns out quicker. I run it at cherry red, this gives us a reasonable usable time from the wire.

Now I am back from the xmas break I can try the fitting of a rectifyer and let you all know the results.

It is not the earth magnetic field as this happens in all positions of the compass.

I believe that over a long perid of time, the earth's magnetic field has magnetized some machine parts. Till they are demagnetized, you will still get the problem.....

A colour TV screen demagnetizer/degausser may help....Talk to an old TV engineer, they do not need them anymore for modern flat screen TVs.

There seems to be a permanent/DC field in the machine's structure or parts. Ideally a gaussmeter/magnetometer would allow the source to be located, can you borrow/hire one?

Alternatively a small compass (Christmas cracker quality OK) can be used. Since the field has to be treated as random-source the compass will need to be used with it's axis horisontal (and rotated at each test point in the horizontal plane).

Good luck with the rectifier hookup, I guess it will give >95% chance of adequate vibration attenuation. However do not use a filter capacitor, little benefit and the high peak current pulses may burn up the Variac.

Is there a magnetic chuck, permanent magnet motor or magnetic clamp nearby?

You can also try changing the tension and/or the length. If you've got lots of these in your production line, and each one must be independently variable, DC power supplies are going to get expensive.

Placing a heated wire inside a suitably-sized quartz capillary tube may also be something to consider. Not only would the tube stop the vibration, but it would eliminate the need to place the wire under tension; the tube itself supplies the rigidity you need. Another advantage is that if your heated wire is exposed to oxygen, scale will form and slough off from the wire during repeated heating/cooling cycles. This scale could contaminate/discolor your product should it come in contact. Not so with quartz. Food for thought.