Measuring wire speed

[malcolm]

With the ideas on here for modifications to SIP welders it would be fun to do some objective measurements during welding before and after changes.

I've got an acoustic analyser that could measure and chart voltages. Could tap off the voltage feed to the torch to see how stable it is, though might have to do something to limit the high voltage spikes from the coils.

For the wire speed I'm thinking of measuring voltage from a little DC motor running against the wire. Does that sound like a sensible way to do things? Is there a type of motor I could look for that turns very easilly and produces a smooth voltage output?

 

[rtbcomp]

For the wire speed I'm thinking of measuring voltage from a little DC motor running against the wire. Does that sound like a sensible way to do things? Is there a type of motor I could look for that turns very easilly and produces a smooth voltage output?

There are DC tachometers on the market, but in my experience they are expensive and unreliable. You could try a permanent magnet motor but you'd need to calibrate it. I'd drive it from the rollers rather than the wire

A lot of DVMs will measure frequency, so you could adapt that mouse method (Sorry I can't find the thread); get a conventional mouse and use the photocell to produce a pulsed output. You'd need to do some sums regarding roller diameter, pulses per revolution, ratio between roller & sensor diameters &c.

 

[Unclefista]

Bicycle speedo ?

Glue the magnets on to the drive roller and calibrate.

They're only £2.99 from Netto

 

[rtbcomp]

Bicycle speedo ?

Glue the magnets on to the drive roller and calibrate.

They're only £2.99 from Netto

Nice one - have they got them in at the moment?

 

[malcolm]

Bicycle stuff is something I hadn't thought of - can you get a DC signal out of them? Keen to record DC volts against time to see the fluctuations during welding - I'm convinced that's the difference between a lovely arc and a spattery one.

 

[rtbcomp]

..... can you get a DC signal out of them? Keen to record DC volts against time to see the fluctuations during welding - I'm convinced that's the difference between a lovely arc and a spattery one.

Probably not, these days all the gubbins will be in a single IC, probably surface mounted with no hope of soldering a wire on!

You could measure the current in the motor if it runs off steady DC, but this would give a qualitative rather than quantitative measure.

 

[Unclefista]

Nice one - have they got them in at the moment?

Errr, I'm not sure, but it's one of the things they regularly get in. They're only cheap wherever you go, the one on my car was £6 but that's only 'cos it's a "branded" one that shows over 99mph, shouldn't be an issue with wire speed

 

[malcolm]

Come up with half an idea - greyshirtguy's mouse wheel would be good for it's lack of inertia. That could possibly operate on either the wire or the wire feed roller. Plug the output of that into a frequency to voltage converter chip (somehow) and hey presto.

While I'm thinking - is there an obvious way of clipping the output voltage of the welder to get rid of any nasty spikes that might blow my measuring equipment up?

 

[rtbcomp]

Some form of opto-isolator would be the safest.

 

[malcolm]

Some form of opto-isolator would be the safest.

Brilliant - I've looked it up and it sounds ideal. I'd not heard of those things before.

 

[irondarren]

Output voltage will vary on a welder depends on the length of the arc. you put a volt meter on your welder malcolm you will realise how unsteady there hand is and how bad you are at keeping a constant arc lol. Well thats what ive been finding

 

[Drains]

Output voltage will vary on a welder depends on the length of the arc. you put a volt meter on your welder malcolm you will realise how unsteady there hand is and how bad you are at keeping a constant arc lol. Well thats what ive been finding

Surely it should be the current varying, not the voltage?

At least, on my set that's what happens. If I point the gun at a piece of plate and move the gun near and away etc, it's only the current that changes. The voltage stays stock still.

 

[rtbcomp]

Brilliant - I've looked it up and it sounds ideal. I'd not heard of those things before.

Thinking about it, Opto-Isolators may not give an output as a function of input, I think they may be more on or off. I retired from the wonderful world of electronics a few years ago, Shenion may have a view on the subject?

 

[rtbcomp]

Surely it should be the current varying, not the voltage?

At least, on my set that's what happens. If I point the gun at a piece of plate and move the gun near and away etc, it's only the current that changes. The voltage stays stock still.

In that case measure the voltage between the torch and the machine, ie the voltage drop accross the wire in the torch lead is proprtio0nal to the welding current.

Easier still, accross the choke if one is fitted?

 

[Luke]

you can get a decent speed measurement off a small optical encoder.


Last year i did a project involving on board processing on a scalextric car, the encoder was accurate to <10mm over 6000mm. that was a 15 hole wheel driven off the back of the motor, which drove the wheels through a 1:3 gearbox, giving 45 pulses per wheel rotation, or a resolution in the region of 1.6 mm.