My volt meter melted when measuring btween exposed metalwork

[martinxxxxxx]

Hi All

I was called to a printing company to look at a 3phase seam welding machine as users were reporting getting shocks from it.

I has a 4mm earth bond to the movable seam section, but no bond on the main table.

So I decided to take a 10mm bond from the connection box on the wall and measure the voltage between it and the metalwork of the machine.

I connected black crock clip to 10mm cable going back to the wall mounted isolator switch earth bonding point and the the red clip to the exposed metal work of the machine.

When the machine was used to make a weld. My connected meter went up in smoke after lighting up like firework.

The machine is a
Digicon Micro Systems Seam welding machine for welding fabrics.

I have lock off the machine and said it is not safe to use.


Any ideas

 

[RF Lighting]

What sort of meter was it? What range was it set on?

 

[Spark123]

Does it use a high current at ELV to create heat?
I imagine you were using the current range when it went wrong!

 

[martinxxxxxx]

I was on VAC 600v range on cheap tenby clamp meter. CATII rated.

I think machine uses HF to create weld.

I need someone who can check machine down here in hampshire.

M

 

[bernardgreen]

Unless it was very poor quality meter even on the amps range it internal fuse should have ruptured before the meter became a fire ball. Unless when the internal fuse blew on high current there was then a high enough voltage to create an arc inside across the fuse holder.

To have that much energy between the machine chassis and an earth point is a definate and serious fault. I would suspect a phase to chassis short that may have taken out a in-adequate earth conductor leaving the chassis un-earthed and live. That suggests that either the original earth ( if fitted ) was too small to operate the over current trips and/or the safety devices were not working, not fitted or by-passed.

 

[bernardgreen]

Have just read that it is RF heating.

You need someone who knows about RF "plumbing" as RF energy does not have the same rules as energy at 50Hz. It is more a black art than a science.

 

[ricicle]

Most DVMs will not read voltages at frequencies above 50hZ anyway.

 

[JohnW2]

Most DVMs will not read voltages at frequencies above 50hZ anyway.

That's not true - at least, not as a generalisation. The specs of many DVMs indicate that they maintain their accuracy at 50Hz up to 1kHz or thereabouts, and most will give readings (albeit probably not accurate) at frequencies much higher than that.

I've just tried with a couple of cheapo DVMs and they both gave fairly credible (and very similar) voltage readings for the output of a signal generator at 100kHz. It's not really surprising, since most of them simply rectify the AC and then measure the resultant DC.

Kind Regards, John

 

[ricicle]

All the Fluke multimeters we ever used in the factory I worked in would never read invertor output voltages at 100 Hz !

 

[THRIPSTER]

I had a meter like that once - it was made by Breville.

 

[ricicle]

 

[martinxxxxxx]

THIS machine did it

it is a FAB-003

 

[JohnW2]

All the Fluke multimeters we ever used in the factory I worked in would never read invertor output voltages at 100 Hz !

They must have been a previous generation of Flukes - although I still think one would have, even then, had to try quite hard to design a voltmeter that couldn't cope with 100 Hz! The specs of current Fluke multimeters mainly look like this:

Untitled

• JohnW2
• 14 Sep 2011
• 1

Kind Regards, John.

 

[ricicle]

Yeah - I suppose it was about 10 years ago !

 

[martinxxxxxx]

The machine welds at 27MHz, but what is that doing on the casing of the machine?

I had thoughts about putting the 10mm bond on the casing, then thought. Nope I'm definitely not competent to work on this machine, so told the owner, who said no-one is and asked what the worse could happen is - so I told him that there could be a dead body on his factory floor and I am advising he locks it off until I find him someone to fix it.

Martin