copper braid/screen/shield dipped in solder bolted on star earth ground?

[coppercone2]

So this miller has a big star earth ground, and they use 1 brass stud inside of aluminum bracket to connect a whole bunch of grounds.

They have 4 ring terminals and 3 pieces of solder braid dipped in solder on one screw.

I am thinking that solder braid/copper screen dipped in solder is not a good thing to put on the main earth ground?? I was thinking I should solder a strip of copper to that, so that there is no solder in the bolt tension.

Surprised to see that from miller in a high end welder.. Is it for some reason acceptable to do this?

[Benta]

I'm not sure what a "miller" is. Do you mean a mill, like for metal cutting/milling?
Anyway, putting tin-based solder under mechanical stress is never a good idea. This includes pressure.
Tin solder cold flows, meaning it will deform over time and lose the connection. This is dangerous from an electrical point of view, but also a fire hazard.
The same thing happens for aluminium, by the way, with the added fire hazard of oxidization. This makes me a bit concerned about your "aluminium bracket".

The correct way is using crimped ring- or fork terminals for the bolt.

[coppercone2]

miller electric dynasty 200 welding machine

The machine is built like I beam in a box, with bracket in middle that PCB and IGBT go on both sides. The front ground for several PCB is a brass stud embedded in this bracket, which has 7 connections consisting of 3 braids and 4 ring crimps.

it was considered a world class TIG welding machine

I thought maybe they used braze, but I see there is rosin flux residue on the braid, meaning it is in fact soldered.

[Ian.M]

Yeah, that's *NOT* *GOOD* and possibly illegal for a safety ground as solder-tinned stranded wire will cold flow and deform under clamping pressure, resulting in a bad connection, as Benta has already mentioned.   It could be converted to solderable ring terminals for functional grounding but if any of those braids is a safety ground, all solder-contaminated braid must be cut back and crimp ring terminals properly fitted.

[bill_c]

Does it show any signs of poor connection or burnt areas?  I'm sure they tested it during the design phase, but there is also a possibility it was designed that way to fail out of warranty.

[coppercone2]

well I don't know what the damage is from, the input rectifier brick (skiip module) was cracked and 3 PCB had burned parts on them, still in progress of working it out. Looks like HV got on the 15V rail.

I did read on a forum somewhere last year that someone was saying service techs were doing something with the ground on that unit, but I don't know what it is. IIRC the unit stops working and the service tech tightens a bolt some where, it could be that one. I learned no more then 3 ring terminals per stud from someone, claims it has vibration issues (like junkers test) when there is too many ring terminals on one bolt. I was surprised to see 7 connections there when I took it apart lol

I don't think its the main safety ground, but I need to check, but anyway it seems like bad practice, since that solder creep could cause a cascade failure for all those ground connections on that bolt? And fuck up EMI behavior? But why am I not surprised from a company charging like 25x parts on a PCB lol. Too cheap for 2$ of lugs... perhaps the money had to go on rock and roll music for the advertisement. 

The one of the ground braids is from MCU board, which connects to keypad board. Not sure if it has another ground, the keypad is membrane, so I guess it might be isolated even without ground by insulation but hmm, I need to see where the top cover grounds to. With the amount of soot/metal particles I see in those units from being around grinders and welding fumes, it might very well be a safety hazard.

Ok, main safety ground goes to another lug on the other side of the device, I guess those grounds are just filter grounds.. but again IDK what happens if you fuck up the ground connection on a 800V bus filter cap, maybe it can make the unit go haywire like a timebomb.


I don't know what the deal is with the front panel membrane thing, does that need to be earth ground referenced or do they trust the membrane in front of the unit? Because that PCB does have a separate ground wire, which goes to the same bolt as the ground strap. I guess the strap is RF and the ground wire is safety (it comes from the cable harness to the bolt (for that PCB), but they both connect to the same stud, which has two pieces of soldered braid in the bolt.

SO I am gonna go out on a limb and say the ground for the front panel, for whatever reason has a braid and wire going to the same bolt, so that seems marginal for operator touch safety?

[Ian.M]

Its a safety ground if any of the braids or heavy ground wires link exterior metal panels, even though the main safety ground stud that has the incoming ground on it is elsewhere.  However from your description it does sound like its primarily signal/EMI shielding grounding.

[coppercone2]

yeah when I disconnect that bolt, the display board and the processor board behind it are not earthed. The display board has switches on it that a membrane is infront of, and I think there is some metal or something inside the plastic. It attaches to I guess a metal plate inside of the front chassis cover plastic piece with 4 rusty steel studs, that bolt on to processor PCB earth plane. I think they coulda used brass or stainless there too, but that plate is not some how connected to the grounded superstructure, its only the molex wire + soldered braid

I can't really evaluate further because I am not familiar with the standards exactly, but it seems that HV could go on the display cable harness I think (control signals). I don't have schematics so I can't really say how well isolated it is or whatever, but I think they should have put solid metal on the straps to stabilize that structure. They do have a green/yellow wire connected to a molex pin, is earth grounding through a molex pin OK anyway ? (not that it matters since it goes onto the same bolt with the soldered braid)

Let me say, I feel alot better about having such problems repairing this unit when I find these sort of problems, and maybe its good that I did not get a working one I never took apart that I would unconditionally trust lol.... how about keyed grounded panel switches instead of membrane shit on shoddy earth for $4000 lol

[Ian.M]

A loose ground could cause all sorts of havoc, including HV getting onto control signals and low voltage power rails if its the ground return for any HV circuit.

[coppercone2]

Well, it was tight when I first took the unit apart, but who the fuck knows what happened or if someone retightened after the problem occurred. But I seriously think amateur hour because its like the first thing you learn in any factory, aka don't put solder on the fucking wires unless they are to be soldered. I never saw this done before, so I made this thread because I thought maybe there was something about flat braid that makes is acceptable to bolt down (it seems unlikely and this thread seems confident). Sucks if you try to learn from taking stuff apart.

[coppercone2]

How does the damage occur from loose ground? I just know that its not acceptable. Is it because the filter caps and MOV are not filtering/burning and you get destructive HV HF/line harmonics on the bus going through the converters and shit? (i have always been paid NOT to find out)

You never learn why because its too expensive and unsafe to do so (they always want a good safety earth for safety and insurance, so its usually #1 priority to get it right)

[Doctorandus_P]

I had a very quick look at those Miller Welders, and just their price point and the fact that they seem to be rented gives me more confidence in them than in a "coppercone2".
So for the moment I just assume they know what they are doing and you're missing some detail.
Maybe a few pictures would help convince me to swing to your side.

[Ian.M]

The loose ground HV problem is when a power ground to the HV circuit develops an intermittent connection together with adjacent grounds on the same stud/bolt.  As the HV collapses due to the  normal load on it, the loose ground can get pushed to the same voltage as the HT but with opposite polarity, as there's nowhere for the load's return current to go.  If the ground to another module is touching it there's a large risk the other module gets fried.

[abquke]

If you look for interference, you'll find it.

If you look for design flaws, you'll find them.

[CatalinaWOW]

Miller is one of the two most reputable brands here in the US so I see a few possibilities.

1.  If the machine was purchased used there could easily have been some field "improvements".  It would be interesting to note if such "improvements" would be caught and corrected at an authorized service center.

2.  Solder flow under load is only a problem if the solder is carrying load.  Is it possible that there is some hidden mechanism in the braid that carries the load and makes the connection?  A welded in washer for example.  Such a situation might not be obvious.  And it might be obviated by possibility 1.

3.  Miller has fallen as so many have in the face of low price competition.

[coppercone2]

ok doctor, you can believe whatever corporate bs you want

here is the manual picture, its a piece of braid with solder on the end that goes over a bolt

why would I make this shit up? there is no hidden mechanism, there should be no 'mayonnaise' in the 'sandwich'

there is a green yellow ground wire in the RC1 connector that goes over the same bolt with a ring connector. there is a big dent in the braid where the other ring terminals were on top of

I don't see the outline of solder for any kind of hidden washer (if there was a washer, it should be clean with a little groove around where the cut hole is in the braid, there is none, just smushed flat solder and copper wire

[coppercone2]

in fact, I looked inside of the hole sideways and I see that its been clearly put in a solder pot then punch pressed or something, you can see the embedded wires on the perimeter surface of the hole (like 'damascus' wire rope steel or something), I am not bringing a microscope into the workshop to try to photograph this. If you have one you can look at it with a loupe and its plainly visible. All the braids are terminated in the same way.

I am telling you it looks factory original and also improper

[coppercone2]

and here is a ebay picture of another welder with the same thing
https://i.ebayimg.com/images/g/VwMAAOSwuWVikGqE/s-l1600.jpg
(lower right near the front of the unit)

as you can see there is no termination like a real ground strap has, which would look like
https://encrypted-tbn0.gstatic.com/shopping?q=tbn:ANd9GcQZnORxFs8wAJkc8FYQ0667PUfJThyjQ7nFyv8sEIAuKGNqpDCWbrHSP8eJNixuLvGvk2xKc6zaoj1NcbmMT8LQsGsUeEd0bATYiavf437T&usqp=CAE

[mag_therm]

Braided copper grounding straps with solder tinned ends, drilled, have been around for 100+ years.
One example is automotive battery ground straps, commonly used up until the 1960's I recall.
Other examples I have seen are industrial and ship board use inside enclosures for RF and electrical grounding of panels, subchassis etc.
On navy ships the electronic racks used to be on rubber shockmounts, bypassed by these straps.

They are still commonly available, search "braided grounding strap"

Solder, as we know, has poor creep and fatigue properties.
Hovever I think soldered braid is bit like polyester fiberglass.
If the coppper braid % fill is much higher than the solder that wicked in, the properties will be more like copper.

[Doctorandus_P]

I'm not saying they did not do it the way you wrote but I did assume you made a simple (and often correct) assumption.

Look at it this way:
If you control the amount of solder so it's not excessive and combine that with enough force to deform the tinned braid enough then the solder will create an air tight seal but it has no further room to creep to, so creep will not be a factor anymore.

So I'm on the same line as mag_therm. If there is enough copper in between, then the tin only fills the extra gaps.

It's true that tinning wires that are clamped is considered bad because of creep, but that does not mean that this creep will always occur with clamped wires.

[Doctorandus_P]

This wire does not look like it's put in a solder pot.
It looks like that the individual strands are tinned or nickel plated before they are braided together.
It also looks like it is a very low current connection.
I've been looking at that picture a bit, and it does not look like that braid is carrying any current at all. The star point seems to be just for safety GND and EMC grounding.

There is no indication anything in that area is overheated or damaged in any way.
I would say the black PCB's on the left side are far more suspect

[coppercone2]

it has solder in it, that end piece is totally solid (even on the corners that did not experience pressure), it feels like used solder wick. I would say the problem with that welder is not mine because its an ebay picture of someone elses welder

How do you control the amount of solder that is being chaotically wicked up into the braid?

I just did a test
1) i clearly see a solder boundary
2) in the middle of the braid the metal pick is easily able to get under one of the weaves of wire
3) in the end of the braid where the bolt is, even in the area that was not under bolt pressure, the sharp pick behaves exactly like if was trying to penetrate a solid solder filled area (aka there is no place for the pick to get between the strands)
4) it bends like a solid piece of metal
5) its really obvious that there is a solder boundary

I never saw any standard that gives an acceptable amount of solder in a ground connection.

[mag_therm]

I should add to my previous post about historic use of solder dipped braid.
In the USA, for fixed wiring covered by National Electric Code,
soldering is still allowed provided the joint wires are twisted mechanically first.

For some ground conductors (I assume ones that might carry a switchboard full fault currnent,
as protected by the incoming overcurrent protection),
soldering is not permitted. Such conductors would be subject to adiabatic heating during a severe ground fault
The joints are often by listed clamps and approved exothermic welding.

I don't have the NEC code here and my description above is approximate and should be confirmed
by reference to the actual code.
Maybe some on here have the code and can give better info.