EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: jwhitmore on November 21, 2016, 03:56:51 pm
-
I've asked a few questions on here before as I'm working on a CAN Bus project on a vehicle for my brother. He asked me the other day if the electronics I'm adding would be OK when welding on the vehicle. Of course I said yes ;-)
I've heard a lot of theories over the years about welding on vehicles and never given it much thought, now I guess I should. I've heard people say that you have to disconnect the positive battery terminal or you'll fry electronics when you're welding. I've heard people say you can't do that any more as modern electronics in a vehicle don't have persistent memory, so if you pull the battery positive terminal you've lost all the settings, so it's not safe to weld at all.
I've not got a clue but I'd have thought that during welding the currents involved would just get conducted to the shortest route to ground, the welder ground. Why welding would effect vehicle electronics I'm not sure. Unless you didn't attach the welder ground connection to the area being welded. In that case I'm not sure that there's any current flow at all. Potentially dangerous if somebody touches the right two parts but a threat to electronics?
-
The simple answer is if the arc welder does not complete the circuit to the grounding clamp it won't arc, but you probably know this. From a pragmatic view point, my friend who owns a welding shop never bothers to disconnect anything when welding mufflers, tailpipes, etc. If it would be a problem he would hear about it from his customers.
-
That's what I thought. I don't imagine that it's an issue, but I'm no expert. Don't know why all those electrons would head to embedded electronics. On the other hand if you did open the wireless radio and stick an Arc welder in there things could get interesting. ;-)
-
The simple answer is if the arc welder does not complete the circuit to the grounding clamp it won't arc, but you probably know this. From a pragmatic view point, my friend who owns a welding shop never bothers to disconnect anything when welding mufflers, tailpipes, etc. If it would be a problem he would hear about it from his customers.
Very bad advice when it comes to modern cars. The magnetic fields are large enough to induce serious spikes and currents into circuitry. You really wouldn't be the first the fry the electronics on a car or a motorbike. A good car mechanic puts a welding surge protector across the battery before welding to a car.
-
Arc welders are typically under 50V. Electronics used in cars are supposed to be automotive qualified which means they're tolerant to 65V. Car electrical systems are notorious for large voltage spikes due to liberal transient response in the charging system and large inductive load dumps like the starter motor. So most of the time an arc welder is not operating above voltage tolerance of the car's electronics to begin with. Even if that's not the case there would have to be a big voltage gradient between the arc welder's ground and battery ground for a potential to occur that might damage electronics. Pretty unlikely that can happen since it would defeat the current required for the arc welder to operate. Now when it comes to RF stuff there may be other considerations, not sure about that one.
-
"Even if that's not the case there would have to be a big voltage gradient between the arc welder's ground and battery ground for a potential to occur that might damage electronics."
Ah now that is food for thought! Thanks for that gem, it's certainly worth looking into.
-
"Even if that's not the case there would have to be a big voltage gradient between the arc welder's ground and battery ground for a potential to occur that might damage electronics."
Ah now that is food for thought! Thanks for that gem, it's certainly worth looking into.
:palm: It is not (all) about the voltage on the welder but the energy in the magnetic and electric fields emitted from the arc and the cables! Also some weld systems use HF/high voltage to start the arc more easely.
-
OK Magnetic and Electric fields can be a problem but most, if not all, automotive electronic devices are normally in metal boxes. I assumed that was to provide some sort of shielding to the electronics?
-
OK Magnetic and Electric fields can be a problem but most, if not all, automotive electronic devices are normally in metal boxes. I assumed that was to provide some sort of shielding to the electronics?
There is tons of unshielded wiring in a car and not all electronics are in metal boxes. The electronics are designed to withstand what happens during normal operation of a car and to meet the requirements but absolutely nothing more than that.
-
There is tons of unshielded wiring in a car and not all electronics are in metal boxes. The electronics are designed to withstand what happens during normal operation of a car and to meet the requirements but absolutely nothing more than that.
[/quote]
Cars have insulated wires. Most of the welding takes place on exhaust systems and chasis parts far away from the delicate parts.
-
Even without the special HF start mentioned, voltages involved in welding are in range of kilovolts to tens of kilovolts. The fact that the secondary winding produces only 50V is meaningless, when the whole concept of arc welding (stick or MIG/MAG) is completely based on the large inductor (separate unit, or built as leakage inductance in the transformer) on the output of the said transformer, and when this inductance is shorted, it creates a "boost converter", allowing the arc to initiate through up to several mm of air.
It's not easy to exactly define the actual voltages during welding, but it's way more than "50V", and they are not only some small side-effect transients; there is a lot of energy there.
-
If the machine is set up for MIG welding using shielding gas, AKA GMAW "Gas Metal Arc Welding", then it's likely configured for DCEN "Direct Current Electrode Negative". If it uses flux-cored wire then it's likely set up for DCEP "Direct Current Electrode Positive".
Let's say it's set up for DCEN. The ground clamp will be connected to some place that has a good ground, which on a car isn't the easiest thing to find. For instance, paint, rust, etc. will be in the way. If it's a truck with a frame, there will be rubber mounts to isolate vibration. In either case the rest of the body or cab will be spot welded together, so it isn't like there is a continuous seam that is conductive, particularly with thin-gage steel. Car companies seem to like local grounds connected with a ring terminal to some place on the body.
So where ever the welder ground is located, the more negative-voltage high-current work electrode is bouncing what was good enough as ground through 16 AWG or whatever local control electronics grounds through the steel cab. At 50 to 150 amps of work current who knows what kind of ground loops and other circulating current could be induced, particularly if the control electronics get involved in the loops. One could see local voltages that are significantly negative in this case. If it's set up as DCEP, then local voltages could be significantly positive. Note that this is a very low-impedance source, so if it does make current flow through the circuit, there won't be any limitation except for the component itself.
The actual impact depends on how well the welding currents are isolated to a particular part of the car, and how well isolated from that the control electronics are located. This is why welding the exhaust together usually isn't a problem, because there is an electrical connection really only where it comes out of the manifold to the engine block. The welder ground would be somewhere over there as they work their way down the pipe, which is electrically isolated from the body by the isolation mounts, along with any undercoating and rust.
It's sort of like the good general rule of keeping the loop area small to prevent electrical interference problems. That is, don't put the ground on one corner of the car while welding on the opposite corner.
-
I've heard people say that you have to disconnect the positive battery terminal or you'll fry electronics when you're welding. I've heard people say you can't do that any more as modern electronics in a vehicle don't have persistent memory, so if you pull the battery positive terminal you've lost all the settings, so it's not safe to weld at all.
current flow at all. Potentially dangerous if somebody touches the right two parts but a threat to electronics?
It's just ridiculous. It would mean that I can't disconnect my car battery from the car to connect it to battery charger at home. Only thing that tends to go away is radio settings, because apparently it's too much of an expense to put 0.5$ flash chip in 20000$ car.
-
The correct answer is to read the car manufacturers guide and follow their recommendations. Usually it will be in an upfitter or repair guide.
-
Right or wrong, but I've personally MIG/TIG and plasma cut on many dozens of vehicles (including late model injected/diesel), never once an issue. No special precautions, battery intact.
-
Even without the special HF start mentioned, voltages involved in welding are in range of kilovolts to tens of kilovolts. The fact that the secondary winding produces only 50V is meaningless, when the whole concept of arc welding (stick or MIG/MAG) is completely based on the large inductor (separate unit, or built as leakage inductance in the transformer) on the output of the said transformer, and when this inductance is shorted, it creates a "boost converter", allowing the arc to initiate through up to several mm of air.
It's not easy to exactly define the actual voltages during welding, but it's way more than "50V", and they are not only some small side-effect transients; there is a lot of energy there.
A Constant Current welding power supply (stick, TIG/GTAW) will have an Open Circuit Voltage around 80V. You have to scratch start them, i.e. you must physically connect the circuit (a TIG/GTAW might also use a HF start) . A Constant Voltage welding power supply (MIG/GMAW) will have an OCV around 50V and the welding wire must contact the work piece to initiate the arc. A CC plasma cutting power supply will have an OCV in the 300V range. The arc might be started by a set of electrodes in the torch being separated by air pressure or by a High Voltage/High Frequency AC signal coupled to the leads. This HF start is on the order or 2kv or so, there is nothing at all in the 10's of kv range and the arc cannot be initiated through air without a HF start, that is just simple physics.
Once the arc is started what happens depends on the welding process. For a TIG/GTAW/Plasma the arc is maintained through the plasma stream and filler material is added cold, i.e. it is not part of the circuit. For GMAW/MIG the wired is your electrode, the most common mode is 'shot arc' welding (less than 30v approx.) where the wire contact the work piece creating a short circuit, heats up and melts opening the circuit and this repeats. There is also a 'spray arc' mode which is done at mere than 30v and higher currents. The arc is started as in short arc med but once initiated the plasma stream is maintained and the filler is transported through the plasma stream.
Practically speaking place the welders grounding clamp close to where you are welding, which is good welding practice anyhow, this way you are not sending welding currents all day over vehicle. Welding is done on cars every day with the battery cables connected, again clamp close to where you are welding. You are not trying to operate the car whilst welding. If I were to use a power supply with HF start I might disconnect the battery but I'm not sure it is needed, primarily to stop the cars computer systems to keep them being confused by electrical noise.
Trying to build control systems that will work while welding/cutting is a different matter and can be challenging. The welding process generates a lot of electrical noise and HF start is problematic (old HF start systems were spark gap transmitters).
-
As stated previously, on modern cars, disconnecting the positive lead comes with a set of problems, such as error codes appearing, electric windows won't close/open properly, etc.
Car manufacturers recommend (stated in service manuals) that if welding is to be carried out on a car, do not disconnect the positive of the car. But they do make a point, by saying to disconnect the fuse for the ABS/ESP braking system. Most likely because the sensors are more sensitive to the magnetic fields welders create.
But if the cars is old enough not to have ABS, the go right ahead and weld to you heart's content, with the battery connected
-
To be honest:
I don´t know one single car manufacturer who not recommends to disable the car battery, if welding is necessary.
If you don´t know how to set the window regulators, then perhaps you should not work on cars. (Or, at least, RTFM!)
As an example what a car manufacturer says: Here is a screencopy from the Ford TIS (Technical Information System) regarding to welding.
(https://www.eevblog.com/forum/projects/vehicle-electronics-and-welding/?action=dlattach;attach=272475;image)
I am curious: Which manufacturer recommends to disconnect the positive lead or simply to pull some fuses?
Are there screencopies or scans available? Then show us some, please.
-
I am curious: Which manufacturer recommends to disconnect the positive lead or simply to pull some fuses?
You NEVER disconnect the Positive lead on a car battery, there is far too much danger of the tool shorting to metalwork while you are doing it. That's why all car manufacturers specify disconnecting the negative lead. ;)
-
With the battery one usually disconnects the negative side first. For welding one might leave the positive side connected. Only disabling some fuses might be enough - there could be a few hidden somewhere.
The problem during welding is two fold.
One is possible ground loops. So car electronics might see a relatively small DC voltage over lines connected to ground at different points of the car.
The second is higher frequency noise and spikes - arc welding produces a hell lot of EMI. Car electronics is supposed to be rather well protected but with RF noise and spikes you never really know where they build up and cause resonances. Most of the time nothing bad happens, but it can blow parts worst case. Not having the electronics powered reduced the chances for damage a lot.
I have seen an arc (not from welding but an arc lamp) to bring down several computers about 10 m away (different fuse) and damage one that was closer by.
-
Last time I had something welded on my car the mechanic put a surge protector across the battery. This makes the most sense to me since that is the likely for any spikes to end up through ESD protection diodes in the semiconductors. A car battery doesn't have a particulary low internal resistance. Removing the battery may be advised because of hydrogen gas but still I'd connect the surge protector across the terminals when welding.
Also removing the positive battery terminal should never be done before the negative is disconnected.
-
Friends car had the RCA cables to the amp in the back and the head deck toasted when getting some exhaust work done.
-
I've seen 50+ toyota electronic speedometers for repair, and the common thoughts when they dropped them off, was either welding on the vehicle, or jump starting backwards,
In this case, they are 1990-2004 era speedos, and they have next to no protection on them, detonating a 6.3V cap that was hooked up in a really stupid way.
-
Having fried an alternator or two and regulator modules too when welding on vehicles I've been of the school of thought to disconnect the battery and alternator. But that was yesterday, today's designs are more robust and component voltage headroom has been increased substantially and higher voltage devices have become much cheaper and more reliable.
These days just ensure your welding Earth/return is as close as possible to the welding IF there's no manufacturers guideline for welding repairs.
-
Thanks to everybody for taking the time to reply. What I've learned is that this is a not insignificant topic. There were a few posts on the line of RTFM, which are fair enough, but missed the main reason for my question. That might well be may fault for not being clear in my communication.
I'm designing the electronics to go into a 20 year old tractor. There are no electronic windows, no ABS no engine management as such. There are currently computers, based on the 64HC11 micro-controller with no persistent memory so battery is never disconnected, for any reason what so ever, or computers forget that they're in a tractor and what it is they do. So whilst there's no electric windows and ABS there is an electronic gear box with 5 clutches, and 4 range solenoids. That stuff has never given a problem when welding and yes we'd always have the welders as close as possible to the area being welded.
So to clarify the reason for this post I'm more of an embedded SW head. Hardware I sort of get the basics, enough to put a SMPS and a PIC processor onto a PCB anyhow. So when the brother asks me if my PCB, which is going into the tractor, will cope with welding I wonder will it? How would I know? How do I ensure that it is resistant to that sort of electrical activity? I can't consult the manual for the electronics as it's me that's designing them! So as I say it's obviously a huge topic all I can do is pop it in a metal box, wrap the box in tin foil then wrap it in an insulating material and then another layer of tin foil. ;-) I should also point out that I'm creating a CAN Bus network of maybe 10 nodes, I might go with the remove the fuse solution at well. My PIC will have persistent memory attached, so I've no problem removing power, as opposed to the existing computer systems.
Once again thanks for all your answers which were very educational !
In summary metal box, tin foil, neoprene, tin foil!
-
The main issue is ground potential difference. If you disconnect the CAN bus from any sensor nodes on or near whatever you are welding, and also disconnect it at the controller node, and your electronics power input circuits are designed to handle full repeated ISO 7637-2 load dump surges, you shouldn't have a problem.