High Current path routing issue

[satiz]

Hi All,

In my design there is a high current path from Relay to output connector. Current rating is 15 to 20 Amps max and voltage is 230 Volt. how can I route this trace ? and what are the things need to take care while routing high current path. please give some suggestion.

 

[KlausST]

Hi,

What can we say without more information?
Short, wide, a lot of clearance.

There are safety regulation for each country. You have to keep on them.

Klaus

 

[Relayer]

Hello satiz,
What program are you using to create your design? Eagle, Altium, Mentor... etc?
As KlausST has suggested: "Short, wide, a lot of clearance" and "safety regulation for each country".
Here is a calculator you can use to determine the width of your track (trace):
https://www.4pcb.com/trace-width-calculator.html
You need to determine the copper thickness you're going to use in your design. This you can get from your PCB supplier. Use this for your initial calculation.
A word of warning, high AC voltage attached to PCB's is not really recommended, but if you have to, you need to ensure the AC carrying tracks are fairly spaced away from other tracks not associated with any AC coupling.
Do NOT bury this track, other than on an outer layer.
I hope the above helps.
Regards,
Relayer

 

[FvM]

You'll define net classes with associated design rules for minimum spacing according to required isolation voltage and minimum trace width according to rated current. Need to consider overvoltage category, environmental conditions and basic/reinforced insulation requirements.Then just route the connections.

 

[gary22pcb]

Short, Wide and Thick. My company commonly uses 2, 4, and 12 ounce thick copper for high current and heat dissipation. There are several calculators out there on the web.

 

[Mattylad]

calculation.
Do NOT bury this track, other than on an outer layer.

Can I ask why not?

 

[marce]

230V is not high voltage... it is quite common to find this voltage on PCBs.
Use the Saturn PCB toolkit it is IPC-2152 compliant
https://www.saturnpcb.com/pcb_toolkit.htm
the one linked to by Relayer is out of date.
Use a copper pour for the connection and if it is a multi layer board you can bury the pour, makes things safer.
What you need to determine are:
Allowable track temp rise above ambient.
Copper weight for the design.
Number of layers.
Are planes present.
What type of equipment is it.

BASIC RULES for spacing can be found in IPC-2221 table 6....
https://www.smpspowersupply.com/ipc2221pcbclearance.html

Provide the above information for a more qualified answer from us please.....

 

[schmitt trigger]

Good websites!
Thanks for sharing.

 

[Relayer]

Can I ask why not?

Mainly due to heat dissipation. Though if the trace is big enough and no plated holes pass near it, it would be reasonably safe.
Regards,
Relayer

 

[Warpspeed]

There is another less obvious reason for keeping high current tracks on the outer layers.
If it ever suffers a catastrophic overload the track may simply explode.

If its on the outside, you can clean away the debris and mess, and run a fat wire instead of the original track.

If its buried on an inner layer, the board may be unrepairable and the whole thing becomes a throw away item.The gas generated, and molten copper effectively blows the board apart, shorting out many other inner layer tracks. The result can be quite spectacular.

Not too bad for a small ten dollar board.

But if its a board a foot square with a great many many expensive components, it then becomes a very expensive repair for someone.

 

[marce]

If a track has burned away the board is unlikely to be repaired these days, cosmetically and functionally it will not be up to spec. It is also very expensive to perform these repairs. It is only likely to be repaired during prototyping and initial testing these days. I would also look for solutions to make sure this does not happen as I would not like to think of a design where this is possible these days, to many class action law suits floating around.
I do a few high current high voltage boards and always tend to bury high voltage tracks on inner layers, for safety and the greatly reduced track to track spacing's. The boards will not blow a track because we tend to allow for fault currents so check the figures using copper trace fusing tables and information, also I only ever work to a 10 deg C above ambient figures to give further de-rating so I am not creating hot spots or traces that are borderline, in fact all high current I do is generally a copper pour rather than a trace or multiple layer bus bar routing (signal and return broadside coupled on multiple layers).

 

[Warpspeed]

If a track has burned away the board is unlikely to be repaired these days, cosmetically and functionally it will not be up to spec. It is also very expensive to perform these repairs.

So what you are saying is that a $5,500 variable frequency drive will be scrapped because a track has burned off the board.

I don't think so.

If the fried tracks are on outer layers it can probably be made to work again, it will look horrible, but the customer gets his drive back working and will be happy.

I have been asked to repair drives that had inner layer tracks blown. The board bulges and blisters, plated through holes are torn away, copper splatter shorts other inner tracks, and the whole thing definitely is unrepairable.

We are in the age of planned obsolescence. Try sourcing a replacement board for anything that has been out of production for a couple of years.

Rarely do circuit board designers end up doing penance in the service department, where they get to repair the problems they themselves created.

 

[KlausST]

Hi,

A high quality device should prevent a trace to burn.

Burned traces usually vaporize some copper...this will condens on every surface around. Mostely invisible. But it makes saftey clearance useless. Suddenly each - usually isolating - surface becomes conductive.
There is an invisible and uncalculable risk of electric shock.

Whenever there is dangerous high voltage around I don't recommend to fix the burned trace.

Again: the danger is invisible.

Klaus

 

[Warpspeed]

Judgement obviously comes into this...

But if the blown track does splatter significant copper around, the solder mask is very likely to limit the damage to the immediate vicinity of the disaster.

While its true, fuses and circuit breakers should normally be used to interrupt fault currents, idiots can be very ingenious in creating unexpected low impedance events <grin>.

Its amazing what 240 volts can do if connected directly across closed relay contacts !!
And one screw terminal block can look just like another, the wires go in just as easily.

I actually watched a guy plug a very expensive studio monitoring loudspeaker box directly into the mains. He noted that the speaker used a cannon socket, and he found a mains cable with a cannon plug on the end. Obviously the right cable to use.

The sound was very loud, but very short.

Maybe the loudspeaker manufacturer was at fault for not providing a circuit breaker on his speaker enclosure.

We all had a good laugh at the time, after all it is only money, as we all worked at the Government TV station, so no real harm done.

 

[marce]

I would be surprised if a single board cost $5,500 in such a device but maybe it did...
But as I said the design should be done so this DOES not happen and the fuse blows first, safety first, that's why there is lots of stuff on track heating and fusing currents, I would take all this into consideration. Just my view well more my philosophy I like to engineer my board layouts so they do the job required, the design I am on now cannot have a track blow under any circumstances such a failure is not an option. I do quite a few layouts like this but apply the same rules to any layout I do. If there is a risk of overcurrent then the board must be laid out to avoid trace vaporisation and must carry the fault current until another safety device (fuse) breaks the current loop.

Agree it is hard to avoid stupidity, I have seen a heavily populated VME rack go up in smoke when someone put the mains into the keyboard connector!!! Don't ask how, we couldn't work it out, the best thing was they wanted it replaced under warranty (back in the 90's).

 

[Warpspeed]

I would be surprised if a single board cost $5,500 in such a device but maybe it did...

I did not mean the actual PCB cost that much, but if it cannot be repaired, and a replacement board is no longer available, the whole drive may need to be junked.
A fairly large brand new replacement motor drive may easily cost that much.

So generally for obsolete equipment, a customer will be happy if it can be put back into service, no matter how ugly the repair. Even if it blows up again after some time, he will realize you are trying to help if he is brought fully into the picture from the start.

Warranty work is entirely different to resurrecting old obsolete equipment.
And sometimes obsolete means only a year or two once a newer model comes out.

Another scenario is that company X gets sued and suddenly without warning ceases trading.

Too bad if your board blows a track, its then either an ugly repair, or the whole thing has to go into the dumper bin. There are sometimes genuine financial motives for repairing boards you would not normally regard as worth the trouble of saving.

 

[FvM]

I agree that in case of obsolete instruments, also development prototypes, generally home-brewed electronic devices and home repair work you possibly try to save all kinds of burned PCBs if ever possibly. Manufacturer service staff and authorized service partners will never do, for serious reasons that have been mentioned.

We got around to this point by Warpspeed's comment that surface power tracks are less problematic when suffering fusing and easier to repair. I still don't see how the consideration can apply to professional electronic devices and as stated by marce, spacing and creepage requirements are often against it.

The original question has been answered quite verbosely anyway.

 

[Warpspeed]

I suppose it comes down to planned obsolescence, versus design for easy repair.

 

[marce]

We use to mend old school "battery chicken light controllers" with bits of wire:smile: same reason... obsolete products, old school design (no digital) and tight ars*d farmers.