Top Trace Removes Bottom Trace? (ANSWERED)

Hi.

I am running a high voltage 240VAC line from part to part and wanted redundancy by forming a thick trace (3mm) on the top layer and ALSO a thick trace from/to the same components on the bottom layer as well. BUT when I drag a trace wire on the bottom layer from the components that already have a trace on the top layer, the top layer trace is automatically removed and only the bottom layer trace is kept.

Why is that and how can I run a trace from two components on BOTH the top layer and the bottom layer at the same time (and even middle layers too)?

Thanks in advance!

I routinely do what you are trying to do, and without difficulty. Many years ago I was bitten by one or two bad vias so now I routinely switch layers with two vias and small redundant redundant tracks. I do not know the answer to your question but if you go to help>about>copy version info and post that, it will help someone else to answer your question.

BTW I am shocked that you would work with 240 VAC. I hope you are not.

If you make a new connection, the KiCad assumes that you have found a better track location and the old track is “obsolete” and removes it.

If you want to keep it then un-check Pcbnew / Route / Interactive Router Settings / Remove Redundant Tracks.

Thanks Bob. Good to know at least it CAN be done. So you are able to run multiple traces on different layers from the same 2 pads?

I am using the latest KiCad 5.1.12.

Ah ha! OK. That is probably it. Thank you so much Paul!

Yes I have that unchecked. Before my previous post I looked unsuccessfully in preferences. But I am sure that I set that router setting once upon a time…lacking memory of what I had done…

That’s where I looked too assuming there was some setting for this. As you had found, it was elsewhere.

Thanks so much guys!

Don’t understand why you are shocked.
When my father told me (I was 9 those time) how transformer works I wound the transformer (7 primary and 1 secondary turns on the middle column of the core). Connected the AC voltmeter to secondary and connected the primary to 220VAC socked and didn’t get expected 31V at output :slight_smile:
Someone in GB who wanted to do the same experiment had to connect to 240VAC.

Hello @Piotr,

I suspect it was @BobZs’ humour at work… mains voltage (240 VAC) can be “shocking”.

What did you get?

A smack in the head?
A lecture?
A blown fuse or tripped circuit breaker?
Magic smoke?
All of the above? :slightly_smiling_face:

Another alternative is to use zones in place of tracks. I prefer zone for routing high voltage and high current as typically I want as much copper as possible to draw heat away from the core conduction path

This way you can duplicate (or assign with V6) to multiple layers. This also offers the benefit of interleaving to produce a lower inductance trace. I did this recently with a 6layer, 3oz copper card to carry 100A at 540V

100A@540V !!! Can that even be done on a PCB? How hot does it get even with 6 layers @ 3oz? I thought at those powers, metallic buss bars would be needed to be soldered on the traces. No?

Also, can someone give me a little education on creepage and clearance? How much distance between traces will be required for 240VAC @ 2A with 2oz copper and 1mm traces?

I was alone at home (ill in bad). Circuit breaker tripped but I didn’t noticed that. I also didn’t noticed any spark when I was connecting my trafo to socket. Those time in Polan frequently happened breaks in power supply for few hours so I said: “Of course when I want to check something they had to switch power off.”

10C temprise :slight_smile: its a matter of providing wide enough copper so the cross sectional area aligns with the loss limits.
I have another PCB that is 16oz copper :slight_smile:

To connect to the traces we typically don’t solder but use a RedCube type conductive pillar from Wurth or Erni (can get up 300A). The downside of using 3oz copper is the process limits and thus the minimum track and gap is a lot larger … I had to redesign some “common circuits” as SSOP and SOT23-6 just do not match fabricator limits.

As to creepage and clearance: IPC-2221 is you friend and the pcb_calculator that is included within Kicad has such separation calculations

Indeed!! :slight_smile: I have worked with more than that. And I have been shocked…

Electrifying experiences! I also know that feeling :star_struck:

I got my start in electronics at the ripe old age of 13 as an amateur radio operator. Perhaps my parents did not know enough to prevent me from putting together a transceiver kit with 750 VDC for the anodes of the transmitter final amplifier tubes.

http://onetuberadio.com/2016/03/15/eico-753-seven-drifty-three-review-1966/

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.