[sdiy] Self-Soldering Circuits
hbissell at wowway.com
Thu Jan 19 18:07:34 CET 2023
This looks "attractive" in a nerdy kind of way but it really isn't a great idea. Thinking this is a "ground plane" is completely wrong.
A good ground plane would be be an unbroken plane under the top layer (or bottom layer, etc) that is intimately connected to all ground points on the top layer with the intent to reduce return path impedance as much as possible. The serpentine trace is not going to benefit at all even if connected to ground only at a single point. The comments about it being an antenna structure are quite reasonable.
You might design it so that the serpentine trace will be grounded at multiple points by attaching vias to the top and populating multiple zero-ohm resistors. that might help damp out any antenna effect, but still does not address the fact that it does not make a useful ground plane.
Now if we consider maybe a six layer board then you might have top layer / layer 2 ground / layer 3 serpentine etc... and try that. It still might work thermally. You will still want a lot of vias from the top to the ground plane.
Anyone willing to make blind vias from top layer to layer 2 ground, then make layer 3 serpentine? Not for low-cost but it might actually allow a true ground plane that functions as a return path.
If you route your PCB as all pairs of conductors (signal in parallel with ground return) it could work OK. No need for a ground plane then... That is a tough PCB design task. (even with Altium etc...)
Google "Daniel Beeker" He's an application engineer with NXP (Freescale) and has many PCB training sessions online that take the approach that traces etc. on a PCB are in effect structures to control magnetic fields (a la Maxwell...) for more info about routing as duplets or triplets, and how return currents flow in a ground plane. You can attend some classes in person if NXP Tech Days comes near where you live.
I specify my ground planes to be 8 mils beneath the top (bottom) layer, that is the minimum that does not cost you extra money at fabrication. So the return path is 8 mils away from the signal path. Closer would be better but this is not bad.
I showed to video to my workmates and we all had a good laugh. It's an interesting concept but the devil is in the details...
ROTFLFMAO to the "your subscription has expired..."
From: Ben <ben at stuyts.nl>
To: Roman <modular at go2.pl>
Cc: synth-diy <synth-diy at synth-diy.org>
Date: Thursday, 19 January 2023 10:40 AM EST
Subject: Re: [sdiy] Self-Soldering Circuits
Yeah! Plus a thermistor in a feedback loop to keep a constant temperature.
> On 19 Jan 2023, at 13:08, Roman Sowa <modular at go2.pl> wrote:
> My first reaction was "what an idiotic crazy idea with no purpose". But then again, since we get internal layers for free (JLC making 8-layer PCB for $2 now) it's no brainer to leave one or two layers for the heater. Although I would put it in small enclosure to prevent unnecessary heat loss, which causes temperature gradients in the components when done in free airflow.
> Another use of this technique is to make ovenized exponential converter.
> With cheap & small DC/DC converter pumping 1A at 0.5V this should be enough if not too much.
> W dniu 2023-01-18 o 23:31, cheater cheater via Synth-diy pisze:
>> This could be cool if you don't want your kit's user to have to use a
>> soldering iron... You do need a power supply that can do 150W-ish
>> Synth-diy mailing list
>> Synth-diy at synth-diy.org
>> Selling or trading? Use marketplace at synth-diy.org
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> Selling or trading? Use marketplace at synth-diy.org
Synth-diy mailing list
Synth-diy at synth-diy.org
Selling or trading? Use marketplace at synth-diy.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Synth-diy