PCB Design Considerations: Trace Width

PCB Design Considerations: Trace Width
www.pcbway.com/ Get 5 boards in about a week for $22!
Infrared Thermometer: amzn.to/3JGe2kI
IPC-2221A Spec Sheet: www.ipc.org/TOC/IPC-2221A.pdf
Please help support this channel at NO COST TO YOU. If you shop on Amazon, please use this link: amzn.to/3706t6c
I will get a small commission and it costs you nothing.
Learnelectronics Merchandise:
Yes!! I now have merch: teespring.com/stores/learnele... Electronics themed T-Shirts!
----------------------------------------------------------------------------------------------------------------
HEY YOU ABOUT TO POST SOMETHING STUPID:
Yes, I know my hands are large/fat. It's because of a heart condition I suffer from called Congestive Heart Failure. Here's a link to some information about CHF: cle.clinic/2TdS2Ux
One of the symptoms is that it causes fluid retention, mostly in the hands, feet/ankles, and face..
--------------------------------------------------------------------------------------------------------------------
Amazon Store:
If you are shopping for electronic components, test gear or consumables please consider visiting my Amazon shop @ amazon.com/shop/learnelectronics
Please check out www.patreon.com/learnelectronics and pledge a dollar if you can. It will go a long way to keeping the channel alive. It costs a bit of money to buy all the items and produce these videos. Your help is appreciated.
---------------------------------------------------------------------------------------------------------------------
Donation:
Or....if you'd like to send a one-time donation you can use this link: paypal.me/learnelectronics
---------------------------------------------------------------------------------------------------------------------
Gear I use:
DMM: Owon B33+ - amzn.to/3qQgH20
Power Supply: Matrix 3206 - amzn.to/37C0fL9
Portable O'scope: YEAPOOK ADS1013D - amzn.to/2ZG4G3x
Bench O'Scope:Hantek DSO2D15
- amzn.to/3ir307N
Soldering Station - KSGER T12 - amzn.to/2NuNdby
Multimeter leads: Probemaster - amzn.to/2ZJ0BLK

[@nigeljohnson9820]

One trick I have used is to take thick copper wire from single strand mains cable, and solder this on top of the copper trace. This solution is better than simply adding a thick layer of solder, as the solder has a much higher resistance than the copper. It is possible to make a very neat finish to the copper wire. It is a good idea to provide two through hole fixing points for the cable. Then solder down one end of the added copper wire. Tack solder the copper wire along the length of the copper trace, to hold the copper wire tight to the pub trace. It is then possible to run solder along the copper wire, between the tacking points. If done well, it produces a clean ridge line along the center of the trace made of the copper wire, covered in a thin layer of solder. This solution is very handy when fitting mains switching relays to PCB, when 13 amp mains current is being switched. Clearly the idea solution is to use short thick copper tracks linking the relay contacts to mains rated power connectors. I have used this technique in the construction of an IR controlled, fused, multisocket main extender.

[@learnelectronics]

Great tip, thanks.

[@kevinscollan8293]

Yes bruv, I've used the exact same method described but with slivered copper wire you can get it cheap on Ebay used for jewelry or craft things.😎💕🖖

[@bblod4896]

You can use solder wick to increase the current handling capability of a pcb trace, just remove the solder mask and tack down the wick on top of the trace. Good lecture Paul. Have a great week.

[@learnelectronics]

That's a great 💡. Thanks for sharing.

[@TRONMAGNUM2099]

Good Job. Excellent tutorial.

[@sinjhguddu4974]

An appropriate tutorial and at the right time. Good work!

[@learnelectronics]

Thank you kindly!

[@trelligan42]

Max current is thickness * width. A longer trace adds more resistance as well, so even signal traces can benefit from width increases if the run is long. That pretty green covering on your traces insulates the copper from the air (while it prevents oxidation) so keeps it hotter. Rule of thumb: make any traces running power and ground to outputs as wide as your budget will stand, and everything else twice as wide as necessary. This is because acid etching of PC boards is a chaotic process, and you may have thinner spots occasionally. [I don't know what method(s) PCBWay uses, but acid is common and cheap, so perhaps likely.]

[@nigeljohnson9820]

Acid etching does not necessarily lead to a reduction in copper track thickness. The standard industrial process for producing chemical etched PCBs is to use a negative chemical resist, defining where copper is to be removed. Before any copper is etched away, tin is electro plated on to the exposed traces, potentially covering them in a relative thick coating of tin metal. The negative chemical photoresist is then removed with a weak alkaline solution or other organic chemical solvent, exposing the copper area between the tracks to be etched away. The tin then acts as a resist to etching with a solution of hydrogen peroxide and sulfuric acid. This technique has the advantage of making the process less time sensitive, as the only copper being removed is between the tin protected tracks. As under cutting of a positive resist is avoided, this method produces much finer features, because there is less chance of finer tracks being eroded by prolonged etching time. This technique also lends itself to larger production runs, where the first layer of resist is screen printed onto the copper coated board, eliminating the photo exposure step. It would be interesting to see the current start of the art practice for producing PCBs. No doubt there are different methods used for smaller production run, to those used for large scale mass production. There are a number of companies that sell laser etching equipment for making prototype boards and preproduction runs of boards. It would be interesting to see the details of the processes used for making the small volume orders, such as those generated by hobby electronics. My guess is that laser etching will play its part, but it must fit into a total process that gives multilayer boards, with plated through holes and tin or gold oxidation protection for the copper, plus the layer of resist. Someone must have posted a KZbin video on modern PCB production techniques.

[@trelligan42]

@@nigeljohnson9820 Interesting. I have seen traces that had 'undercut' in places, and in fact solved a few problems with hairline cracks by noticing these. I fixed them by covering with solder or wire soldered onto it. My only foray into PCB creating was a home kit I used three times, and undercutting with trace narrowing was a problem.

[@nigeljohnson9820]

@@trelligan42 when using a positive photo resist contact printing, undercutting is a problem. It is the result of a number of processing issues, such as poor contact between the photomask and the un-exposed photo resist coated copper board during UV exposure, or not having a sufficiently opaque photomask, resulting, in a thin resist or one that is covered in microholes. (The holes being the result of using a laser printer to make the photomask.) The professionals who once used the positive resist technique, quickly discovered the need for the photomask to be truly black, unfortunately the lithographic film needed to generate this high level of contrast is very expensive and has a very short shelf life. It was once possible to buy a drum photo plotter that manufactured these masks in high resolution, unfortunately this too was very expensive instrument. The contact problem between the mask and the board was solved using a vacuum table to suck the film into tight contact with the unexposed board. The only way for hobbiests to control mask contrast is to double up the mask film, to improve the contrast ratio, and limit the UV exposure time, and use a less aggressive etching chemical, than ferric chloride solution, such as cuprous chloride and hydrochloric acid solution, or ammonium/ sodium per-sulphate. The latter, known by the trade name of fine etch. It too also has a very short shelf life once mixed. Process temperature control and air stirring of the etchent also improves the quality of the PCBs that can be produced. Heating the etching solution to 35 to 45C, controlled to 1 degree, improves the soeed and consistency of the etching reaction. The copper 1 chloride solution also has the advantage of being self buffering and regenerative, when blown with air, Some hobbiests have found a solution to the problem by using a laser plotter, to vaporize a paint based resist. The method shown in some KZbin videos is to coat the virgin copper board in spray on car paint. The laser plotter is then used to burn off the paint from the copper areas to be etched. This method can produce very high quality, fine, copper traces. It is a technique that is becoming less expensive, as cheap laser plotters become available. The limitations of such methods is that at best they can only produce two layer boards, which may still require through hole drilling. I personally find the production of home brew PCBs very satisfying, but sometime frustrating. Most of my boards are single layer, double sided, with a copper "earth" (0 volt) backplane. It is surprising the level of circuit complexity that can be achieved by this method. Currently, I am using a CNC router to produce most of my PCBs, as this makes it easy to include the through hole drilling step. Boards that have high voltage (mains) circuits, are still chemically etched, with the through holes drilled in the first step. All my boards are passivated with electroless tin plating over the copper.

[@blic-sx9ix]

I use Yankee units for traces! I set my design rule to 25 mil traces and 25 mil clearance for all my through hole projects. If I mount a relay cube on the board and route to a connector at edge of board I use 100 mil top and bottom layer and air gap coil/load barrier. That gives me 5+ amps on the load with one oz. copper.

[@oldwrench4213]

Excellent presentation. Valuable information.

[@NA-ms8rl]

amazing... thanks for ur time..

[@TheOleHermit]

Thanks for confirming what we already knew, Paul. But, what about the fun stuff, like finding the maximum current that the different tracks could have handled, before fusing? Isn't that what no one really knows, but needs to understand? That would be an impressive and memorable video to behold and share. Everyone loves watching sparks and destruction, especially on a slo-mo loop. JS 😎

[@mstrxxx0072]

i have the same probleme , iam trying to understand if my track could habdle the current ?

[@stoxel]

Now I feel robbed of an opportunity for a PCB project to end in smoke and fire. Thanks, Paul.

[@killer2point057]

"PCB Design Considerations: Trace Width Part II - The Burn Through Baby!!!!" Thank you as always for a great video!

[@MarcelRobitaille]

Bigger is not always better. Sometimes, traces should be sized for impedance matching or not accidentally creating an antenna with a big trace

[@schmetterling4477]

A wider trace doesn't create a better antenna. That's just a misconception about the physics of electromagnetic fields. A mismatched trace creates an antenna. Whether I have a 90 Ohm trace with a 90 Ohm termination resistor or a 30 Ohm trace with a 30 Ohm termination makes virtually no difference. It does, of course, take a lot more power to drive the 30 Ohm trace with the same voltage, if we are talking about logic with constant thresholds rather than analog RF systems with constant power. For logic it is therefor much better to use a higher impedance design.

[@createpk]

good representation. Could you please mention the copper thickness you used in these PCBs?

[@waynegram8907]

What is the perfect temperature that should be expected that is at operating voltage and current for any trace width? you didn't mention what is suppose to be the perfect temperature for each trace width at each wattage

[@learnelectronics]

It's not about a set temperature, it's about temperature above ambient.

[@waynegram8907]

@@learnelectronics So any PCB Trace that is measuring above room temperature is considered BAD?

[@phuonganh6445]

what Temperature Rise ?

[@samk2630]

First of, thanks for another great video. Secondly, can you talk about multilayer boards and thermal management? Thank you, Sam

[@learnelectronics]

No, that's not my specialty. I'm sorry to disappoint.

[@samk2630]

@@learnelectronics you never disappoint! Thanks anyway.

[@michaelpadovani9566]

I downloaded free software called Saturn (or something like that) that tells you info like this. I was really interested how big to make vias (something not mentioned in this video but same idea regarding current capacity).

[@MrDoneboy]

93.4 C, Paul!

[@sirnukesalot24]

I'll bet the typical engineering firm has standardized the maximum trace wattage they design to. I wouldn't be surprised if there's a class of consumer electronics never exceeds 1/4 Watt or even 1/8 Watt on any trace.

[@johnramsay1827]

I thought that was interesting. And it look like Dogly thought so too.