PCB design consideration(s)

From DIY contraptions to sophisticated FreeEMS-specific designs! Plus general hardware development!
Post Reply
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

PCB design consideration(s)

Post by jharvey »

I had some thoughts that might come in handy for when PCB design is underway. I would put FET's and other power chips on their backs and surface mount them to the PCB (even though that are thru hole parts). No bending leads and creating stress points that can fatigue and cause weird resistance issues. Many people underestimate the strength of the coppers adhesive.

It's a quick note, and premature, but might as well start somewhere. Are there any other ideas out there that might come in handy when thinking about PCB design?
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: PCB design consideration(s)

Post by MotoFab »

jharvey wrote:I had some thoughts that might come in handy for when PCB design is underway. I would put FET's and other power chips on their backs and surface mount them to the PCB (even though that are thru hole parts). No bending leads and creating stress points that can fatigue and cause weird resistance issues. Many people underestimate the strength of the coppers adhesive.

It's a quick note, and premature, but might as well start somewhere. Are there any other ideas out there that might come in handy when thinking about PCB design?
Surface mounting TO220 package power transistors onto a board is a mechanically sound method for sure. For it to be successful, there are a couple of things to consider.

Depending on the heat dissipation requirements, a large copper area of the board must be devoted to heat sink area. And since the heat isn't conducted to the outside through direct attachment to the enclosure, the ambient temperature within the enclosure is a much larger component the heat dissipation equation.

If the drivers are on a separate board, and the Rdson of the drivers is very low, then the power dissipated across the junction is also low. But either way, the heat from the drivers becomes part of the internal ambient heat. And that heat cannot get to the outside of the aluminum enclosure very quickly. That doesn't mean it will not work, it only means that the numbers either say it will work, or not.

Regarding the end-user-assembly, most users do not have a soldering iron with enough heat capacity to flow the backside of a TO220 part in a reasonable time. This doesn't mean a small iron turned up to a high temperature.

About ideas for pcb design, I agree it's sort of premature. I don't really have any ideas because the circuit is still sort of blank sheet at this time.

At least I think it is, or maybe I think it should be until other related decisions have been made.

- Jim
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: PCB design consideration(s)

Post by Fred »

Nice idea Jared :-)

And great post Jim!
MotoFab wrote:Surface mounting TO220 package power transistors onto a board is a mechanically sound method for sure. For it to be successful, there are a couple of things to consider.
Another one is that a TO220 laying down uses a LOT of space. The DPAK stuff might be more appropriate and can be had with very low RDson and high current handling too.
Regarding the end-user-assembly, most users do not have a soldering iron with enough heat capacity to flow the backside of a TO220 part in a reasonable time. This doesn't mean a small iron turned up to a high temperature.
Good point!
About ideas for pcb design, I agree it's sort of premature. I don't really have any ideas because the circuit is still sort of blank sheet at this time.
I agree, premature, and yes, it more or less is blank at this stage, yes.
At least I think it is, or maybe I think it should be until other related decisions have been made.
Agreed :-) Though the time for PCB's is coming in the medium term so some decisions will have to be made in that same time frame.

I have some ideas about boards, but they can be summarized fairly quickly :
Not enough room on a single 160x100 eurocard for what we need.
Therefore two eurocards required to be sandwiched or similar.
Must accept the Tech Arts card in IDC socket form.
Tech Arts card should sit close to one end such that DB9 can be flush for users that can't afford to upgrade to USB
High V/I inputs should be separated AND isolated from low V/I inputs.
High V/I inputs should be separated AND isolated from low V/I outputs.
Inputs should be separated and isolated from outputs.

All of this is discussed in more detail (and more clearly) in the wishlist thread above.

Obviously this is for our first version for those who don't want to do fine pitch smd work themselves and before we have someone keen and willing to manufacture to our designs. At this point Tech Art = plug in CPU which is the main point of using it. It's very convenient for what we want to do.

Later anything is possible. I personally would like to see a microsquirt competitor out of this project as I just love small things of all sorts. Cam intends to design an SMD single board version for hand soldering also. The world is our oyster, but we need a rock to get started on ;-)

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: PCB design consideration(s)

Post by MotoFab »

Fred wrote:Another one is that a TO220 laying down uses a LOT of space. The DPAK stuff might be more appropriate and can be had with very low RDson and high current handling too.
All the really good parts are often only available in DPAK or D²PAK. Still though each part'll need typically a minimum 25mm² of board area for the heat sink, and that's in addition to the part outline area.

-Jim
Last edited by Fred on Wed Jun 25, 2008 8:38 am, edited 1 time in total.
Reason: Fixed quote tag
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: PCB design consideration(s)

Post by Fred »

Yeah, understood. This puts a serious damper on trying to SMD 12 drivers and fit everything else on too then doesn't it. I knew there were requirements, but I wasn't sure about just how much each needed. Thanks :-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: PCB design consideration(s)

Post by jharvey »

I wasn't quite clear when I said surface mount. I didn't picture mounting the back of it. I was imaging just the leads such that tab can be mounted securely to the heat sink. The colder the better. Then the leads would hold the PCB down and some form of standoff/spacer would keep the power devices from pealing of the copper. There would be nearly not up and down stress, only side to side, which would be minimized by proper length and width measurements.

Well it's off to read the wish list in more detail. I skimmed it so far, and I'm trying to read it better as I go.
Post Reply