Opinions on design requested
HighFlight
I'm a novice and this is my first attempt at designing a PCB. If you don't
mind, could you give me some feedback on the feasibility of my design.
Also, suggesting on the resistors and LED's would be incredibly helpful.
The 5v voltage regulator design I'm stealing from the current board. It is
outputting a clean 5v. I'm planning on breadboarding the design...but a
starting point would be awesome.
The background is that this a 12V (boat) light control. The current one
doesn't work and I wan to replace it with a simpler design. I'm using a
P227EE1C DPST push button switch but still trying to keep the "cute" LED
indicators.
Any feedback is greatly appreciated.
Thanks.
Jon
http://www.malenfant.us/ShoreThing/Lights
Richard Henry
"HighFlight" <highfl...@yahoo.com> wrote in message
news:ubKdnRn16M_...@comcast.com...
Is there a schematic?
What are the loads?
What about the old design doesn't work?
Leon Heller
HighFlight wrote:
> I'm a novice and this is my first attempt at designing a PCB. If you don't
> mind, could you give me some feedback on the feasibility of my design.
It's not a good idea to have those acute angles where the tracks leave
the pads and where tracks meet.
Leon
--
Leon Heller, G1HSM
Email: aqz...@dsl.pipex.com
http://www.geocities.com/leon_heller
HighFlight
layout, which is pretty simply I think. The lights use a common + and
switched -.
The loads are light. They are just small 12v, very low wattage each lights.
Each circuit controls a single light (they used the shotgun approach to
lighting the cabin).
The current board kinda works when I first insert it. But after a few
seconds, all the circuits shut off and never come back on. If I pull the
board out, wait a few seconds, and put it back in, it works again for a few
seconds. I though it might be a thermal problem so I replaced everything
but the resistors...but the same thing happened. Can a resistor have
thermal issues?
Thanks.
"Richard Henry" <rph...@home.com> wrote in message
news:BnQpb.1469$mb5.580@fed1read02...
HighFlight
I'm not sure where you are talking about. Sorry, although I can solder
halfway decently, I got nutt'n otherwise. Could you give me an example of
where you mean.
Also, do you think with small 12v lights, that the switch is ok...and the
design in general is ok?
Thanks very much!
Jon
"Leon Heller" <aqz...@dsl.pipex.com> wrote in message
news:3FA7D462...@dsl.pipex.com...
Richard Henry
> No. I don't have any schematics. But the link I included shows the board
> layout, which is pretty simply I think. The lights use a common + and
> switched -.
>
> The loads are light. They are just small 12v, very low wattage each
lights.
> Each circuit controls a single light (they used the shotgun approach to
> lighting the cabin).
>
> The current board kinda works when I first insert it. But after a few
> seconds, all the circuits shut off and never come back on. If I pull the
> board out, wait a few seconds, and put it back in, it works again for a
few
> seconds. I though it might be a thermal problem so I replaced everything
> but the resistors...but the same thing happened. Can a resistor have
> thermal issues?
>
After sending my earlier reply I looked at the layout more closely and
realized you were controlling each light with a switch and had a LED in
parallel to show which light should be on. I'm guessing the F's are fuses.
Why was the old board so much more complicated?
If you are building only one of these, I can't see any problems with it, as
long as you get the LEDs in the right way.
For componenet selections, look in google.com. For 5V, and normal red LEDs,
330 ohms usually is about right.
George Patrick
> I'm not sure where you are talking about. Sorry, although I can solder
> halfway decently, I got nutt'n otherwise. Could you give me an example of
> where you mean.
>
> Also, do you think with small 12v lights, that the switch is ok...and the
> design in general is ok?
>
> Thanks very much!
>
> Jon
>
Don't route a signal that has an angle less 90 degrees if possible.
These are acceptable: | /
| /
------- -------
This is unacceptable: \
\
\
-----
The reason is that contaminents can get caught in the acute angles and
corrode, possibly causing long term reliability issues.
--
+--------------------------------+----------------------------------+
| George H. Patrick, III | Resources for PCB Designers on |
| geo...@pcb-designer.com | the Web - The Designer's Den |
| George.H...@tektronix.com | http://www.pcb-designer.com |
+--------------------------------+----------------------------------+
| Take what you like and leave the rest... My opinion ONLY. |
+-------------------------------------------------------------------+
HighFlight
etching process and am going to give that a go. Should be interesting. I'm
going to breadboard the circuit first. Thanks very much for the LED and
resistor suggestions. I will start with that.
The complications with the current board are just around the use of what I
have come to learn are J/K flip flop circuits and MOSFETs. I tried
re-creating the use the MOSFETs without luck. Switching the ground just
seems odd to me...but what do I know! I guess they just wanted to be cute
and use momentary switches. Heck...I just want it to work!
Thanks again for your help.
Jon
"Richard Henry" <rph...@home.com> wrote in message
news:tCWpb.1823$mb5.1203@fed1read02...
HighFlight
sides of the switch. I figure two paths were better than one. Considering
I'm going to etch it, I could just make it one solid area rather than making
the 'V' shape.
Thanks very much for the input.
Jon
"George Patrick" <m...@pcb-designer.com> wrote in message
news:ndWdnenmVeK...@comcast.com...
Walter Harley
> Ok...gotcha. The close angles were me just trying to take advantage of
both
> sides of the switch. I figure two paths were better than one.
Considering
> I'm going to etch it, I could just make it one solid area rather than
making
> the 'V' shape.
Indeed, the more copper you can leave on the board, the faster your etching
will go and (to massively oversimplify) the more current you'll be able to
handle. There's really no reason to have a couple of traces on an otherwise
bare board - it's probably better to have a couple of bare lines separating
large areas of copper. Sometimes when you do it this way, you can just etch
the bare spots with a Dremel tool or razor blade - you don't even need to
use chemicals at all. A rendition of a simple board done this way is at
http://www.cafewalter.com/cafewalter/fetpre/basspre2.htm.
By the way, if you're going to etch this yourself, then you'll have a much
easier time if you don't lay it out double-sided. Do it single-sided, and
use jumpers where you have to. With a little bit of thought you can
probably get it so that you need few if any jumpers; but even if you need a
lot, it will still be a lot faster and easier than trying to get both sides
of the board aligned and making sure that every via (the places where one
side connects to the other) is actually connected.
Jim Meyer
1, Your 5 volt regulator, VR1, only has connections to two of its
leads.
2, The dropping resistors for the LEDs all connect to each other at
one end, but then they only go to a capacitor. LED current will only
flow while the capacitor is charging (brief flash).
3, You really don't need to drop the 12 volts down to 5 volts just
to run some LEDs. Use the 12 volts directly by raising the dropping
resistors' values.
Jim
HighFlight
1. Yeah, I was going to jump the lead in into the VR using a diode like the
current card. I should have made a circuit diagram...sorry.
2. Wow...I totally misread the current card. After your comment, and
looking deeper, I realized the ground and output are jumped with the
cap...which I then followed with "DUGH!". Thank you on that one!
3. Well that would certainly make things less complicated. What size of
resistor could I use? I have 10K, 1K, and 3.3K, and 330, and 220ohm
resistors. Can they be strung together? Parallel or serial?
Thanks!
Jon
"Jim Meyer" <jme...@nektonresearch.com> wrote in message
news:21ede509.03110...@posting.google.com...
HighFlight
need a few jumper. You are right, that will make the process much simpler.
Thanks.
Jon
"Walter Harley" <wal...@cafewalterNOSPAM.com> wrote in message
news:bo9r37$haf$0...@216.39.172.65...
HighFlight
resistor. Much easier!
Thanks!!!
Jon
"HighFlight" <highfl...@yahoo.com> wrote in message
news:xamdnURRh7D...@comcast.com...
James Meyer
>
>3. Well that would certainly make things less complicated. What size of
>resistor could I use? I have 10K, 1K, and 3.3K, and 330, and 220ohm
>resistors. Can they be strung together? Parallel or serial?
>
>Thanks!
>
>Jon
>
You are starting with 12 volts and the LED will subtract 2 volts all by its
self leaving 10 volts remaining. If you use a 1K value for the dropping
resistor, Ohms Law (V = I x R) tells us that you will limit the LED current to
10 mA which should make the LED pleasantly bright. The 3.3K value will be fine
if the 1K value makes the LEDs bright enough to keep you awake at night.
Heck, with a few transistors and a photo-resistor, you could make the LEDs
automatically get brighter in the daytime and dim at night.
Jim
HighFlight
Thanks to everyone who made suggestions. I have a simpler and cheaper
design now between making it single layer and getting rid of the VR and just
using between a 1K to 3.3K resistor for the LED's.
My only other question is the switch. Is the
http://rocky.digikey.com/WebLib/E-Switch/Web%20Data/P197%20Series.pdf safe
to switch the raw 12V source in the boat? I'm using 2mm traces on the board
and it's switching ground with a fuse between the switch and the light.
Thanks.
Jon