Could someone spare the time to give this a quick design-look-over?

[KarenColumbo]

Attention: n00b-question (again)

Since I started my presence here with two quite ambitious projects I realized I gotta get the knots out of my brain first. So I decided to build a few quick PCBs to get the juices flowing, so to speak.
One of them is the legendary Catalinbread SFT 2, just because. The others are the AMZ Mosfet Booster, a quick and dirty paraEQ and a darkglass experiment.
I recently finished (not quite) the SFT 2 PCB - schem on one screen, ExpressPCB on the other. Here's the result:



Could someone give it a quick glance, design-wise? I haven't ckecked all the connections yet, so all I ask is for some of you dear experts to point out to me serious and obvious design flaws, if you would.

Especially: Are the traces thick/thin enough? Should I fill the "empty" spaces with copper planes? Shall I make the ground connection "Full Circle" or is this not advisable?

It's not that I don't google like heck to collect articles and How-tos and What-nots, but it takes me quite a while to absorb them - and errors don't become obvious to me as they become to an experienced builder.

[diegocw]

Hi, I believe the Diode should be pointing in the oder direction.


Enviado desde mi iPhone utilizando Tapatalk

[KarenColumbo]

Hi, I believe the Diode should be pointing in the oder direction.


Enviado desde mi iPhone utilizando Tapatalk

THX - noted and done!

[Transmogrifox]

I see certain component footprints are overlapping.  Double-check your actual capacitor dimensions to make sure this is ok.  You have plenty of room on the board to adjust these parts if needed.

May as well copper fill the secondary side of the board and make it into a ground plane.  It's an easy thing to do that doesn't cost anything.  Then all your through-hole components going into the ground ring on top will stitch to the ground plane and each component will get the shortest possible (straight-line) path back to the board's ground connection.

In more complex designs it can help a lot with layout because you don't have to get creative to route something into ground.  Just punch it down to ground plane wherever the part best fits.  And then you also get a really good low-impedance ground throughout the circuit.

As for your question about making a full ground ring -- for an audio circuit I don't think you would be able to tell any difference, but a good rule-of-thumb is to do what you did.  Any kind of enclosed loop will make a better antenna for RF energy.  In most audio circuits this will be naturally filtered away and inaudible, but in high-gain and/or nonlinear circuits this can get demodulated into audible noise.

The non-closed loop like you have is still an antenna so the "best" thing is a ground plane (which is still an antenna).  The advantage of the ground plane is lots of copper, low impedance so the RF currents in the ground don't generate very large voltage differentials across the ground.

The worst I think is the loop, though, because it provides a return path for RF currents so you can have RF loop currents around the ground creating RF voltage differentials across the circuit ground connections.

Hopefully I haven't confused you more.  Here's a summary of ground connections:
Worst = ground loop
Good = ground ring (not enclosed) (like you have)
Better = Star ground (all grounds make the shortest path straight back to a single point for each component)
Best = Ground Plane (this is pretty much a glorified star ground)

By the way, I like your signature .  I have read several Terry Pratchett books -- good equal-opportunity satire for all.

[KarenColumbo]

Thx, Fox, very much - will give these a good hard think!!

[PRR]

This may be minor but it bothers me.

Top-left, you have a big storage capacitor, connected "the long way" through signal connections.

Would you build the road to the strip-mine through the neighborhood?

The cap can be connected *direct* to the power highway.



The larger picture is much like packing the tent, the cooler, the stove, the kids, and the dog in the minivan, without interference (as Fox says), fuel-drip on the dog, and easy access if a kid gets sick. No-excess space is a lesser goal (you could bring the TV, or use a smaller car). Or the snowsuit, books, and nice-shirts in my closet. Have to ponder various arrangements.

Probably makes no difference here.

[KarenColumbo]

hehehe, yes, I see thx for pointing this out! As with the way those data flows - there's so much I have to learn yet I promise I will ponder all the variations of how to escape the minefields in this crowded neighbourhood. Good design seems to have a whiff of alchemy to it. Well ... I started all this, I'll see it through to the bitter ... output.

[KarenColumbo]

v1.3
I didn't dare to go below 0.018 with the clearings - I'm etching those myself, dunno how exact my equipment (Laserprinter, UV lightbox and cheap etching frame) will work.
I checked the overlapping components - but it's just the outlines. I deliberately took huge capacitors to have clearance between their legs (like so few of them have) and the transistors, pots and input/output/9V/switch pins will be socketed. Maybe one or the other tone cap.

[antonis]

Sometimes it's good to make "planes" not only for GND but also for some neighbour items..

i.e. possitive leg of up left electro + resistor lower leg + cap to SW4_5 upper leg..

Also, ground plane continuities via 0.1mm paths are troubles on etching..

[KarenColumbo]

I see what you mean. Already did that, but with simple rectangles that make it hard to solder into. Now that I know how to make those planes (and "thermal pads") I'll have a review of this one. Can't wait to etch this
I took 0.020" as clearance - worked out fine. And if I take the time to setup the mask tidily it works out. Takes some crap versions, but I'm getting there.
Well, without further ado - this is gonna be build til sunday

[jonnyeye]

Please check the size of your components! All of the small capacitors are drawn very small, with 0.1" pin spacing and 0.1" maximum width. It is far more common for radial film caps and mono ceramic caps to have a 0.2" pin spacing, and the width can vary.  You may have trouble physically fitting the capacitors into place.  Also, the input protection diode and the resistor below VR2 look to be too small to be placed without standing up (assuming 1/4W resistors and 1N400x diode). There is a lot of empty space on your board, it shouldn't be difficult to make more room. The large space given for the electrolytic caps (which I doubt will be anywhere near that big!) bothers me much much less, though you may have to bend the leads to fit a 0.2" pin spacing.

Either check your datasheets, or have the parts in hand to measure their size.  A good test before etching (especially while learning how to layout) is to print out a copy and physically poke the leads through the paper. If things smash into one another or won't fit you can make changes before committing to copper.

[Transmogrifox]

Please check the size of your components! All of the small capacitors are drawn very small, with 0.1" pin spacing and 0.1" maximum width. It is far more common for radial film caps and mono ceramic caps to have a 0.2" pin spacing, and the width can vary.  You may have trouble physically fitting the capacitors into place.  Also, the input protection diode and the resistor below VR2 look to be too small to be placed without standing up (assuming 1/4W resistors and 1N400x diode). There is a lot of empty space on your board, it shouldn't be difficult to make more room. The large space given for the electrolytic caps (which I doubt will be anywhere near that big!) bothers me much much less, though you may have to bend the leads to fit a 0.2" pin spacing.

Either check your datasheets, or have the parts in hand to measure their size.  A good test before etching (especially while learning how to layout) is to print out a copy and physically poke the leads through the paper. If things smash into one another or won't fit you can make changes before committing to copper.

Although sometimes committing them to copper and getting your hands on all the parts is a nice way to learn.  It's all a matter of whether you want to get it right the first time, or use the experience for improving future projects. 

I'm of the philosophy to waste (spend) the time and resources to see the outcome just to really learn by experience.  Some people (like my brother-in-law) do a lot of research and planning up front and nail it the first time.  When it comes to building hobby projects the value is in how much you enjoy the process.

If you really enjoy doing your research up front and nailing it the first time, then you really need to take your time to study datasheets and use your caliper to check ever component, and edit your custom footprints to match the actual parts you intend to use.

If you're more of cowboy like me, you just etch the damn board and see how it works out.  You shove a mess into a box and play guitar.  Then you revise your design and build a better one if the mess-in-the-box causes problems.

The expense in the first method is time (planning and research).  The expense in the second method is materials, but (in my experience) significantly less time...and a lot more fun because I enjoy sitting down with a soldering iron and some parts a lot more than I like looking at datasheets and measuring dimensions....so for me the spent time is not time wasted, but time spent doing what I enjoy doing the most.

[KarenColumbo]

Please check the size of your components! All of the small capacitors are drawn very small, with 0.1" pin spacing and 0.1" maximum width. It is far more common for radial film caps and mono ceramic caps to have a 0.2" pin spacing, and the width can vary.  You may have trouble physically fitting the capacitors into place.  Also, the input protection diode and the resistor below VR2 look to be too small to be placed without standing up (assuming 1/4W resistors and 1N400x diode). There is a lot of empty space on your board, it shouldn't be difficult to make more room. The large space given for the electrolytic caps (which I doubt will be anywhere near that big!) bothers me much much less, though you may have to bend the leads to fit a 0.2" pin spacing.

Either check your datasheets, or have the parts in hand to measure their size.  A good test before etching (especially while learning how to layout) is to print out a copy and physically poke the leads through the paper. If things smash into one another or won't fit you can make changes before committing to copper.

Although sometimes committing them to copper and getting your hands on all the parts is a nice way to learn.  It's all a matter of whether you want to get it right the first time, or use the experience for improving future projects. 

I'm of the philosophy to waste (spend) the time and resources to see the outcome just to really learn by experience.  Some people (like my brother-in-law) do a lot of research and planning up front and nail it the first time.  When it comes to building hobby projects the value is in how much you enjoy the process.

If you really enjoy doing your research up front and nailing it the first time, then you really need to take your time to study datasheets and use your caliper to check ever component, and edit your custom footprints to match the actual parts you intend to use.

If you're more of cowboy like me, you just etch the damn board and see how it works out.  You shove a mess into a box and play guitar.  Then you revise your design and build a better one if the mess-in-the-box causes problems.

The expense in the first method is time (planning and research).  The expense in the second method is materials, but (in my experience) significantly less time...and a lot more fun because I enjoy sitting down with a soldering iron and some parts a lot more than I like looking at datasheets and measuring dimensions....so for me the spent time is not time wasted, but time spent doing what I enjoy doing the most.

Thank you for looking into this mess I made with open eyes! I will do both of your suggestions: I have to build up a personal component database anyway, because it saves time and opens up possibilities. I love fiddling around on the virtual PCB, routing, tracing and things of that nature - all the time thinking about how those tiny electrons want to enjoy their constant bumping into each other from input to output And I have a box with failed projects - I think a bit more than half of my spontaneous boards never lit up in any way But it's fun to try and bet on myself - more fun, even, when everything looks clean and nice, though.

Will update this thing soon - want to commit it to Na2S2O8 a.s.a.p.

[Transmogrifox]

Awesome   Have fun Na2S2O8.

Different stuff but it reminds me of a rhyme:

Johnny was a chemist's son,
but Johnny is no more.
What Johnny thought was H2O
was H2SO4.

Keep your etchant solution and your glass of whiskey in distinctly different types of containers. The distinction will be vitally important when both containers are nearly empty.