PCB with repetitive design

Easily done with an EDA package that supports multi-channel designs. You can create the repeated part as part of a hiearchical schematic and instantiate it as many times as necessary, then send that to the PCB layout, place the parts carefully, then route it. Some packages (Altium, for example) allow you to copy placement and layout between channels using predefined chunks of PCB ('rooms').

Or, if it's simple, just cut and paste on the schematic and lay it out manually and save that part of the learning curve.

For sure there are. Timing too.

Speaking of which, roughly how much current can 80 mils trace in 2 oz copper handle? We need over 10A.

A useful rule of thumb is that 0.5 oz copper has a sheet resistance of 1 milliohm per square at 25 C. Your trace will be about 1/(4*0.08) = 3.1 milliohms per inch. At 10A, you'll be dissipating 300 mW per inch of trace, and that will get toasty. Maybe put a thermal pour on it. ;)

Cheers

Phil Hobbs

You will have a 60C rise in conductor temperature for your stated conditions. It will fuse in one second at 42A.

There are only 4 distinct structures repeated several times. Am I correct in assuming I need a list of parts and coordinates? The repetition is sli ghtly imperfect. Units have unique tests for sequential binary (not unary) addressing.

probably cheap DRAM. Are there any special voltage or current requirement s?

Of course. It sounds like you need to go back and sort out your circuit des ign first.

NT

Den onsdag den 13. august 2014 21.43.08 UTC+2 skrev snipped-for-privacy@gmail.com:

depends,

-Lasse

There will still be weak points in the layout. Just trying to estimate how much current it can handle. Perhaps 5A?

http://173.224.223.62/motor/bldc.jpg

There are only 4 distinct structures repeated several times. Am I correc t in assuming I need a list of parts and coordinates? The repetition is sl ightly imperfect. Units have unique tests for sequential binary (not unary ) addressing.

, probably cheap DRAM. Are there any special voltage or current requiremen ts?

To what extent does layout depend on the exact (pinouts of) the components? Can you connect repeated components with a formula? How do you specify w ord buses? Can a program figure out the sides of the board?

A multiplexer for 100 rows would be the width of an address and 100 pins lo ng.

I have Verilog simulations of the units and raster diagrams, but they're no t polished. If it would it help to take a look [1]. There are 5 PNG's and related.

So I've got to saturate 100 transistors? Or however many?

All the wide tracks on that have room for extra cu in at least some places. You can also solder along the tracks.

NT

with a high temperature material, but that's a lot of heat to be burning.

Yes for the bottom traces, but not for the top traces. Probably need to do filled via as well.

get kicad or saturn pcb - both have a calculator for that - and do it yourself

To specify a wide track, you can specify a "hinge / corner" with a reflecti on vector. The spacing and angle of incidence, plus the reflection vector, plus the angle of reflection, determine the reflection spacing. The refle ction vector is a pseudovector. You also need a sign, but you could just i nvert the vector to "toggle". Neighboring "hinges" also need to be rotated if a hinge is dragged.

I have no idea what the units are going to look like once I get the actual parts. I expect it will have to be redrawn if I put it on silicon.

You could probably just arrange some muxes to get the "Eq" results. Though the switching muxes do need the unary ports.

I am too lazy to do either. Google is faster.

According to this table, i need 4 oz of 50 mils or 3 oz of 80 mils for 10A.

At GenRad we used a "bus bar" that was a heavy flat piece of copper material mounted perpendicular to the PCB with thru-holes mounting "pins" about every inch. I can recall occasions where we ran 50A !! ...Jim Thompson

Or require PCB maker to plate up to 2 oz min.

At 72V and 15A, we would need 5 oz copper and heat sinks. Assuming a 50mOhm MOSFET, each active one would draw 12W and two would be active at a time.

Power estimate at: http://173.224.223.62/motor

We probably need to space out the MOSFET and add heat sinks.

Top? I don't see anything on that pcb other than the red grid that can't have more copper area added. In fact I'd say that layout shows someone hasn't thought this through properly.

NT

Remember that solder has much higher resistivity than copper, so soldering along the top will not do a lot to reduce the resistance and therefore will not change the power dissipation appreciably. It will add thermal mass, which will help if the average current is a lot less than the peaks, and it will help with cooling. But widening the copper tracks will make a bigger difference to the real problem.