.4mm to 2.54mm Routing

Glenn Jones schrieb:

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

https://dl.dropboxusercontent.com/u/2238318/esp32hb.zip

 

Any suggestions?

Manual routing is the key to all autorouter problems.

Seriously.

Tilmann

On 16/09/2016 5:14 p.m., Glenn Jones wrote:

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

https://dl.dropboxusercontent.com/u/2238318/esp32hb.zip

 

Any suggestions?

 

Glenn

Even if you get a trace between the QFN pins you will be up for the

expense of 2mil trace and space. The nature of your design would appear

not to justify that.

Warren

warrenbrayshaw schrieb:

On 16/09/2016 5:14 p.m., Glenn Jones wrote:

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

Even if you get a trace between the QFN pins you will be up for the

expense of 2mil trace and space. The nature of your design would appear

not to justify that.

I didn't download the zip, however from his text I assume he just wants

a track from the QFN to the header, not traces between two QFN pads at

0.4 mm pitch.

Tilmann

On 9/16/2016 2:10 AM, Tilmann Reh wrote:

Glenn Jones schrieb:

 

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

https://dl.dropboxusercontent.com/u/2238318/esp32hb.zip

 

Any suggestions?

 

Manual routing is the key to all autorouter problems.

Seriously.

 

Tilmann

 

I was afraid you were going to say that

Glenn

On 9/16/2016 2:10 AM, Tilmann Reh wrote:

Glenn Jones schrieb:

 

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

https://dl.dropboxusercontent.com/u/2238318/esp32hb.zip

 

Any suggestions?

 

Manual routing is the key to all autorouter problems.

Seriously.

 

Tilmann

 

I was afraid you were going to say that

Glenn

On 9/16/2016 5:33 AM, Tilmann Reh wrote:

warrenbrayshaw schrieb:

 

On 16/09/2016 5:14 p.m., Glenn Jones wrote:

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

Even if you get a trace between the QFN pins you will be up for the

expense of 2mil trace and space. The nature of your design would appear

not to justify that.

 

I didn't download the zip, however from his text I assume he just wants

a track from the QFN to the header, not traces between two QFN pads at

0.4 mm pitch.

 

Tilmann

 

Correct.

Glenn

On 9/16/2016 5:33 AM, Tilmann Reh wrote:

warrenbrayshaw schrieb:

 

On 16/09/2016 5:14 p.m., Glenn Jones wrote:

I seem to be having a heck of a time getting the autorouter to route

between .4mm pitch 48 pin qfn smd and 2 x 2.54mm pth header strips.

 

Even if you get a trace between the QFN pins you will be up for the

expense of 2mil trace and space. The nature of your design would appear

not to justify that.

 

I didn't download the zip, however from his text I assume he just wants

a track from the QFN to the header, not traces between two QFN pads at

0.4 mm pitch.

 

Tilmann

 

Correct.

Glenn

On 16/09/16 17:48, Glenn Jones wrote:

On 9/16/2016 2:10 AM, Tilmann Reh wrote:

 

Manual routing is the key to all autorouter problems.

Seriously.

 

Tilmann

 

 

I was afraid you were going to say that

 

Why afraid? Manual routing - particularly of localised bits - is a lot

easier than setting up the autorouting rules!

CadSoft Guest wrote:

I was afraid you were going to say that

 

Why afraid? Manual routing - particularly of localised bits - is a lot

easier than setting up the autorouting rules!

I agree, manually routing for the most part is much simpler, the autorouter isn't good enough or stable enough to be used for full on routing. For me I can cause it to abruptly exit within minutes of starting to use it, that's just not an acceptable level of stability for me.

Back to the OP's question, I have downloaded the design and have confirmed that for no obvious reason it just won't route from the QFN to anywhere else. I don't know if this is something odd in the library part (I had a look but couldn't see anything) or something relating to the grid, etc. I couldn't see anything on keepout/restrict layers. Unfortunately I can't keep the autorouter open for long enough without crashing for me to do any in depth analysis of why it is not working.

Manual routing of this is simple though if you rotate the IC by 45 degrees and move it down slightly and then all of the signals just flow nicely out to the connectors so not using the autorouter shouldn't be a big problem.

Best Regards,

Rachael

Hi!

You can't manually route to/from the QFN48 as well. Just run a DRC on

the unrouted board and you'll get a ton of errors. The package already

violates the minimum clearance rules. So the autorouter can't reach the

SMDs without violating the rules even further, hence it leaves them out.

So, to solve the problem, first check the DRC parameters. Then go to the

schematic and check the CLASS settings. Finally, ensure that the NETs

attached to ESP32 belong to the proper CLASS.

Rene