I lost track of the conversation today-- what components are you using for your updated design?Quick update before I go to bed: I may just get all surface mounted parts from Farnell, but it looks like they only take credit cards. I’ll see…
If so, I’ve found the same chips in SOIP (and QFN for the ATmega), so I imagine the circuit should be identical.
NleahciM:
Yes - the fact that it worked once guarantees that nobody will ever have any issues with going completely outside of spec. Great advice. It’s not like there is [tons of information about proper USB trace routing available on the Internets.[/quote]Well, for the 90 ohm differential impedance lines on a 4-Layer board it’s easy, 7.5mil trace width, 7.5mil trace spacing with the ground plane on the 2nd layer. On a 2-Layer board to keep the differential impedance within spec, the trace width and spacing becomes impractically large due to the additional FR4 material between the traces and ground plane.
In this case on 1oz, 2-Layer, 62mil thick FR4, for your D+ and D- lines to have the trace impedance of 45 ohms, they need to be a rather large 126.89mil wide. To have the 90 ohm differential impedance you need a WHOPPING 787.4mil spacing between your D+ and D- lines. Somehow I think, in this case, keeping to the spec is a touch impractical.
Note Since the figure for the spacing between the D+ and D- lines is so big, here is the site I used to calculate my figures.
The units required are metres, so the figures were converted to/from mils to metres.
Individual Trace Impedance Calculator : http://www.technick.net/public/code/cp_ … microstrip
Width : 0.003223m (3.223mm, 126.89mil)
Height Above Return Plane : 0.00157m (1.57mm, 62mil, BatchPCB’s FR4 thickness)
Trace Thickness : 0.00003556m (0.03556mm, 1.4mil, 1oz copper)
Relative Permittivity : 4.7 (Dielectric constant of FR4)
Differential Impedance Calculator : http://www.technick.net/public/code/cp_ … strip_diff
Characteristic Impedance : 45.0032102500151 ohms (Output from previous calculator)
Height Above Return Plane : 0.00157m (1.57mm, 62mil, BatchPCB’s FR4 thickness)
Space Between Traces : 0.02m (20mm, 787.4mil)](http://www.usb.org/developers/docs/hs_usb_pdg_r1_0.pdf)
I'm planning on using an HPIC B 595 (can't remember the exact number) for the (common) cathodes, two HC74595s for the anodes and an ATmega168 for the main control. The USB chip is the FT232RL.Vraz:
I lost track of the conversation today-- what components are you using for your updated design?
NleahciM:
Generally speaking, when routing two layer boards that have a ground layer, you try to route as many traces as possible on one layer so that you have a nearly solid plane on the other layer.
Ah, okay. I’ll have a go with that using SMT components, but I don’t know how easy it’l be.
I did google it, and read the wikipedia article on both that and microstrips, but neither really seemed to answer my question in the context of laying PCB tracks for USB signals. I guess I'll keep looking...Life lesson here: when you don’t know what something means (ie “differential pair”) google it. You’ll get the information faster, learn more, and irritate fewer people.
Sadly, at the moment I believe most of them are probably more accurate than anything I could make myself until I get used to Eagle a bit more, but I am trying to make as much as possible myself.Eagle footprints are, in general, terrible. Make your own. Always.
I am printing the circuit in 1:1 (which I’ve checked) and making sure the footprints line up with components under magnification, which I’d hope should alleviate major problems.
I would not recommend routing under the NC pins, there is not enough information in the datasheet. If do find the need for the space, what are you planning to do, remove the pins?Random:
So, with regards to my previous questions, can I route tracks under the NC pins, and do all the grounds or just one need to be connected?Thanks!
Yes, you should connect all the grounds, they are high current paths off the die. The chip will not meet the current specs if they are not connected.