USB Micro-B Connector (Receptacle) - DFM Considerations

[MixedSignalSystems]

Hi all,

I'm working on a product at the moment that will include a USB Micro-B Receptacle (aka USB Micro Connector).

I'm choosing an appropriate connector at the moment and something has caught my attention.

Most, if not all, USB Micro connectors are surface mount components. However, many of the USB Micro connectors use additional through-hole tabs to secure the component to the PCB.

When using these components, is the assumption that the PCB will be wave soldered or that the through-hole tabs will be hand soldered?

The product that I'm working on is likely to run to a few thousand units - too many to hand solder and perhaps too few for wave soldering.

I anticipate that a standard reflow process will be used. Does it make sense to stick to a component without any tabs (such as the 047346-0001 from Molex) or am I missing something?

Molex 047346-0001 details:
http://www.molex.com/molex/products/datasheet.jsp?part=active/0473460001_IO_CONNECTORS.xml

Many thanks,
Kieran

[Precipice]

When using these components, is the assumption that the PCB will be wave soldered or that the through-hole tabs will be hand soldered?

Paste-in-hole does the job - and you definitely want all the tabs and hold-down assistance you can possibly get. Glue, case pressing down from the top, PCB tracks approaching the pads from both sides if possible.

While Micro-USB trumpeted it's lifetime improvement over USB-mini, that was seemingly for the actual contact pins. The connectors themselves are a bloody liability. If actual users are going to use the product (and drop it with USB cables plugged in) then brace for product returns.

Any advice and super-good connectors would be welcome. I really don't mind if they're bigger than standard, with massive mounting lugs - just something with a chance of survival.

[sacherjj]

We also looked at potting the micro-USB connectors for extra strength, until we finally just abandoned them all together.  There are epoxy and UV cured potting that won't flow through the small openings in the jack and provide the physical strength boost they need.  micro-USB 3.0 are even worse.  The cables are thicker and put more force on the connector and the span it has to handle if further.

I can't wait for USB 3 Type-C to become main stream with dual ended, non-directional cables.

We couldn't make a scenario where they are likely to last for 5-10 years of use for charging, so we are not just using good old barrel plug.

[mikeselectricstuff]

Automated selective soldering may be another option, but if you can do it with paste that's the easiest.

[MixedSignalSystems]

Hi guys,

Thank you for your thoughts and advice.

Plenty for me to think about. I may go for a USB micro connector with tabs and use the solder paste approach for production.

I'll need to talk nicely to the industrial designer to see if I can add some support through the plastics.

Cheers,
Kieran

[westfw]

There was a related discussion on AVR Freaks a while ago:  http://www.avrfreaks.net/forum/caution-flimsy-atmel-ice
Recommendations included:
KyCon KMMX-BSMT35S-BTR: http://anacapa.kycon.com/gz/website/Products/USB/usb1.html
Hirose Connectors in their ZX series of Micro-USB connectors.  part number ZX62D-B-5PA8
FCI 10103592
These are connectors with "long" tabs, where "long" means about 0.8mm (still not enough to go all the way through most boards.)

[Bassman59]

While Micro-USB trumpeted it's lifetime improvement over USB-mini, that was seemingly for the actual contact pins. The connectors themselves are a bloody liability. If actual users are going to use the product (and drop it with USB cables plugged in) then brace for product returns.

This is completely true. The Micro-USB B connector is a total piece of shit, with a lifetime of maybe a hundred mating cycles (based on my testing) before a cable won't stay in place and your phone won't charge. You may hate the Apple Lightning cable for its cost and "non-openness" or simply because it's from Apple, but it can't be plugged in wrong and seems a whole lot more robust.

[ehughes]

I have a product out in the wild (for Agriculture Purposes) where the customer wanted micro-USB.

You ABSOLUTELY want one with a through hole mounting connections.   Our 1st rev had the SMT pads.   No amount of glue, potting, epoxy etc. with survive the Gorillas out in the wild.

At your quantity,   having selective solder will be a minimal cost.   If your CM doesn't have selective solder,  get a new CM.

Micro-USB is shit for reliability.   The line that it is better than mini-B is 100% bogus.  The standards body who made that statement obviously never supported anything in the field.

[Precipice]

I don't know if you could do something like mount the connector on a little PCB (or flex ribbon?), and _let_ it move? I mean, it's going to move, whether you like it or not - so try to go with the flow?

I've been using Molex HSSTAC connectors recently for automotive jobs - they're a pretty decent implementation of hardened mini-USB, not sure they'd be well received as a constantly plugged/unplugged connector, but as a way of shifting Gbit+ video around, they did the job. No idea if they'll ever gain a micro-USB variant - might make sense for the increased current, and the shell would avoid a lot of the grumbles that come with the naked connector.

http://www.molex.com/molex/products/family?key=hsautolink&channel=products&chanName=family&pageTitle=Introduction

[MixedSignalSystems]

You ABSOLUTELY want one with a through hole mounting connections.   Our 1st rev had the SMT pads.   No amount of glue, potting, epoxy etc. with survive the Gorillas out in the wild.

At your quantity,   having selective solder will be a minimal cost.   If your CM doesn't have selective solder,  get a new CM.

Thank you. I've come to that conclusion that the SMT only option just isn't going to be worth the risk. Also, my customer isn't entirely opposed to using Mini-USB and the more I look into this, that might be the way forward for me.