FLiP vs Propstick USB
Rumple
Posts: 38
in Propeller 1
The description of the new FLiP board sort of implies that it is pin and code compatible with the Propstick (with the exception of two 68K pull-downs). Is this true?
If so, great (if only for the power-over-USB feature). If not, what would I have to change?
If so, great (if only for the power-over-USB feature). If not, what would I have to change?
Comments
30 was XI (Xtal input), and with FLiP is NC (Not connected)
31 was NC (Not connected), and with FLiP is USB 5Vout (assuming you are powering from usb!)
If you handle those 2 pins, and also factor in the two 68K pull-downs you mentioned on P26/P27), then your good to drop the FLiP in place of the PropStickUSB.
The FLiP module I2C bus has both clock and data (P28 & P29) pulled up with 3.9K resistors.
PropStickUSB didn't have the clock pin (P28) pullup, and used 10K on P29.
This should be an improvement with FLiP that ought not impact external circuits, unless you've added an external pullup to either of those pins.
Ie. if you previously have added external resistors for I2C, you likely won't need them anymore.
Yes, it's gone from a mini to a micro. That may result in a bit of rewiring in some applications if I ever need to replace an existing Propstick with a FLip. But I have embedded Propsticks in dozens of projects over the last ten years and I haven't yet seen one fail. Although I have managed to blow a few up on the breadboard...
I'm in two minds over this mini-micro thing....
Micro is the new charging standard, so it's become more of a 'go to choice', but it is more fragile. Cables seem to fail first.
I've seen Vendors choose to stick with mini over micro for that reason, and some I notice even use fully thru-hole Minis.
I never use a pure-smd USB part, at the least the mounting pins (prefer 4 over 2) must have holes, and poured copper.
I thought about the USB connection, it would be a shame if it failed.
With the module securely planted in the breadboard, a trip over the cable could rip the thing completely apart.
I've thrown out 2 Prop Plugs since starting the rover project. Both times I kicked the USB cable, or it got caught on something. The USB connector rips the pads off the PCB making it not worth the effort to repair. Now that I can program everything wirelessly, I don't have to deal with the issue.
Wireless programming would solve so many issues, and certainly adds a lot of excitement to working with any mobile robot.
I think this is a really good reason to use Micro. AFAIK, that was also a design goal of the usb micro connector... that the cable would fail first.
ie. much easier to change a cable, than the on-board connector.
One issue is that fully through hole versions add significant cost to a product- especially relevant when retail is <$50. Most customers want lower costs
For higher end (higher cost) products, I don't disagree. A fully through-hole part will be stronger.
That said, for most users the SMT version with 4 through-hole mounting pegs will be as robust as the product lifetime.
Sure, there will always be edge cases. Although breaking these connectors is surely gonna need a lot of force? More than might be reasonably expected? .. Even in that respect, I like the strain relief shown by MikeDYur above. Neat and simple option for "tough" users.
More generally, it's good to see Parallax has been using those robust 4-pegged TH/SMT connectors jmg mentions for a fair while across its product range. Both in Micro and Mini format. I guess the old PropPlug was designed just before the change, but I think Parallax has been careful to choose the most robust connectors.
That said, Sure - Wireless is the future! And a wireless FLiP would be very cool!
If it helps, there are adapters available (Micro USB to Mini USB) that would save replacing your existing usb cable. With regard the USB side, no other re-wiring needed.
After reflow, the holes will fill to 80-90% with solder (IPC-610 requires >75%). Of all the ones we place, I have yet to hear of a damaged connector from normal use. We place them on commercial and industrial products. As long as the posts are soldered at least 50%, they are not weaker than a full TH version.
Full SMT micro USB connectors, the type with gull wing mounting pads, are notorious for ripping off, so you rarely see those on new designs. The DIY field still uses them though.
In some earlier boards I'd used a gull-wing SMD part, no end of problems, because these connectors get lots of use, in and out, twisting, or even hanging by the cable. I'd resorted to adding a strain relief wire on top of the connector, first photo below. Helped some, but takes a 2nd assembly pita. The problems have mostly gone away with the staked connector.
Andrew, I didn't know that was called "intrusive reflow". Okay! One issue with shrouded connectors is that they are difficult to inspect with X-ray or optical. Do you have special tricks for inspection?
µUSB is definitely the ubiquitous option these days.
µUSB is definitely the correct choice nowadays, period. The only reason it won't be, is if the designer doesn't want to take the time to design the board correctly for manufacturing. Besides, nearly everyone has µUSB cables laying around. Mini USB are far less common now.
edit: Here's a decent "cover it all" page for Intrusive Reflow: http://www.mtarr.co.uk/courses/topics/0226_pip/index.html
Martin has some really good info on his site: http://www.mtarr.co.uk/courses/topics/index.asp
The FCI connector data sheet does suggest a pcb layout, attached, and it does in fact specify narrow slots for the the stakes as well as larger pad area toward the centerline. They aren't really pins, as they are only 0.6mm (26 mils) long, so they don't extend very deep into a 62mil circuit board. It looks like the FLIP adheres quite tightly to that layout. The slots are plated through, but no pads on the opposite (smooth) side of the board.
I wasn't comfortable with plated slots, so in my boards I've used holes of 32mil and 40mil diameter for the two sizes of stakes, into which they fit snugly but with clearance at the sides. There is a lot of pad to the side with the notion that solder from the larger pads will wick by surface tension into the hole to help secure the stake. That is one of the suggestions in your mtarr reference. There has been no problem with solder dripping thru, which is good, because there is the entry for an SD card on the other side.
I use such PAD layouts and also copper-pour flood over pads both sides, to give the highest copper adhesion areas.
Think of it as belts and braces....
Good point about highest copper adhesion areas. It doesn't do any good to have a part soldered to a pad if the pad itself lifts off the board. But about USB connectors, I believe the shell is not supposed connect to the end circuit ground. It is supposed to connect only to the cable shield. I see it is connected to Vss on the FLIP, no big deal.