I have to ask: Are SMT pin headers problematic? I rarely see them in the wild. I've been designing a P2 module that should fit into most applications using the FLIP. Obviously when I hand solder the prototypes I'll need a fixture to hold the two headers with the right spacing. But could they shift during reflow soldering?
@SaucySoliton said:
I have to ask: Are SMT pin headers problematic?
Usually yes. You really want to order the ones with pegs.
If not, slim down (narrow down) the pads of some of the middle pins in the row to help surface tension keep the headers straight. Pegs is really the way to go though.
My own rationale about these modules is this: such modules are not targeting people who split the cent in two when calculating a design cost, but rather those who target reliable and low-maintenance operation in an industrial or similar environment. I am not saying that cost is not an important factor, but for non-DIY and hobby stuff it is probably of lower importance than things like reliability or compatibility. That's why I don't normally look seriously at DIL-style modules - those are not reliable enough in socket and also compromising on size to keep two-row pins. My personal experience and tests over a number of years have shown that nothing comes close to the PLCC sockets in terms of balancing reliable operation in the field and overall size.
With my contact, boards come to under $3 in qty of 1k. That for my design of the Stamp I mean, which is 8 layers and 70um copper on all layers and gold finish. They have manufactured such and similar boards for me in the past. For a simpler design those prices will likely fall further.
Not going to publish it here but can give it to Parallax in a DM should they want to talk to that company.
@Rayman the two words together comprise our registered trademark BASIC Stamp. Stamp is not a trademarked name by Parallax, nor could it likely be approved as one.
However, it could pose a problem for us if the term P2 Stamp were used and we decided in the future to create a Stamp module from the Propeller 1 or 2. Market confusion, your prior work with the name, etc are all examples. Therefore, we would prefer you choose a different name for the product.
Going to build some this weekend with out RTC, flash, and HyperRam.
These things are often needing rework to fix.
Got the thinner stencil and that might fix this issue, but a lot of applications don't need these things, or might want different kinds.
Think will leave uSD on though. That can't really fail and allows for booting without anything else...
At least, think it can boot from uSD. Haven't actually tried that yet.
Really hoping the thinner stencil saves rework on the protection mosfets.
The solder syringe was able to just get enough solder through the stencil to work, but that is a pain.
@Rayman said:
Hmm... Works with SpinTools IDE with uSD installed.
Maybe messed up Prop Tool somehow...
Hmm. Doesn't work with PNut with uSD installed either...
Guess it's in same boat as Prop Tool...
First guess here is that it could be reset timing pulse differences in the loader SW along with pullup/downs seen on P61,P60,P59. Maybe P59 is being pulled down and/or locking out the serial window? If you pull up P59 weakly does the problem disappear?
If the SD boot option isn't disabled (Disable with a pull-up on P59) then a bootable SD card has only a 100 ms timing window for serial to get control first. It's not surprising that downloader programs haven't been well tuned to catch this correctly. Both Eric and Macca have belatedly put extra effort into getting it right in their respective downloaders.
Something must be pulling P59 high in full build, but don't know what.
Don't see any change in the P59 voltage when uSD is present or not, very strange...
Seems the Boot1 option resistor (R2) has been left off in the "core" build. Could fix that and force serial only boot, but not liking that...
Thinking that maybe the "core" build should include the flash chip.
Maybe that would fix whatever is going on here...
Comments
[deleted by JM]
@JonnyMac yeah, don’t think my operation is much of a threat…
I’m liking SWAP though
Just learned what SWAP means in electronics parlance a few years ago and it fits nicely here...
Here's an interesting thing...
https://www.technologicalarts.com/products/plcc-adapter-84-pin
Says the two rows are spaced 0.3" apart, which guess means will work on breadboard...
I have to ask: Are SMT pin headers problematic? I rarely see them in the wild. I've been designing a P2 module that should fit into most applications using the FLIP. Obviously when I hand solder the prototypes I'll need a fixture to hold the two headers with the right spacing. But could they shift during reflow soldering?
Maybe this VLA module is not such an expensive idea after all. https://forums.parallax.com/discussion/comment/1566654/#Comment_1566654 Only place I've seen smt headers is these modules: http://janus-rc.com/CF-embedded.html
The end section is a test to figure out how wide an 80 pin module would be if using SMT headers.
Tried reflow once with headers thinking they might self align… they didn’t …
So probably need something to help…
Usually yes. You really want to order the ones with pegs.
If not, slim down (narrow down) the pads of some of the middle pins in the row to help surface tension keep the headers straight. Pegs is really the way to go though.
My own rationale about these modules is this: such modules are not targeting people who split the cent in two when calculating a design cost, but rather those who target reliable and low-maintenance operation in an industrial or similar environment. I am not saying that cost is not an important factor, but for non-DIY and hobby stuff it is probably of lower importance than things like reliability or compatibility. That's why I don't normally look seriously at DIL-style modules - those are not reliable enough in socket and also compromising on size to keep two-row pins. My personal experience and tests over a number of years have shown that nothing comes close to the PLCC sockets in terms of balancing reliable operation in the field and overall size.
With my contact, boards come to under $3 in qty of 1k. That for my design of the Stamp I mean, which is 8 layers and 70um copper on all layers and gold finish. They have manufactured such and similar boards for me in the past. For a simpler design those prices will likely fall further.
Not going to publish it here but can give it to Parallax in a DM should they want to talk to that company.
@Rayman the two words together comprise our registered trademark BASIC Stamp. Stamp is not a trademarked name by Parallax, nor could it likely be approved as one.
However, it could pose a problem for us if the term P2 Stamp were used and we decided in the future to create a Stamp module from the Propeller 1 or 2. Market confusion, your prior work with the name, etc are all examples. Therefore, we would prefer you choose a different name for the product.
Thank you for asking us!
Ken Gracey
"P2 Stamp" is a hard name to beat and it fits with Parallax.
Going to call the @knivd one "P2 SWaP" for:
Size
Weight
and
Power
Which are needs that this module can address.
Going to build some this weekend with out RTC, flash, and HyperRam.
These things are often needing rework to fix.
Got the thinner stencil and that might fix this issue, but a lot of applications don't need these things, or might want different kinds.
Think will leave uSD on though. That can't really fail and allows for booting without anything else...
At least, think it can boot from uSD. Haven't actually tried that yet.
Really hoping the thinner stencil saves rework on the protection mosfets.
The solder syringe was able to just get enough solder through the stencil to work, but that is a pain.
New chips from parallax were packaged nicely
They can probably fit eight chips in the box the way they’re doing it now
Yep, P2 SWAP is distinctive. I like it.
Started testing with uSD and seeing something strange...
Seems FlexProp can load with uSD inserted, but Prop Tool cannot.
Have to figure this one out, very strange...
Hmm... Works with SpinTools IDE with uSD installed.
Maybe messed up Prop Tool somehow...
Hmm. Doesn't work with PNut with uSD installed either...
Guess it's in same boat as Prop Tool...
First guess here is that it could be reset timing pulse differences in the loader SW along with pullup/downs seen on P61,P60,P59. Maybe P59 is being pulled down and/or locking out the serial window? If you pull up P59 weakly does the problem disappear?
If the SD boot option isn't disabled (Disable with a pull-up on P59) then a bootable SD card has only a 100 ms timing window for serial to get control first. It's not surprising that downloader programs haven't been well tuned to catch this correctly. Both Eric and Macca have belatedly put extra effort into getting it right in their respective downloaders.
This is with the core build where I left off a bunch of parts. Maybe I left something I shouldn’t have….
Not seeing pullup on P59 anywhere... Guess that's something user needs to add...
Yeah, pullup on P59 lets it work with PropTool when uSD is inserted. Wish new about this earlier, but oh well...
Something must be pulling P59 high in full build, but don't know what.
Don't see any change in the P59 voltage when uSD is present or not, very strange...
Seems the Boot1 option resistor (R2) has been left off in the "core" build. Could fix that and force serial only boot, but not liking that...
Thinking that maybe the "core" build should include the flash chip.
Maybe that would fix whatever is going on here...
Or Chip sorts out Parallax's downloader so it can intervene in that 100 ms pre-boot window when a bootable SD card is present.