I don't know what I'm missing, but it seems that the 80-pin MicroBit connector is very expensive.
Yes, they do seem pricey.
If the PCB can manage an 80 pin edge, maybe it can also manage a 40 pin PiHeader on the other side ?
That would allow Pi Displays to be usefully added to a board, that lacks any display now.
If you want to use an edge connector, why not go with the SO-DIMM (DDR2 200 pins) one?
Used by e.g. the RPI compute module
And connectors are cheaper and easier to get
I noticed that Ken showed an updated photo of the P2 Edge in the latest video (with Chip discussing sound-related matters). But does anyone have a link to the video in which the P2 Edge was originally mentioned (if one was posted)? And does anyone know of any price speculation, or can anyone venture a range? I believe that Ken said that the P2 Edge might be coming down the pipe towards the end of this year, around the time of a volume production agreement and some other announcements. So I'll wait for all that, but I'm just trying to get up to date on what info is already out there (even if just speculation). Thanks.
I enjoyed the video with an update from Ken and a discussion of Jon McPhalen's work.
As to the cost of the Edge module, Ken said, “As low as we can make it.” That could help get them out there into as many hands as possible and jump start development to build up some P2 momentum.
From the images released so far, there are a few things I have questions about:
Q1: Is there just one 3V3 regulator on the Edge for 8 I/O's, or are there 8 of them, one for each I/O section? If just one, then we could pass in 3V3 through the connector, I assume, for other I/O sections. But could there be 8 (or maybe 4, one for each side of the P2)? It would seem convenient if all 3V3 regs were on the module. The number of 3V3 regs is not clear to me so far.
Q2: Are there two different breakout boards in the works? From the images, it looks like there's a narrow one and maybe a wide one that has eight 12-pin headers (similar to the Eval board). But maybe I'm missing something or conflating the images somehow.
Q3: For the P2 Edge – Breakout image, why does it look like the connector is keyed when the images of the Edge board's micro:bit pads don't seem to be? Perhaps the 3D model is just a placeholder. Sorry, I'm not familiar with the micro:bit connector.
Q4: Does the connector really cost around 10X as much as the connector that Chip referenced? That is, are we really talking $7 to $10 bucks just for a connector? I would hope that a connector could be under $2 bucks. I trust that Chip is involved in this decision-making process, as the success of the module could have quite an impact on the success of the P2. If a connector change is needed, it would seem that the rest of the module could stay the same.
Q5: If a variant of the module was made that had HyperRAM on it, would the connections to the micro:bit connector be the same except for possibly dropping one or more lines that were connected to the HyperRAM chip? I think it would be good to have a variant like that. I recall that there was a design for a P2-Hot module that included a DRAM chip. That suggests that its worth considering adding external RAM. Also, as rogloh mentioned, a variant could be made just by having some modules populated with the BGA part and some not (though that likely would require a somewhat larger module, whether one wanted HyperRAM or not). If I understood rogloh's comment correctly, the HyperRAM chip would float all of its pins if the CS line was not active, making his idea for a possible jumper on the Edge module worth considering. But perhaps it's believed to be too risky to commit to a particular HyperRAM chip at this point, and you likely want to keep things simple in the beginning. If so, that's understandable.
Conclusion: I'm really glad that Parallax is working on such a module. I feel that it is very important to getting the production P2 parts off to a running start. Even if such a module is mostly used by hobbyists and folks doing low-volume production runs, that could still bring the P2 to the attention of some of the big players. After all, if they see that the P2 has been used successfully in various projects and in various ways, then they are more likely to consider using it. At any rate, I think that this thread is an important one and could stand a few more comments.
For experimentation, I would also like to see a P2 variant of the Propeller Project Board USB https://parallax.com/product/32810. Very useful features, leave any further peripherals to the user and maybe lower cost?
I like the look of this P2 Edge for breadboard experiments/embedded development. It might be handy if Parallax also had a version of this board with a HyperRAM BGA footprint included on it, and could optionally then include that part later on a more expensive variant sold with the RAM populated.
I do see Parallax has their 12 way breakout module pins on the breakout accessory with could then potentially attach to a HyperRAM module, but people might like to use this Edge as an embedded device in some other system that goes into an enclosure etc. If they did that and didn't want expansion boards hanging off breakouts etc, then to make use of the high speed RAM with the P2 they'd need to design HyperRAM onto that board which gets complex if the wiring is long, and BGA is hard for average users to play with. Nicer if it's already present.
Keep it in mind anyway. It may not be feasible to include that BGA footprint into the space available, but if you could it would be useful down the track. You could fit a jumper on it's CS pin to enable the RAM only when needed in case you are worried about shorting the memory signals when people don't intend to use it. All its pins would float then.
My thoughts too, except I was thinking of 2 HyperRAM chips with a P2Edge++ designation.
People initially see the cost of the CPU board (Cheaper = Better), and only later discover accessory costs like carrier boards.
It really must be as bare bones as possible, but not feel like something's missing (missing the 1.8V core voltage regulator or boot memory would be in that category).
Oh. but one thing. Reconsider the color of the blue LEDs. They're not cool anymore: white, yellow-green, amber? Fine. But not blue!
I agree there. They are hard on the eyes also. Also, please don't run the LEDs so bright that we have to tape over them, like on the Propeller Demo board.
I think it is probably pretty important to have USB on the P2Edge as well similar to the QuickStart board, so you can just connect it to your laptop via a USB cable similar as you would expect, like on ESP32 host boards, Teensy, Arduino, etc.
I have projects waiting and would buy 10 of them today if they were available. Even if they are not designed with every feature, we can figure it out, we are engineers!
I think the Edge is going to be a great way to get the P2 into commercial projects; the key -- for me, anyway -- is helping developers develop with the Edge. I'd like to propose a P2 Edge Development Board that looks like the attached image. It's not a PDB, but it would help developers create circuits for the Edge while maintaining compatibility with P2 Eval accessories.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Not shown in idea graphic: 5v power supply components and USB interface components.
I think the Edge is going to be a great way to get the P2 into commercial projects; the key -- for me, anyway -- is helping developers develop with the Edge. I'd like to propose a P2 Edge Development Board that looks like the attached image. It's not a PDB, but it would help developers create circuits for the Edge while maintaining compatibility with P2 Eval accessories.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Wow! That's a great idea, Jon. You've got everything covered there. I think we should make it.
I think the Edge is going to be a great way to get the P2 into commercial projects; the key -- for me, anyway -- is helping developers develop with the Edge. I'd like to propose a P2 Edge Development Board that looks like the attached image. It's not a PDB, but it would help developers create circuits for the Edge while maintaining compatibility with P2 Eval accessories.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Wow! That's a great idea, Jon. You've got everything covered there. I think we should make it.
What do you think about also making a version called the P2Edge++ with 2 HyperRAM chips on it for fast double-buffered frame buffer? I don't know how essential it is to be on short traces for best performance, but since they are BGA devices, it's not something most of us can't solder anyways.
I too think it a breadboard based breakout would be great. Bigger, to me, is better when it comes to bread boarding space, so no objection to it being big from me.
I think the Edge is going to be a great way to get the P2 into commercial projects; the key -- for me, anyway -- is helping developers develop with the Edge. I'd like to propose a P2 Edge Development Board that looks like the attached image. It's not a PDB, but it would help developers create circuits for the Edge while maintaining compatibility with P2 Eval accessories.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Wow! That's a great idea, Jon. You've got everything covered there. I think we should make it.
What do you think about also making a version called the P2Edge++ with 2 HyperRAM chips on it for fast double-buffered frame buffer? I don't know how essential it is to be on short traces for best performance, but since they are BGA devices, it's not something most of us can't solder anyways.
I too think it a breadboard based breakout would be great. Bigger, to me, is better when it comes to bread boarding space, so no objection to it being big from me.
Yes, we could put those two HyperRAM chips on and still have P0..P31 free.
Will there be a small PCB for hobbyists, containing just the P2 and decoupling capacitors and maybe a 20MHz oscillator, surrounded by and wired to pins on a 0.1" grid?
Why cant the P1 and P2 be sold in a simple, straight format like the 4809?
Thank you, I have a flip module, but I prefer the idea of
a module that I can solder directly into a project (without
including 32 holes, or a custom pcb with a P1 chip), like
the Arduino Every. Right now, I need something in between
a Flip or a P1 ic for my current project.
Yes, a P2 would be long (about 3 1/2" x 5/8" (89mm x 16)).
That is longer than a 1/2 sized bread board. But it would
be compact. The board could be made from a slightly thicker
fiberglass? to increase stiffness.
At first glance, common sense tells you that is too long, but
someone will make one anyway. Why not try a few prototypes?
Bill M.
As bare bones as possible to keep the $$ down would be great.
Could leave vacant footprints for extra flash or eeprom.
Indeed. You should grab KiCad or another PCB program and demonstrate exactly what you want. I use DipTrace, but any tool that can produce Gerbers will do. When your design is done, send the Gerbers for fabrication. I've been using JLCPCB lately and am happy with the price and turn-around.
Don't count on anyone else wanting what you want, until you show a willingness to build it yourself.
Yes, we could put those two HyperRAM chips on and still have P0..P31 free.
I'd recommend shuffling the expansion pin groups a little. I'm still wary of exposing the prop2 oscillator circuits to dangly wires on expansion connectors. The associated I/O group (P28-P31) can destroy the prop2's boot ability if damaged.
If keeping P0-P31 free is wrong, I don't want to be right, lol. Not only do many instructions span a full port but it also just sounds better to have a full port free. I'm willing to donate a couple of Rev B chips for destructive testing of the oscillator circuit if there's interest. Well, if the risk to the oscillator is high, at least the Edge is a replaceable module. And hopefully the Edge will be stable when plopped into a fixed design. Fingers crossed, anyway.
I'm bringing it up because the Edge with dual HyperRAM has only half the I/O pins exposed. Better to choose the more rugged option.
There was definitely a number of killed Eval Boards due to an internal short on VIO pin (V2831) common to the oscillators. General assumption was this came from spikes on the exposed VIO at the associated accessory header and not from the I/O pins themselves ... but this was never proven in any way as far as I know. Besides, if the I/O is exposed then so will be the associated VIO.
I think the Edge is going to be a great way to get the P2 into commercial projects; the key -- for me, anyway -- is helping developers develop with the Edge. I'd like to propose a P2 Edge Development Board that looks like the attached image. It's not a PDB, but it would help developers create circuits for the Edge while maintaining compatibility with P2 Eval accessories.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Wow! That's a great idea, Jon. You've got everything covered there. I think we should make it.
What do you think about also making a version called the P2Edge++ with 2 HyperRAM chips on it for fast double-buffered frame buffer? I don't know how essential it is to be on short traces for best performance, but since they are BGA devices, it's not something most of us can't solder anyways.
I too think it a breadboard based breakout would be great. Bigger, to me, is better when it comes to bread boarding space, so no objection to it being big from me.
Yes, we could put those two HyperRAM chips on and still have P0..P31 free.
I'm really glad you are looking into this. With a couple of HyperRAMs on a P2Edge, this could be the basis for quite a nice development platform and free us from the bloat & lack of control and security of the PC/Mac/Linux platforms. I like the idea of serving up an IDE from a WROVER to a web browser too, but since I grew up in the 80's, I'm quite biased towards a simpler yet very capable system. I'm sure @Cluso99 and others would agree with me here. I also can't wait to try out TAQOS with some of the demos I have been drooling over from Peter almost 1.5 years ago!
Other features that would be nice, like built in USB2Serial are good to have, especially if a PropPlug won't do the 8Mbit you are playing with, and microSD to host the whole self-hosted dev platform hanging on a single USB connection accessible from a serial terminal. These two might be better for the other side of the edge connector to keep the cost down, and I could even bodge that hardware on with a bit of vero board.
Comments
Yes, they do seem pricey.
If the PCB can manage an 80 pin edge, maybe it can also manage a 40 pin PiHeader on the other side ?
That would allow Pi Displays to be usefully added to a board, that lacks any display now.
Used by e.g. the RPI compute module
And connectors are cheaper and easier to get
Although I agree with @"Peter Jakacki"
As to the cost of the Edge module, Ken said, “As low as we can make it.” That could help get them out there into as many hands as possible and jump start development to build up some P2 momentum.
From the images released so far, there are a few things I have questions about:
Q1: Is there just one 3V3 regulator on the Edge for 8 I/O's, or are there 8 of them, one for each I/O section? If just one, then we could pass in 3V3 through the connector, I assume, for other I/O sections. But could there be 8 (or maybe 4, one for each side of the P2)? It would seem convenient if all 3V3 regs were on the module. The number of 3V3 regs is not clear to me so far.
Q2: Are there two different breakout boards in the works? From the images, it looks like there's a narrow one and maybe a wide one that has eight 12-pin headers (similar to the Eval board). But maybe I'm missing something or conflating the images somehow.
Q3: For the P2 Edge – Breakout image, why does it look like the connector is keyed when the images of the Edge board's micro:bit pads don't seem to be? Perhaps the 3D model is just a placeholder. Sorry, I'm not familiar with the micro:bit connector.
Q4: Does the connector really cost around 10X as much as the connector that Chip referenced? That is, are we really talking $7 to $10 bucks just for a connector? I would hope that a connector could be under $2 bucks. I trust that Chip is involved in this decision-making process, as the success of the module could have quite an impact on the success of the P2. If a connector change is needed, it would seem that the rest of the module could stay the same.
Q5: If a variant of the module was made that had HyperRAM on it, would the connections to the micro:bit connector be the same except for possibly dropping one or more lines that were connected to the HyperRAM chip? I think it would be good to have a variant like that. I recall that there was a design for a P2-Hot module that included a DRAM chip. That suggests that its worth considering adding external RAM. Also, as rogloh mentioned, a variant could be made just by having some modules populated with the BGA part and some not (though that likely would require a somewhat larger module, whether one wanted HyperRAM or not). If I understood rogloh's comment correctly, the HyperRAM chip would float all of its pins if the CS line was not active, making his idea for a possible jumper on the Edge module worth considering. But perhaps it's believed to be too risky to commit to a particular HyperRAM chip at this point, and you likely want to keep things simple in the beginning. If so, that's understandable.
Conclusion: I'm really glad that Parallax is working on such a module. I feel that it is very important to getting the production P2 parts off to a running start. Even if such a module is mostly used by hobbyists and folks doing low-volume production runs, that could still bring the P2 to the attention of some of the big players. After all, if they see that the P2 has been used successfully in various projects and in various ways, then they are more likely to consider using it. At any rate, I think that this thread is an important one and could stand a few more comments.
For experimentation, I would also like to see a P2 variant of the Propeller Project Board USB https://parallax.com/product/32810. Very useful features, leave any further peripherals to the user and maybe lower cost?
My thoughts too, except I was thinking of 2 HyperRAM chips with a P2Edge++ designation.
I agree there. They are hard on the eyes also. Also, please don't run the LEDs so bright that we have to tape over them, like on the Propeller Demo board.
I know it looks really big -- it's not, though; it's just a little bigger than two P2 Eval boards placed side-by-side.
Not shown in idea graphic: 5v power supply components and USB interface components.
Wow! That's a great idea, Jon. You've got everything covered there. I think we should make it.
What do you think about also making a version called the P2Edge++ with 2 HyperRAM chips on it for fast double-buffered frame buffer? I don't know how essential it is to be on short traces for best performance, but since they are BGA devices, it's not something most of us can't solder anyways.
I too think it a breadboard based breakout would be great. Bigger, to me, is better when it comes to bread boarding space, so no objection to it being big from me.
Yes, we could put those two HyperRAM chips on and still have P0..P31 free.
Yeah, please make it. I want a couple. Great idea!
Why cant the P1 and P2 be sold in a simple, straight format like the 4809?
a module that I can solder directly into a project (without
including 32 holes, or a custom pcb with a P1 chip), like
the Arduino Every. Right now, I need something in between
a Flip or a P1 ic for my current project.
Yes, a P2 would be long (about 3 1/2" x 5/8" (89mm x 16)).
That is longer than a 1/2 sized bread board. But it would
be compact. The board could be made from a slightly thicker
fiberglass? to increase stiffness.
At first glance, common sense tells you that is too long, but
someone will make one anyway. Why not try a few prototypes?
Bill M.
As bare bones as possible to keep the $$ down would be great.
Could leave vacant footprints for extra flash or eeprom.
Don't count on anyone else wanting what you want, until you show a willingness to build it yourself.
ends after ~$40.
Maybe swap P24-P31 for P32-P39 or P48-P55.
There was definitely a number of killed Eval Boards due to an internal short on VIO pin (V2831) common to the oscillators. General assumption was this came from spikes on the exposed VIO at the associated accessory header and not from the I/O pins themselves ... but this was never proven in any way as far as I know. Besides, if the I/O is exposed then so will be the associated VIO.
I'm really glad you are looking into this. With a couple of HyperRAMs on a P2Edge, this could be the basis for quite a nice development platform and free us from the bloat & lack of control and security of the PC/Mac/Linux platforms. I like the idea of serving up an IDE from a WROVER to a web browser too, but since I grew up in the 80's, I'm quite biased towards a simpler yet very capable system. I'm sure @Cluso99 and others would agree with me here. I also can't wait to try out TAQOS with some of the demos I have been drooling over from Peter almost 1.5 years ago!
Other features that would be nice, like built in USB2Serial are good to have, especially if a PropPlug won't do the 8Mbit you are playing with, and microSD to host the whole self-hosted dev platform hanging on a single USB connection accessible from a serial terminal. These two might be better for the other side of the edge connector to keep the cost down, and I could even bodge that hardware on with a bit of vero board.
Did that gain any traction? and will a pinout for the edge connector
be released before the Edge is available?
Possibly a connector part# too, if the connectors are not yet available
from Parallax.
Bill M.
Was there any problems with right angle connectors?