Mine arrived, and sadly I'm not able to do much with it just yet (contract running longer than expected), and my wife had it in her hands, "So? Vat is dis?" inquisition style.
I grinned and said it's an advance on that larger contract payment. Problem solved.
Now I just gotta get the ugly work done quick. Looks like Chip is getting close!
On the second pass(board), px worked fine, but again I was mostly unable to load a program onto the resulting P1V.
I did get my spin program running a couple of times after fiddling with the usb lines. Once working, it looked stable, through several load cycles.
I could not identify a pattern that lead to occasional success and I had mostly fails. I had to quit last night and won't be able to play with it today. Before quitting, the usb driver was failing with a code 10... something about a file object waiting. I could get px to function again by trashing the driver and rebooting. But the error then returned a couple of times.
I wonder if main power needs to be on, before plugging in the usb line? And whether NOT doing this might damage the board? ... Is power cycling with the usb line in place an issue?
I am also wondering if Port settings have any effect... I didn't see any but right about the time I was starting to look at it, I started to have driver issues.
I won't know for another week. I took a side job that will pay well, but... you all know how these things can go.
Rjo, you really should try another machine. Or, do you have a spare disk? Install it, and load your OS onto it plain vanilla to test. Disconnect the current one entirely.
A family member got something so sneaky, I was only able to discover it this way. Kept thinking it was some issue or other.
According to this post earlier in the thread, international shipping is $20.
I would imagine there are some left - out of the 30, there maybe 10 out in the wild unless a number of peoples have bought them "secretly" without saying anything in this thread.
A quick email to the sales folks shoudl be able to get that question answered.
Today I was getting things running on the new Prop 1-2-3 FPGA board and I ran into a real problem.
I couldn't get the PLL's to work.
After trying everything I could think, it dawned on me that maybe there was some pin that we didn't connect properly on the FPGA. This happened initially with the VBAT pin and it prevented the FPGA from loading. It turns out that pin A1, which (very fortunately) is the ball closest to the corner indicator on the package, needs to be connected to a 2.0K resistor which goes to GND. It was not connected to anything on the -A7 board, as it was one of 8 pins we believed were optional no-connects. Google found me an Altera document that said otherwise today.
I managed to find a 1k resistor and I soldered a single strand of stranded wire onto it. I plugged the resistor into a nearby header and bent the wire strand so that it pushed against the A1 ball on the FPGA. This got the PLL's working.
We were about to run the -A9 boards on the pick and place tomorrow or Wednesday, but now we realize that we must rev the PCB first. There are also some benign VREF pins on the FPGA that will get grounded. Also, some pi filters will be made for the PLL power pins.
This a pain! These FPGA's are full of subtle gotcha's. When we got the -A7 board working, I tested out every connection to I/O's, but just assumed the PLL's would be fine, since their power was connected. Turns out on these Cyclone V devices there are more pins to mind than on earlier devices.
Anyway, this sets the -A9 boards back a few weeks and we need to figure out how to handle all your -A7 boards that have non-working PLL's. To make those boards work, two things can be done: solder a resistor and a wire together like I did, or just feed in a clock of the desired frequency into the SMA connector. Neither is pretty. For those of you with -A7 boards, we can make a clock generator that plugs onto the board somewhere and send you one, or we can offer you a new -A9 board for, say, $100. You can keep your old -A7. I don't know what else to propose. Is one better than the other?
OK... let's say hypothetically that I put my P123 on a conductive surface... which I very well know not to do....
and I order another p123... not caring about the LED's. Do I get a pre-surgical or post surgical unit?
It's probably five weeks out. Daniel will be making the PCB changes this weekend. He went back to school in Texas, so we don't have him full time, anymore.
Well, now that the P2 code for FPGA is available, I am clear that the BeMicroCVA9 may never fit into the scheme of things as the actual Verilog code won't be shared.
I am okay with that. I am also happy that developers have boards that will allow them to move ahead with software while the P2 works closer to production. I can look forward to the day when the actual P2 goes on sale.
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
Well, now that the P2 code for FPGA is available, I am clear that the BeMicroCVA9 may never fit into the scheme of things as the actual Verilog code won't be shared.
I am okay with that. I am also happy that developers have boards that will allow them to move ahead with software while the P2 works closer to production. I can look forward to the day when the actual P2 goes on sale.
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
Our plan is to make P2 open source. We need to get the design ironed out before we do that, though. The BeMicro A9 can be supported. I just need to get ahold of one.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
I know what you mean. I have to run Win7 in a VM under VirtualBox on my Mac. There is a Linux version of Quartus though. Have you tried that?
Well, now that the P2 code for FPGA is available, I am clear that the BeMicroCVA9 may never fit into the scheme of things as the actual Verilog code won't be shared.
I am okay with that. I am also happy that developers have boards that will allow them to move ahead with software while the P2 works closer to production. I can look forward to the day when the actual P2 goes on sale.
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
Our plan is to make P2 open source. We need to get the design ironed out before we do that, though. The BeMicro A9 can be supported. I just need to get ahold of one.
Well, I could give you one. I will just order it from Arrow and have it sent to Parallax. But I am NOT in any rush to deal with the P2. Keep your developers happy.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
I know what you mean. I have to run Win7 in a VM under VirtualBox on my Mac. There is a Linux version of Quartus though. Have you tried that?
I finally got the Linux version of Quartus II v15.0.2 running right on my Intel Quad 64bit, and suddenly the computer is unstable -- appears to be the power supply is resetting and forcing a reboot every minute or two.
If you do indeed desire to run Quartus II in Linux, it seems REHL, Fedora, CentOS, or maybe even SUSE will be easier to get installed. The Debian side of Linux seems to not work well with the installer software, at least not Debian Jessie 8.1 amd64.
I have spent about a month creating a triple boot on my computer to explore how to best install Quartus II. I am now happy with it in Fedora 22. But I guess I should run out and buy a new power supply before I do anything more.
.....
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
.....
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
Yeah, it is back to $150. I saw that after I posted the higher figure.
I have been following the BeMicroCV and BeMicroCVA9 prices and availability as I wa considering buying another BeMicroCV.
Arrow's web site can be a bit odd at time. They seem to have typos creep in.
They did have the CVA9 for $219.00 the previous time I looked. and the CV(A2) which is backordered had two prices - $49 and 42.19. Now, I suppose this was all something wrong in their database.
++++++++++++
Even with the Quartus II v15.0.2 installed correctly in Linux, everyone will need Windows to use the additional software executibles for the P2 on the P 1-2-3 FPGA from Parallax. Expect to at least install WINE.
Personally I would rather see the Propeller 1-2-3 FPGA-A9 effort go forward and the BeMicroCVA9 code to come later. It seems that is the shortest possible timeline to getting the P2 to market. I have a whole separate FPGA project to explore in the meanwhile that is teaching me VHDL, Verilog, and how to use the software. The BeMicro boards all me to learn Quartus II, and the P1V.
Jac Goldsmit seems to have something in mind for the P2 on the CVA9, and it is very important to resolve the i/o pins in a good way. One of the CVA9 40pin connectors (J4) is not fully used, while the 80 pin edge card connector seems like it would be needed to make a P2 on the CVA9 really feature rich. Those added features are going to require a custom board that may never be as useful as the Propeller 1-2-3 FPGA.
So it seems to be a potential development fork that just distracts everyone.
Personally I would rather see the Propeller 1-2-3 FPGA-A9 effort go forward and the BeMicroCVA9 code to come later. It seems that is the shortest possible timeline to getting the P2 to market.
??
Reality: The 1-2-3 FPGA-A9 is not even available yet, whilst users have the BeMicroCVA9 in their hands now.
If you really want the shortest time-line, the not-yet-available 1-2-3 FPGA-A9 is not the right answer.
Parallax wants the widest test coverage, and the XX-A9 gives that - the build differences between differing A9 boards are not that great, and are a set-up-once type of difference.
Well, if having the BeMicroCVA9 provide Beta testing does work for Parallax, I am all for it. I just don't feel that I know enough to be a worthwhile part of the Beta testing. I don't want to be a distraction. And it certainly seems to be predominantly a Windows-based effort at this point in time.
While the BeMicroCVA9 appears to have lots of i/o in the form of J1, J4, and the 80 pin edge connector; J4 is mostly empty -- not the same as the BeMicroCV-A2.
Tranquilo, Loopy. Chip isn't supporting the BeMicroCVA9 just for you. I rather doubt you even entered his mind when he declared a few weeks ago that the BeMicro CVA9 would be one of the supported boards.
Tranquilo, Loopy. Chip isn't supporting the BeMicroCVA9 just for you. I rather doubt you even entered his mind when he declared a few weeks ago that the BeMicro CVA9 would be one of the supported boards.
We'd do it for Loopy!
I will make it a point to acquire one of these soon so I can make a file for it.
I'm trying to decide between the A7 and A9 version... Is the A7 version with resistor fix now available? Now that there's a new P2 image, what will be the cog# and RAM differences between A7 and A9?
Comments
Mine arrived, and sadly I'm not able to do much with it just yet (contract running longer than expected), and my wife had it in her hands, "So? Vat is dis?" inquisition style.
I grinned and said it's an advance on that larger contract payment. Problem solved.
Now I just gotta get the ugly work done quick. Looks like Chip is getting close!
I did get my spin program running a couple of times after fiddling with the usb lines. Once working, it looked stable, through several load cycles.
I could not identify a pattern that lead to occasional success and I had mostly fails. I had to quit last night and won't be able to play with it today. Before quitting, the usb driver was failing with a code 10... something about a file object waiting. I could get px to function again by trashing the driver and rebooting. But the error then returned a couple of times.
I wonder if main power needs to be on, before plugging in the usb line? And whether NOT doing this might damage the board? ... Is power cycling with the usb line in place an issue?
I am also wondering if Port settings have any effect... I didn't see any but right about the time I was starting to look at it, I started to have driver issues.
If you don't have this happening, I'm thinking that I may have a virus, which hasn't been exactly killed yet.
rich
Rjo, you really should try another machine. Or, do you have a spare disk? Install it, and load your OS onto it plain vanilla to test. Disconnect the current one entirely.
A family member got something so sneaky, I was only able to discover it this way. Kept thinking it was some issue or other.
According to this post earlier in the thread, international shipping is $20.
I would imagine there are some left - out of the 30, there maybe 10 out in the wild unless a number of peoples have bought them "secretly" without saying anything in this thread.
A quick email to the sales folks shoudl be able to get that question answered.
Hope you join us for the fun!
I only skimmed through the topic and missed $20 for intl.
I shall write back in a couple of days after it has been ordered
Frank
22 in stock as of 2:20 EST.
Could someone from Parallax post a working .rbf file of the P1V for the P123 board?
Thanks,
Rich
Ozpropdev ran my .rbf and it worked fine... so that's not the issue:)
I couldn't get the PLL's to work.
After trying everything I could think, it dawned on me that maybe there was some pin that we didn't connect properly on the FPGA. This happened initially with the VBAT pin and it prevented the FPGA from loading. It turns out that pin A1, which (very fortunately) is the ball closest to the corner indicator on the package, needs to be connected to a 2.0K resistor which goes to GND. It was not connected to anything on the -A7 board, as it was one of 8 pins we believed were optional no-connects. Google found me an Altera document that said otherwise today.
I managed to find a 1k resistor and I soldered a single strand of stranded wire onto it. I plugged the resistor into a nearby header and bent the wire strand so that it pushed against the A1 ball on the FPGA. This got the PLL's working.
We were about to run the -A9 boards on the pick and place tomorrow or Wednesday, but now we realize that we must rev the PCB first. There are also some benign VREF pins on the FPGA that will get grounded. Also, some pi filters will be made for the PLL power pins.
This a pain! These FPGA's are full of subtle gotcha's. When we got the -A7 board working, I tested out every connection to I/O's, but just assumed the PLL's would be fine, since their power was connected. Turns out on these Cyclone V devices there are more pins to mind than on earlier devices.
Anyway, this sets the -A9 boards back a few weeks and we need to figure out how to handle all your -A7 boards that have non-working PLL's. To make those boards work, two things can be done: solder a resistor and a wire together like I did, or just feed in a clock of the desired frequency into the SMA connector. Neither is pretty. For those of you with -A7 boards, we can make a clock generator that plugs onto the board somewhere and send you one, or we can offer you a new -A9 board for, say, $100. You can keep your old -A7. I don't know what else to propose. Is one better than the other?
Sorry about this pain in the rear.
Here is a picture of my resistor/wire fix:
and I order another p123... not caring about the LED's. Do I get a pre-surgical or post surgical unit?
I think I have a quick and easy solution. Please see my discussion in Chip's initial thread about this.
Rich
Thanks,
Chris Wardell
C.W.
Yes. If you wanted to buy one, we could fix that A1 ball with the resistor.
I am okay with that. I am also happy that developers have boards that will allow them to move ahead with software while the P2 works closer to production. I can look forward to the day when the actual P2 goes on sale.
I will just continue to use the BeMicroCVA9 to explore the P1V and its adaptations. BTW, the price went up - no longer $150USD, it is $219USD. So the Propeller 1-2-3/A9 FPGA looks a bit more attractive. The BeMicroCVA9 just won't keep up.
It has also become very obvious that exploring FPGAs is more easily done in Windows. Linux support is likely to never be equal to what Altera and other FPGA companies provide in Windows.
Our plan is to make P2 open source. We need to get the design ironed out before we do that, though. The BeMicro A9 can be supported. I just need to get ahold of one.
Well, I could give you one. I will just order it from Arrow and have it sent to Parallax. But I am NOT in any rush to deal with the P2. Keep your developers happy.
I finally got the Linux version of Quartus II v15.0.2 running right on my Intel Quad 64bit, and suddenly the computer is unstable -- appears to be the power supply is resetting and forcing a reboot every minute or two.
If you do indeed desire to run Quartus II in Linux, it seems REHL, Fedora, CentOS, or maybe even SUSE will be easier to get installed. The Debian side of Linux seems to not work well with the installer software, at least not Debian Jessie 8.1 amd64.
I have spent about a month creating a triple boot on my computer to explore how to best install Quartus II. I am now happy with it in Fedora 22. But I guess I should run out and buy a new power supply before I do anything more.
https://parts.arrow.com/item/detail/arrow-development-tools/bemicrocva9#MFez
Yeah, it is back to $150. I saw that after I posted the higher figure.
I have been following the BeMicroCV and BeMicroCVA9 prices and availability as I wa considering buying another BeMicroCV.
Arrow's web site can be a bit odd at time. They seem to have typos creep in.
They did have the CVA9 for $219.00 the previous time I looked. and the CV(A2) which is backordered had two prices - $49 and 42.19. Now, I suppose this was all something wrong in their database.
++++++++++++
Even with the Quartus II v15.0.2 installed correctly in Linux, everyone will need Windows to use the additional software executibles for the P2 on the P 1-2-3 FPGA from Parallax. Expect to at least install WINE.
Personally I would rather see the Propeller 1-2-3 FPGA-A9 effort go forward and the BeMicroCVA9 code to come later. It seems that is the shortest possible timeline to getting the P2 to market. I have a whole separate FPGA project to explore in the meanwhile that is teaching me VHDL, Verilog, and how to use the software. The BeMicro boards all me to learn Quartus II, and the P1V.
Jac Goldsmit seems to have something in mind for the P2 on the CVA9, and it is very important to resolve the i/o pins in a good way. One of the CVA9 40pin connectors (J4) is not fully used, while the 80 pin edge card connector seems like it would be needed to make a P2 on the CVA9 really feature rich. Those added features are going to require a custom board that may never be as useful as the Propeller 1-2-3 FPGA.
So it seems to be a potential development fork that just distracts everyone.
??
Reality: The 1-2-3 FPGA-A9 is not even available yet, whilst users have the BeMicroCVA9 in their hands now.
If you really want the shortest time-line, the not-yet-available 1-2-3 FPGA-A9 is not the right answer.
Parallax wants the widest test coverage, and the XX-A9 gives that - the build differences between differing A9 boards are not that great, and are a set-up-once type of difference.
While the BeMicroCVA9 appears to have lots of i/o in the form of J1, J4, and the 80 pin edge connector; J4 is mostly empty -- not the same as the BeMicroCV-A2.
We'd do it for Loopy!
I will make it a point to acquire one of these soon so I can make a file for it.
Is the A7 version with resistor fix now available?
Now that there's a new P2 image, what will be the cog# and RAM differences between A7 and A9?