 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : RP2040 Successor
| Author | Message | ||||
| PhenixRising Guru  Joined: 07/11/2023 Location: United KingdomPosts: 857 |
"Moreover, it has been developing a successor to the RP2040, which Upton implied will sport a more powerful architecture, additional RAM and an improved General-Purpose Input/Output (GPIO) offering." The link |
||||
Grogster Admin Group  Joined: 31/12/2012 Location: New ZealandPosts: 9306 |
Interesting!  Smoke makes things work. When the smoke gets out, it stops! |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 9110 |
From the Raspberry Pi Investor relations IPO document The interpretation of RP235X is thought to be: 2 : Cores - 2 3 : Core Type - 32-bit Cortex M3 ? 5 : RAM - 512KB plus any stack, scratch and additional banks X : Non-volatile storage undefined |
||||
| Bleep Guru Joined: 09/01/2022 Location: United KingdomPosts: 509 |
And have they sent you one yet for 'stress testing'? as I suspect that MMBasic running flat out, driving large colour LCD displays, USB, SPI, I2C, SD cards, games controllers, Interupts, Sound...... must make very extensive use of almost all of the hardware. :-) Not that you would be allowed to say anyway! ;-) Edited 2024-05-23 04:54 by Bleep |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9306 |
@ Peter - assuming that they do this, would you consider porting the PicoMite FW over to the new chip? I only ask, as I know you have said on the record, that you don't want to do any more MMBASIC ports....and I don't blame you! Would the extra RAM and/or SDRAM, allow for more video memory giving even just 8-colour MODE-1 on the PM? Smoke makes things work. When the smoke gets out, it stops! |
||||
| Amnesie Guru Joined: 30/06/2020 Location: GermanyPosts: 396 |
I am pretty happy with the RP2040, for more demanding tasks there is the CMM2...  We will see how much more the new RP235x will cost. For me this is a huge point to consider, since the RP2040 is next to nothing when it comes to $$$. Greetings Daniel |
||||
| Volhout Guru Joined: 05/03/2018 Location: NetherlandsPosts: 4223 |
Personally I like the CMM2 for it's speed. But I really LOVE the PicoMite for it's huge MMBasic command set, and it's PIO's. And I LOVE to squeeze every bit of performance out of the small thing. But this was only possible because during the last 2-3 years there has been a "pico-fever" in this community. With Peter generating new versions, and 10+ people using them, testing, debugging, issuing idea's. There was a cyclone of energy whirling around this forum. Not a single day went by without someone communicating on the picomite (myself included). I understand that RP have understood the limitations of the RP2040 and increased RAM in the new chip (and changed the ARM's from M0+ to M3) and IO pins. Question is if this is enough to bring the "pico++ fever" back in this community. One thing that may need to happen is HDMI/DVI to future proof the Maximite Line. Volhout Edited 2024-05-23 21:32 by Volhout PicomiteVGA PETSCII ROBOTS |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 3800 |
They could fix the analog stuff! John |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 9110 |
IFF the new chip is pretty much completely compatible from an SDK perspective then I will probably support it. As John says, a key issue is whether they have fixed the ADC. Also, it will be interesting to see if it is an M3 core and if it will overclock like the M0+. If it only runs at stock speeds then it won't be as fast as the overclocked RP2040 and certainly STM32 M3 chips don't overclock anything like the RP2040. The M3 chip has a more complex pipeline structure so code operation timings will probably be less determinant the RP2040. Assuming it is worth doing then I would expect to add two new video modes 640x480x16-colour and 320x240x256-colour. 640x480*256 colour takes too much RAM to be really worth doing assuming like the RP2040 lots of ram is needed to load firmware from the slow external flash. |
||||
| PhenixRising Guru Joined: 07/11/2023 Location: United KingdomPosts: 857 |
On-board DACs (12 bit) would be nice. Our PWM's 0.1% resolution limit is a disappointment (for me at least). I read that FPU is an M3 option but this would also be nice. |
||||
| Volhout Guru Joined: 05/03/2018 Location: NetherlandsPosts: 4223 |
Pwm's are 16 bit, but at 44kHz they are 11 bit. Lower the frequency, or cascade 2 pwm's. A DAC is just as good as it's Vref. Volhout PicomiteVGA PETSCII ROBOTS |
||||
| PhenixRising Guru Joined: 07/11/2023 Location: United KingdomPosts: 857 |
OK, I'm confused  ; I'm @18KHz but I only have a total range of 1000 if the PWM is 0 to 100% with a resolution of 0.1%, right? Or am I missing something? |
||||
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
I can see why you could think that as the manual says:-But I had assumed that meant you could use any float rather than a limit of 0.1. To test it filter the PWM and feed it to a 16 bit DAC and see what happens as you vary the PWM by 0.01%. EDIT The extra detail for servos implies higher resolution. Edited 2024-05-24 08:14 by phil99 |
||||
| Volhout Guru Joined: 05/03/2018 Location: NetherlandsPosts: 4223 |
@Phenix, This is off-topic, but short explanation: A Picomite PWM is in essence 16 bit. It runs on 63MHz (PicoMiteVGA) or 66MHz (PicoMite), essentially half of the designed CPU speed (no overclocking). The counter wraps roughly every 63MHz / 65536 = 1kHz. So if you program a PWM for 1kHz it is 16 bit. If you program a PWM for 18kHz it can count up to 63MHz/18kHz=3500. This is almost 12 bit. The counter counts 0...3500, then wraps to 0 again. The PWM is determined by a hardware compare of the counter. When set to 1000, the PWM output is high from 0...1000, and low from 1000....3500. These are integer numbers. The (floating point) value you use in the MMBasic PWM command calculates the compare value as a percentage of the maximum count (in this case 3500). The floating point variable is 64bit, and far more accurate that the value written in the compare register. But essentially you get 12 bit @ 18kHz, regardless the number of digits you use in the floating point variable. Volhout P.S. You could cascade PWM's (i.e. use 1 PWM at 12 bits, and a second that you program in 4 bits, both running at 18kHz, and sum up the analog voltages of the 2 PWMs with an R-4096R combiner (i.e. 1k to PWM1A and 4.096M to PWM1B). But you need a real stable 3.3V to make this work fine. Edited 2024-05-24 16:56 by Volhout PicomiteVGA PETSCII ROBOTS |
||||
| PhenixRising Guru Joined: 07/11/2023 Location: United KingdomPosts: 857 |
Thanks guys. I'm not dwelling on this for now because I'm only using PWM for my desktop H-Bridge test-rig. I will be using SPI DACs for my +/- 10v motor command. I need to push forward with other developments because I have just committed to a 100% PicoMite control project...Excited |
||||
 Print this page |
