#avr Logs
May 23 2020
#avr Calendar
12:06 AM day_ is now known as day
06:35 AM Smidge204: Does anyone here have experience making/using STimuli files for Atmel Studio simulations?
06:36 AM nohit: whaddaya think
06:37 AM nohit: you keep asking about atmel studio
06:41 AM Smidge204: :D
06:41 AM Smidge204: "Keep" asking is a bit harsh; the number of times I've asked about anything *specific* about the IDE is pretty small!
06:42 AM matrixstorm: one of the great advantages of AVR is its open source gcc integration
06:42 AM matrixstorm: don't use windows
06:45 AM Smidge204: I don't think the IDE is really relevant here, I only mention it because I'm not sure if there's anything special about the stimuli files it uses
06:54 AM nohit: and one of the great disadvantages of AVR is that its dead
06:55 AM nohit: like this channel
06:55 AM Smidge204: :p
06:55 AM Smidge204: Is this where you tell us to visit [redacted] instead?
06:58 AM nohit: nah, that'll get you banned these days
06:59 AM Smidge204: heh
07:17 AM matrixstorm: it might be dead for industy - but on the pther hand, industry is a whore doing that what brings the $$
07:17 AM matrixstorm: ;-)
07:17 AM matrixstorm: long lives the avr ;)
07:30 AM nohit: sure but some people want food on the table
07:36 AM nohit: and can you clarify a bit, "industry" should keep on using outdated, overpriced AVRs because...long live the avrs ?
07:54 AM matrixstorm: yes that was a bit simplified - i know (it was meant to be provokatove)
12:41 PM rue_mohr: have to say tho, its easy to just throw an avr at something with a few minutes coding on it and be done
12:41 PM rue_mohr: other things, say like stm32, might be more powerfull, but have a lot more overhead
12:41 PM rue_mohr: and the more gears there are in a system, the more there is to go wrong
12:45 PM matrixstorm: I like the structure/order in AVR peripherals
12:45 PM matrixstorm: in ARM everything has different IO
12:46 PM rue_mohr: and dear god, how many clock sources should need to have to be turned on to do ONE SIMPLE THING
12:46 PM matrixstorm: ;)
12:46 PM matrixstorm: what may be interesting as alternative
12:46 PM matrixstorm: do you know about the padauk micros ?
12:47 PM rue_mohr: eprom state machines!
12:47 PM matrixstorm: ;)
12:47 PM matrixstorm: hard wire 74HCT logic ;)
12:48 PM matrixstorm: padauk micros are very cheap
12:48 PM matrixstorm: but also small
12:48 PM rue_mohr: damn, I need a LARGE rom on my resistor substitution box
12:48 PM matrixstorm: maybe some level below AVR above mealy machines ?
12:48 PM rue_mohr: like 4M of tables alone...
12:49 PM Smidge204: rue_bed: How man bits is it, again?
12:49 PM matrixstorm: AVR?
12:49 PM rue_mohr: it looks like, if I get fancy, I'm into it for 3.8Mbits or so
12:49 PM Smidge204: The resistor box he's making
12:50 PM matrixstorm: ah, kk
12:50 PM rue_mohr: I got waaay too complicated :/
12:50 PM matrixstorm: put a SPI Flash to it storing the table stuff
12:50 PM matrixstorm: you can get them up to 16MiB for little $
12:50 PM rue_mohr: but then I sold myself on the benefits of the complexities
12:50 PM rue_mohr: I have some up to 32M
12:51 PM matrixstorm: yes, but they are not so cheap anymore
12:51 PM matrixstorm: ;)
12:51 PM rue_mohr: over the last few days I been pulling 8M chips off hard drive controllers
12:51 PM matrixstorm: ;)
12:51 PM rue_mohr: I think the controllers are FPGA, but cant find a lookup
12:51 PM matrixstorm: on my disks these only have 4M = 512KiB
12:51 PM matrixstorm: no
12:52 PM matrixstorm: Marvell?
12:52 PM rue_mohr: there are some with fpga yea
12:52 PM matrixstorm: most of them are ARM, too
12:52 PM matrixstorm: Thumb
12:52 PM matrixstorm: bust SPIs also sores S.M.A.R.T
12:52 PM rue_mohr: hmm, dont have one around, fiber channel drives
12:52 PM rue_mohr: and some SAS
12:52 PM matrixstorm: EVA storage
12:52 PM matrixstorm: ?
12:53 PM matrixstorm: HSV6000 ?
12:53 PM rue_mohr: just have a pile of drives
12:53 PM rue_mohr: seagate...
12:53 PM matrixstorm: yes
12:53 PM matrixstorm: seagate is HPE primary vendor
12:54 PM rue_mohr: ah
12:54 PM matrixstorm: yeah these ones use marvell controller
12:54 PM rue_mohr: ?
12:54 PM matrixstorm: and they can swap the bpards for different Interfaces
12:54 PM rue_mohr: is it an fpga?
12:54 PM matrixstorm: no
12:54 PM rue_mohr: yea, I noticed that the sas and fc controllers are swappable
12:54 PM matrixstorm: should be ARM
12:54 PM rue_mohr: huh
12:54 PM matrixstorm: is it an Marvell ?
12:54 PM matrixstorm: they should be lasered
12:55 PM matrixstorm: there should be some mosfets, motor driver, 32BIT DDR-RAM and the marvell controller
12:56 PM rue_mohr: one is LSI
12:56 PM rue_mohr: the other is an M logo :)
12:56 PM matrixstorm: :)
12:57 PM rue_mohr: 1 sec...
12:57 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1160249.jpg
12:57 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1160250.jpg
12:58 PM rue_mohr: those are the two controllers
12:58 PM rue_mohr: arm?....
12:58 PM matrixstorm: yes, the second one is marvell
12:58 PM matrixstorm: it seems to be the smaller 300GB FastClass HDDs
12:59 PM matrixstorm: have those thing at thouthands at work
12:59 PM matrixstorm: ;)
12:59 PM rue_mohr: 146G and up to 600G
12:59 PM matrixstorm: yes, fast class
12:59 PM matrixstorm: they are also available in slower 1TB++ near line
12:59 PM matrixstorm: but ocb look different then
12:59 PM matrixstorm: pcb
12:59 PM rue_mohr: so far, the flash is the only thing I can make use of
12:59 PM matrixstorm: the crystal ist 105MHz rectangle
12:59 PM rue_mohr: It would be neat to be able to re-use the controller
01:00 PM rue_mohr: have to clock in/out all the i tho ;.
01:00 PM rue_mohr: :/
01:00 PM matrixstorm: no way, marvell does not provide any stuff
01:00 PM matrixstorm: like the memory map
01:00 PM matrixstorm: or the boot sequence
01:00 PM rue_mohr: yea
01:00 PM rue_mohr: 62.58Mhz on the marvel
01:01 PM rue_mohr: 106.25 on the LSI
01:01 PM matrixstorm: kk
01:01 PM matrixstorm: would you like a foto of the near line PCB
01:01 PM rue_mohr: oh, marvel has chips on both these
01:01 PM matrixstorm: I have some laying around here
01:01 PM matrixstorm: ;)
01:01 PM rue_mohr: hah
01:02 PM rue_mohr: I have 54 of these boards in a bucket waiting for me to strip off the eeproms
01:02 PM matrixstorm: I checked these boards
01:02 PM matrixstorm: for me the most worthy was the SPI flash
01:02 PM matrixstorm: followed by the coils
01:02 PM rue_mohr: yup, I grabbed a few inductors and diodes too
01:02 PM rue_mohr: I'm not sure on the smps controller P/Ns
01:02 PM rue_mohr: hard to use them tho
01:03 PM matrixstorm: no
01:03 PM matrixstorm: these are mosfets
01:03 PM rue_mohr: so, they have two ceramic vibration? sensors?
01:03 PM matrixstorm: for these I think I have a datasheet
01:03 PM matrixstorm: the PWM is done by the controller
01:03 PM rue_mohr: it looks like a few of the supplies have their own controller
01:04 PM matrixstorm: maybe you can reuse the RAM
01:04 PM matrixstorm: but the controller is useless
01:04 PM rue_mohr: I was really surprised to see the (are they vibration sensors?) hooked up to LM358 amplifiers
01:04 PM matrixstorm: I also don't know what the ceramics are
01:05 PM rue_mohr: I grabbed a few of the ram chips, but I dont have BGA abilities
01:05 PM matrixstorm: maybe G-sensors
01:05 PM matrixstorm: but could be temp too
01:05 PM matrixstorm: no idea
01:05 PM matrixstorm: I reballed them
01:05 PM rue_mohr: k
01:05 PM matrixstorm: but they are IMHO useless, since they are small 128Mib?
01:05 PM matrixstorm: and they are 32bit
01:06 PM rue_mohr: yep
01:06 PM rue_mohr: I lothe dram
01:06 PM matrixstorm: ?
01:06 PM rue_mohr: too much work
01:06 PM rue_mohr: too many issues
01:07 PM rue_mohr: I wonder if there are datasheets on the smooth controller
01:07 PM matrixstorm: what has too many issues?
01:07 PM matrixstorm: reballing RAM
01:07 PM rue_mohr: 3 phase motor driver and coil driver
01:07 PM matrixstorm: well thats easy, its just 0.45mm balls
01:07 PM rue_mohr: DRAM
01:07 PM matrixstorm: I had to reball CPU with 347 0,1mm balls
01:07 PM matrixstorm: ;-)
01:07 PM rue_mohr: ouch
01:10 PM matrixstorm: https://imgur.com/a/NHaq92Z
01:10 PM matrixstorm: https://imgur.com/a/g7oZURO
01:11 PM matrixstorm: its 14mmx14mm
01:11 PM matrixstorm: Allwinner H5
01:11 PM matrixstorm: so you see, DRAM is not a problem
01:11 PM matrixstorm: but most of the time it is not worth the efford
01:13 PM matrixstorm: https://imgur.com/a/O06HgOd
02:44 PM Smidge204: The fact that so many instructions don't work with R0 thourgh R15 is really frustrating
02:44 PM matrixstorm: yes, but it has sth. useful about it
02:44 PM matrixstorm: in most firmwares you can reserve r2 to r8
02:45 PM matrixstorm: an use it as fast accessable memory
02:50 PM rue_mohr: so many?
02:50 PM rue_mohr: long time since I did assembler, but I dont recall having any issues
02:50 PM rue_mohr: I used to name them
02:50 PM rue_mohr: A, B, C, D, E, H, L...
02:56 PM Smidge204: You can't access the lower registers for most things. Can't put values into them directly, or use any immediate values. Can't set or clear bits directly. Can't load from or store to data/program/registery space from them.
02:57 PM rue_mohr: you can AND with immediate...
02:57 PM rue_mohr: sure you can load values right into them
02:58 PM rue_mohr: are you talking about the new ones microchip screwed up or the real atmel ones?
02:59 PM Smidge204: Both. The instruction set hasn't changed
02:59 PM Smidge204: If you try to "LDI R0, 0x42" you'll get an invalid register.
03:01 PM Smidge204: Basically any instruction that uses immediate data as an operand can't be used with R0 thourgh R16
03:05 PM Smidge204: R15*
03:10 PM rue_mohr: duh
03:10 PM rue_mohr: you stay over.. what was it agian 5?
03:10 PM rue_mohr: you cant do it with 0
03:11 PM Smidge204: ?
03:11 PM rue_mohr: the bottom like 4? registers are special
03:12 PM Smidge204: No?
03:12 PM rue_mohr: ; R 0 ->15 are gen. purp.
03:12 PM rue_mohr: .def A = r16
03:12 PM rue_mohr: .def B = r17 ; and logic instructions
03:12 PM rue_mohr: .def C = r18
03:12 PM rue_mohr: .def D = r19
03:12 PM rue_mohr: stay over 15
03:12 PM Smidge204: That's what I've been saying
03:12 PM rue_mohr: thught you said all of them
03:13 PM Smidge204: "R0 thourgh R15"
03:13 PM rue_mohr: ah k
03:13 PM Smidge204: I'd like to use them but they're not as flexible and it's a PITA for most things
03:13 PM rue_mohr: been a while, I thought it was like 4
03:14 PM rue_mohr: just be glad you have more than 1 register!
03:14 PM Smidge204: If it was just one it wouldn't be as frustrating! This is like they're tempting you with something you can't have
03:16 PM rue_mohr: I wonder if its a design oops
03:17 PM Smidge204: I'm guessing that the interconnections just got too out of hand
03:18 PM rue_mohr: I have to make resistor modules