#garfield Logs
Aug 15 2021
#garfield Calendar
12:03 AM Tom_L: https://www.avnet.com/opasdata/d120001/medias/docus/7/AES-S6MB-LX9-G-User-Guide-Configuration-Guide-ISE-12_4-13_1.pdf
12:10 AM Tom_L: nothing specific to me yet
12:12 AM rue_mohr: I dont have a programer that impact can work with, so..
12:12 AM Tom_L: trying it
12:13 AM Tom_L: i somewhat recall doing some of these steps but not that often
12:13 AM Tom_L: and it's doing what it's supposed to now
12:14 AM Tom_L: you have to create a prom file and describe it to impact then it generates the mcs file
12:14 AM Tom_L: after that you select the specific device
12:16 AM Tom_L: what is BPI?
12:18 AM rue_mohr: I dont know....
12:19 AM Tom_L: A byte peripheral interface (BPI) flash is used to store the FPGA bitstream that will be loaded automatically at power-up.
12:19 AM Tom_L: can't find one that supports the 4x spi stream
12:19 AM rue_mohr: you dont have a 4x
12:19 AM rue_mohr: you have a 1x
12:19 AM Tom_L: i do
12:19 AM Tom_L: have a 4x
12:19 AM rue_mohr: no, there is only 1 serial data line
12:20 AM Tom_L: ok
12:20 AM Tom_L: then i don't :)
12:20 AM Tom_L: wanna follow along?
12:20 AM Tom_L: start xilinx impact
12:27 AM Tom_L: with a project open
12:27 AM rue_mohr: that chip only does 1x
12:27 AM Tom_L: i believe you
12:27 AM rue_mohr: ok
12:27 AM rue_mohr: with a project open...
12:27 AM Tom_L: generate target PROM/ACE File
12:27 AM Tom_L: left menu
12:27 AM rue_mohr: this is in impact?
12:27 AM Tom_L: no
12:27 AM rue_mohr: what?
12:27 AM rue_mohr: ok
12:27 AM Tom_L: but it will take you there
12:27 AM Tom_L: in impact
12:27 AM Tom_L: create PROM File
12:27 AM Tom_L: step1
12:27 AM Tom_L: SPI Flash -> configure single FPGA
12:27 AM Tom_L: click on the green arrow to the right of it
12:27 AM Tom_L: next window select 32M
12:27 AM rue_mohr: but you have 64
12:27 AM Tom_L: then Add storage device
12:29 AM Tom_L: then generate file
12:29 AM Tom_L: after that i had to open project file
12:29 AM rue_mohr: k, so
12:30 AM rue_mohr: hows it written
12:30 AM Tom_L: it writes it thru the fpga i think
12:30 AM rue_mohr: will that process write it too?
12:30 AM Tom_L: that's what the little diagram shows anyway
12:30 AM Tom_L: yes
12:30 AM rue_mohr: huh
12:30 AM Tom_L: once you verify it's your device
12:31 AM Tom_L: click on it and it goes green
12:31 AM Tom_L: then 'program' from the right menu
12:31 AM Tom_L: i remember this now
12:31 AM Tom_L: but you only have to do it once per project
12:31 AM Tom_L: that's why i forgot
12:31 AM rue_mohr: :)
12:32 AM Tom_L: i had been selecting an exhisting mcs file so it was looking at a different project
12:32 AM Tom_L: but once you define the flash chip to it, you can either program that or the fpga
12:32 AM rue_mohr: ah
12:32 AM Tom_L: whichever one you turn 'green' in the diagram
12:34 AM Tom_L: been too long...
12:34 AM rue_mohr: :)
12:35 AM rue_mohr: so
12:35 AM rue_mohr: the idea that occured to me latley was regarding an H bridge
12:35 AM Tom_L: so it does change the .bit format to .mcs i think
12:35 AM rue_mohr: to ahve the fpga generate all 4 control lines
12:36 AM Tom_L: programmable Hbridge
12:36 AM rue_mohr: well
12:36 AM Tom_L: i may have to stop
12:43 AM Tom_L: can barely see the screen now
12:43 AM rue_mohr: if you have 4 control lines you can avoid that sillyness with coast/brake etc
12:43 AM rue_mohr: oh its late for ya isn't it?
12:43 AM Tom_L: 12:30
12:43 AM rue_mohr: midnight or so?
12:43 AM rue_mohr: its only :23, your clocks a bit off :)
12:43 AM Tom_L: it still loaded the wrong file
12:43 AM rue_mohr: !!!
12:43 AM rue_mohr: maybe give it a go in the morning?
12:45 AM Tom_L: i did it once with a different project
12:51 AM Tom_L: https://www.avnet.com/opasdata/d120001/medias/docus/7/AES-S6MB-LX9-G-User-Guide-Configuration-Guide-ISE-12_4-13_1.pdf
12:52 AM Tom_L: that appears to be the process
12:52 AM Tom_L: older ise webpack but it looks similar
01:04 AM Tom_L: ok. i give up
01:04 AM Tom_L: gnite
01:07 AM Tom_L: wireless kbd certainly doesn't agreee with it
01:16 AM rue_mohr: the 50Mhz maybe?
04:02 AM polprog: back home
04:02 AM polprog: it didnt work with that bitfile for me though :/
04:56 AM rue_shop3: hu?
04:56 AM rue_shop3: we need to do you too
04:56 AM rue_shop3: polprog, you do that in 2 stages
04:56 AM rue_shop3: you awake now or what?
05:01 AM rue_shop3: second one machined
05:21 AM polprog: yeah, its 12:00 here noon
05:45 AM rue_shop3: 3:30am
05:46 AM rue_shop3: I'm machining up the bearing blocks
05:46 AM rue_shop3: looks like 1 thou is a not press fit
05:49 AM rue_shop3: polprog, the bit file
05:49 AM rue_shop3: do you have time for me to tell you how to use it
05:50 AM rue_shop3: ok , or not
06:05 AM polprog: im soldering a board
06:05 AM polprog: so not now sadly
06:14 AM polprog: thts some screwed up pcb
06:14 AM polprog: https://i.ibb.co/hKxTskq/IMG-20210815-125318.jpg
06:14 AM polprog: :( the right one
06:14 AM polprog: solder doesnt want to stick at all no matter how much flux i put
06:15 AM polprog: i barely soldered that stm32 but i think ill give up and order the pcbs somewhere else, or email the fab house with a complaint
09:34 AM spline: I've had that, too, with boards
09:34 AM spline: and trying to put wire on (wire too) - I was wondering the same
09:35 AM spline: my experience comes more from the welding end where you can use different alloys/fillers you know right away when something isn't wetting out
09:35 AM spline: I normally use a 60/40, I think? I like leaded solder for personal projects. I can keep my iron (old 80s one a friend gave me but uses weller-style tips). I normally don't use any flux
09:36 AM spline: I had some junk chinese wires off some cheapo board once and like the shit chinese solder wasn't sticking for shit. I cleaned up a little, flux'd and use my leaded solder vs the normal 37 Tin/Antimony? and it wetted out so much better. Same diam
09:36 AM spline: always wondering how to debug that stuff
10:27 AM Tom_L: ok i generated a config that will program the flash. i hope i can remember the steps
10:40 AM Tom_L: 2nd time success
10:40 AM Tom_L: it's a 1 step fwd 2 step back process
10:41 AM Tom_L: configure target device and add the xilinx chip
10:41 AM Tom_L: generate target prom which is where you select the flash size
10:41 AM Tom_L: back to configure target device which will have the xilinx chip in it now
10:42 AM Tom_L: and right click on the chip and add a spi flash
10:42 AM Tom_L: there is where you select the part number once you pre'defined the size above
10:42 AM Tom_L: you must create and save a couple config files along the way
10:43 AM Tom_L: so the previous process can access them
10:45 AM Tom_L: after you're all done from then on you must select the project in impact separate from the main project
10:45 AM Tom_L: it's an impact project file .ipf
10:55 AM Tom_L: meh, it's still not doing what it's supposed to
11:03 AM Tom_L: ok i repeated it successfully but now i need to remember the steps again
11:05 AM Tom_L: impact has it's own separate project file
11:34 AM Tom_L: after you save it all and reopen is when it gets lost
11:51 AM rue_mohr: yes, I recall impact uses a different project file
11:52 AM rue_mohr: I thought impact was only a pain for what I was using it for (making the svf file)
11:52 AM Tom_L: getting the right order of events is a trick
11:53 AM Tom_L: and repeating it is yet another one
11:53 AM Tom_L: it asks if i want to reload the .bit file but it doesn't program that one
11:54 AM Tom_L: it's as if you're not supposed to update your code
11:57 AM rue_mohr: hmm
12:00 PM Tom_L: it's one step i can't figure out
12:00 PM Tom_L: how to let impact know the bit file is different
12:01 PM rue_mohr: for what I do, it watches it and knows
12:01 PM Tom_L: well it knows it's different but there is one step getting left out to regenerate the flash file
12:04 PM Tom_L: i can't figure it out
12:08 PM Tom_L: i guess don't use the flash until you're done testing, just write to the fpga...
12:09 PM rue_mohr: hmm
12:11 PM Tom_L: impact crashes alot
12:13 PM Tom_L: if you delete the project and start from scratch it works
12:14 PM Tom_L: if you update the program and regenerate the .bit file it doesn't load it to flash
12:22 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/xilinx/New_Board/spi_programming.pdf
12:23 PM Tom_L: talks about the way you do it i think
12:25 PM rue_mohr: I didn't have it crash
12:25 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/xilinx/New_Board/ise_tutorial_ug695.pdf
12:25 PM Tom_L: covers impact
12:25 PM rue_mohr: would you like to see the process on linux?
12:26 PM Tom_L: no
12:26 PM Tom_L: i don't use linux much and it would be more of a pain at this point
12:26 PM rue_mohr: ... are you sure....
12:26 PM Tom_L: there _is_ a solution
12:26 PM Tom_L: i just have to find it
12:26 PM rue_mohr: if the programmer is supported, its just two commands...
12:27 PM Tom_L: find a windows solution like that and i'm all in
12:41 PM rue_shop2: hmm, maybe an SPI pwm motor driver
12:41 PM rue_shop2: 8 bits with a direction/speed setup
12:43 PM Tom_L: settings let you fix the impact config file where it looked for the previous project
12:46 PM Tom_L: worked that time. let's try to repeat it
12:54 PM Tom_L: nope
12:57 PM Tom_L: meh, i probably won't mess with this that much more anyway
01:14 PM rue_mohr: yea, the command line thing I do on linux works really well...
01:14 PM rue_mohr: making that file using impact to program the older cpld, was a complete pain
01:14 PM rue_mohr: speaking of which, I need to adjust the on time down to 1 min
01:18 PM Tom_L: there are just an awful lot of options
01:19 PM Tom_L: it does support quite a few chips and configurations
01:24 PM rue_mohr: but none of the cables I have tho :/
01:26 PM Tom_L: parport one is dead simple to make
01:29 PM rue_mohr: I'v never found an fpga/cpld howto that says how to get into it all for less than $20
01:29 PM rue_mohr: everything says "buy this $87 starter board!"
01:29 PM rue_mohr: and like, wtf
01:36 PM Tom_L: this nice board i got was really cheap
01:36 PM Tom_L: i can't find it anymore
01:38 PM rue_shop3: under $50?
01:38 PM Tom_itx: https://www.ebay.com/itm/114807858644?hash=item1abb14e5d4:g:WU4AAOSwl8NVXJK9
01:38 PM Tom_L: i gave less than that for it
01:38 PM Tom_L: but i really like it
01:38 PM rue_shop3: hmm
01:39 PM rue_shop3: thats got the programmer cable built in?
01:39 PM Tom_L: no
01:39 PM Tom_L: jtag plug
01:39 PM rue_shop3: ah
01:39 PM Tom_L: my other one does
01:40 PM rue_shop3: so, over $50 with a programmer if you dont know what your up to
01:40 PM Tom_L: what was yours?
01:42 PM rue_shop2: too much!
01:42 PM rue_shop2: I think it was like $43
01:43 PM Tom_L: and you got nothing but the chip
01:43 PM rue_shop2: but I didn't know I didn't need it
01:43 PM rue_shop2: I was buying into it
01:43 PM Tom_L: i knew i didn't have much to connect to mine so i looked until i found that
01:44 PM rue_shop2: there doesn't seem to be a low budget into anywhere
01:44 PM rue_shop2: for those willing to solder :)
01:44 PM Tom_L: it was $20 something when i got it
01:44 PM rue_shop2: how much was the jtag programmer?
01:45 PM Tom_L: about the same
01:45 PM rue_shop2: oh
01:45 PM Tom_L: low 20
01:45 PM rue_shop2: I need to post you a pic of
01:45 PM Tom_L: i got a schematic for the parport jtag
01:45 PM Tom_L: webpack honors it
01:45 PM rue_mohr: I wonder if its the same as the one Imade
01:46 PM Tom_itx: http://tom-itx.no-ip.biz:81/~webpage/xilinx/digilent/dio1_sch.pdf
01:46 PM Tom_L: you just gotta pick thru it to find it
01:47 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1200632.jpg
01:47 PM Tom_L: it's just a buffer chip
01:47 PM rue_mohr: that was 3:30am
01:47 PM Tom_L: so i heard
01:48 PM Tom_L: .001 is an interferrence fit
01:48 PM Tom_L: something more than that is a press fit
01:48 PM rue_mohr: I got one to 1 thou and I can just about push it in
01:48 PM Tom_itx: http://tom-itx.no-ip.biz:81/~webpage/xilinx/digilent/dxc95_sch.pdf
01:48 PM Tom_itx: that must be it
01:49 PM rue_mohr: I forgot, on the last two, to confirm the depth before I pulled them out
01:50 PM Tom_itx: 74AC244
01:50 PM rue_mohr: so two of the bearing pockets are a bit shallow
01:50 PM Tom_itx: http://tom-itx.no-ip.biz:81/~webpage/temp/xilinx/xilinx1.jpg
01:50 PM rue_mohr: mmm cmos
01:50 PM Tom_itx: well one of those 2
01:50 PM Tom_itx: one is internal one is external
01:50 PM Tom_itx: with a switch
01:50 PM rue_mohr: I think you were working with that one not too long ago
01:51 PM Tom_itx: it's been quite a while
01:51 PM rue_mohr: working on tapping the jtag for programming another thing
01:51 PM Tom_itx: cause i don't remember where the parport cable is
01:52 PM Tom_itx: i doubt you have any 244s though :)
02:01 PM Tom_L: are those bearing holders gonna work ok?
02:02 PM Tom_L: you could realign it and go a bit deeper
02:02 PM rue_mohr: I hope so, there a lot of work
02:02 PM rue_mohr: not reasonable amount of effort to re-align them
02:02 PM rue_mohr: I dont worry about it
02:02 PM rue_mohr: I wont
02:02 PM rue_mohr: I'm thinking of using sphereical adapters to the pipe
02:03 PM rue_mohr: so the alignment isn't critical
02:11 PM Tom_L: i still don't get how they're being used
02:14 PM rue_mohr: the X is being driven on chain
02:14 PM rue_mohr: two chains, one on each side of X
02:14 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1200633.jpg
02:15 PM Tom_L: ok
02:15 PM rue_mohr: there are 1.5" pipes I'm using as drive shafts for the sprockets
02:15 PM rue_mohr: these are the bearing locks for those shafts
02:18 PM rue_mohr: blocks
02:19 PM rue_mohr: the ID of the bearings is *just a bunch bigger* than the pipe, so I have to make collars
02:21 PM Tom_L: so if i use this board, just send things to the fpga
02:21 PM Tom_L: and if i like it, then write to the flash
02:21 PM Tom_L: i can't figure out how to fix it
02:23 PM Tom_L: there are too many files in this environment and i'm missing out on one
02:38 PM rue_mohr: hmm
02:39 PM rue_mohr: hey
02:39 PM rue_mohr: I wonder if you can use openocd
02:43 PM Tom_L: i'm not that worried about it
02:43 PM Tom_L: i shouldn't be writing to the flash that much anyway
02:43 PM rue_mohr: I really was not impressed with impact
02:43 PM Tom_L: it's just that one last step i'm missing
02:44 PM Tom_L: i'm sure it's there, i just haven't found it yet
02:44 PM Tom_L: and i stopped looking
02:45 PM rue_mohr: what do you think of my idea of a 4 output H bridge controller?
02:45 PM rue_mohr: the delays between switching can be put right in
02:45 PM Tom_L: wouldn't take much of a cpld
02:46 PM Tom_L: you'd want to use a cpld not a fpga for that
02:46 PM rue_mohr: well, you also want safe defaults on the fets
02:46 PM Tom_L: or that, then you might get away with a fpga
02:46 PM Tom_L: google stmbl
02:47 PM rue_mohr: brushless?
02:47 PM rue_mohr: yes, but
02:47 PM Tom_L: it can do several different things
02:47 PM rue_mohr: the problem I constantly come up against is the tristate thing with driving dc motors
02:47 PM rue_mohr: you end up needing fet drivers that have delays and on and on
02:48 PM Tom_L: i dunno how they did it but it's fairly popular with lcnc guys
02:48 PM Tom_L: i don't have an interest in it
02:48 PM rue_mohr: having 4 control lines to programmable logic seems great
02:48 PM rue_mohr: every project like that seems to need lots of money
02:49 PM rue_mohr: my project budgets are $20
02:49 PM Tom_L: https://www.pcbway.com/project/shareproject/stmbl_servo_motor_drive.html
02:49 PM Tom_L: list of support targets
02:49 PM Tom_L: i'm aware
02:50 PM rue_mohr: I know I can fit a pwm gen in the small cpld too
02:52 PM Tom_L: mesa has one with vhd source
02:52 PM Tom_L: also a stepgen
02:55 PM Tom_L: that jtag seems to like to interfere with my keyboard
02:55 PM Tom_L: i moved the kbd dongle to another port
03:07 PM Tom_L: https://wiki.trenz-electronic.de/pages/viewpage.action?pageId=10617700
03:07 PM Tom_L: for me
03:07 PM Tom_L: you can peek if you can't resist
03:09 PM Tom_L: maybe too old a version
03:51 PM rue_mohr: ma came by, havn't seen her in a while now
03:52 PM Tom_L: she eat all your hidden food storage?
04:09 PM Tom_L: i recall a complaint previously..
04:45 PM rue_mohr: heh
04:46 PM Tom_L: i can't find much current on this topic
04:47 PM Tom_L: so as a temporary solution, i will leave the flash undefined and just use the fpga until i find something i want to keep
07:53 PM rue_shop3: it bugs me that I can do it so easy now
07:56 PM Tom_L: do what?
07:57 PM rue_shop3: I can program that flash super easy on my software
07:57 PM rue_shop3: 2 command lines
07:57 PM Tom_L: oh
07:57 PM rue_shop3: I know cause its hooked up like mine
07:57 PM Tom_L: it's really not a big deal as while you're debugging you would generally write to the fpga anyway
07:58 PM Tom_L: so much faster
07:58 PM Tom_L: there's likely a windoz solution out there
08:00 PM rue_shop3: openocd maybe
08:01 PM Tom_L: https://gnutoolchains.com/arm-eabi/openocd/
08:02 PM Tom_L: bak in a few
08:20 PM rue_shop3: just... openocd is a beast to figure out how to use
08:21 PM Tom_L: what did you say you used?
08:45 PM rue_shop3: I use xc3sprog
08:48 PM Tom_L: just for spartin6?
09:12 PM rue_shop3: well, I think it does other things too, but yea
09:12 PM rue_shop3: its an old abandoned project, but it works
09:12 PM rue_shop3: openocd is current, but honestly, nobody knows how to use it
09:13 PM rue_shop3: its like a media encoder, nobody knows how the command line options work and they keep changing every version, so anything you find online is completely useless
09:14 PM Tom_L: open source kinda sucks that way
09:15 PM Tom_L: this works, for now i just have to recreate the project which only takes a min or so
09:16 PM Tom_L: probably less painful than trying to find another windows solution
09:27 PM Tom_L: that was odd
09:56 PM rue_shop3: I got the 4th pillow block borred
09:57 PM rue_shop3: I have to put the mounting holes in next
09:58 PM rue_shop3: I can 2d print a template that can help me align the holes to the same places on them
10:01 PM Tom_L: nice
10:01 PM Tom_L: simple block of wood etc would do for that
10:04 PM Tom_L: what size chain are you gonna use?
10:04 PM Tom_L: and sprockets.. where you gettin them?
10:44 PM rue_mohr: bicycle
10:44 PM rue_mohr: but its hard to get here, I might go 50/50 chain/cable
10:44 PM Tom_L: you find 4 the same size?
10:45 PM rue_mohr: yes, but I only need 2 sets of 2
10:46 PM rue_mohr: this afternoon was quite the bunch of visitors.
10:46 PM Tom_L: you seem to have quite a few come by
10:47 PM rue_mohr: before covid it would get insane
10:47 PM Tom_L: looking for what?
10:47 PM rue_mohr: driveway packed and people parking across the street
10:47 PM rue_mohr: visiting and doing things
10:47 PM Tom_L: good you got plenty of friends
10:47 PM rue_mohr: hmm
10:48 PM Tom_L: i get calls when somebody needs something
10:48 PM rue_mohr: when counted it doensn't seem like lods
10:48 PM rue_mohr: heh
10:48 PM Tom_L: a couple or so would do quite a bit for me
10:48 PM Tom_L: the machine shop guy and another
10:51 PM Tom_L: lookin over some of the .v .vhd files i have
10:52 PM Tom_L: serial tx rx
10:52 PM Tom_L: don't think i've ever used that one
11:25 PM rue_mohr: ? programs?
11:40 PM aandrew: wow
11:40 PM aandrew: glad I searched for NiMH parallel charging
11:41 PM rue_mohr: ?
11:41 PM aandrew: was just going to hook a bunch of 6S packs in parallel and slow charge them
11:41 PM aandrew: apparently that's a bad idea
11:41 PM rue_mohr: they like fast charge dont they?
11:41 PM aandrew: https://endless-sphere.com/forums/viewtopic.php?f=14&t=3241&p=47328#p47328
11:42 PM rue_mohr: oh hah, yea dont mix ages
11:42 PM aandrew: they can fast charge, but with multiple cells in parallel the end of charge peak can be missed
11:45 PM rue_mohr: hmm