#garfield Logs

Jul 17 2021

12:02 AM polprog: [05:40] aandrew: the hamamatsu ones I used were 5V
12:09 AM polprog: well, these were hamamatsu :p
12:09 AM polprog: morning rue :)
12:09 AM polprog: its a TC103D cvd
12:09 AM polprog: ccd*
12:09 AM polprog: i have datasheet for TC102D
12:09 AM polprog: I just want to have it running first
12:09 AM polprog: then i was thinking about mounting a lens to it and have a small thing that just saves rows of pixels to an SD card
12:09 AM polprog: for strip photography
12:09 AM polprog: are you moving aandrew ?
12:17 AM Wings: Hamamatsu, nice. I just pulled a bunch of their end-view PMTs out of CO2 fluorescence analyzers.
12:47 AM rue_shop3: polprog, I have a few, so you have me interested
02:03 AM rue_shop2: JTAG loc.: 0 IDCODE: 0x59616093 Desc: XC95288XL Rev: E IR length: 8
02:03 AM rue_shop2: YES!
03:07 AM rue_bed: what I hear as I nod off isn't thoughts, its people talking
03:08 AM rue_bed: hmm
03:08 AM rue_bed: just a quick rest, I need to eat
03:51 AM polprog: Im positive that once i make the signals 12V it would work
03:52 AM polprog: ohh you got jtag runing yay
03:53 AM polprog: I guess you can program it with ISE?
03:55 AM rue_shop3: should be able to
03:55 AM rue_shop3: just hooking up an oscillator module and an lED
03:56 AM rue_shop3: hey wait its the forward votlage of an led...
03:56 AM rue_shop3: wait, why dont I use an SMT RGB
03:57 AM Tom_L: morning
03:57 AM rue_shop3: not yet
03:57 AM rue_shop3: oops, supper!
04:08 AM polprog: haha
04:08 AM polprog: tonight
04:08 AM polprog: last night i finished at 3 and ate the leftover pizza because I got hungry in the meantime
04:35 AM rue_shop2: :)
04:35 AM rue_shop2: ok, so, some quick flashing led code
05:01 AM polprog: nice
05:01 AM polprog: repurposing a cpu from a thing is something id like to do
05:32 AM rue_mohr: thats a CPLD...
05:33 AM rue_mohr: now that I have it flashing an led, I'm not sure what to do withit
05:33 AM rue_mohr: https://twitter.com/RueNahcMohr/status/1416341551286951939 <-- Tom_L !
06:18 AM rue_mohr: gngiht
07:07 AM Tom_L: :)
07:55 AM polprog: I found out how to prepare the flash payload for the spartan 6 boards we have rue_shop3
08:00 AM Tom_L: a bit file
08:08 AM polprog: no
08:10 AM polprog: it doesnt work if you flash the bit file, you have to use iMPACT to make an .mcs file
08:10 AM polprog: for me
08:16 AM polprog: oh crap, it doesnt appear as jtag now
08:18 AM polprog: ok it works now
08:26 AM polprog: oh there is a simpler way
08:26 AM polprog: I can just check "Create bit file" in Generate Programming File properties
08:26 AM polprog: and flash the .bin file
08:28 AM polprog: https://polprog.net/rozne1/ircjunk/fpga/prepare_flash_payload.txt
08:57 AM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/emc/xilinx/xilinx14_install_index.php
09:04 AM Tom_L: mine is more specific to mesa boards
09:37 AM polprog: the linear CCD seems to rect to light but not very much
09:45 AM Tom_L: maybe a certain wavelength of light?
09:46 AM polprog: no, when i shine a laser on it, the entire signal shifts, while I expect it to make a small dent in the line where the dot position is
09:46 AM polprog: The datasheet is not very helpful
09:49 AM Tom_L: hmm
09:51 AM polprog: Hmm, maybe the SP needs to be a separate signal
09:51 AM polprog: the sample and hold gate
09:55 AM polprog: hmm, RS pulses are nonexistent
09:55 AM polprog: they are too short to go through the uln2003
11:25 AM rue_mohr: polprog, I could have helped ya, there is a bit file to make the fpga into a flash programmer and then use it to write a bit file
11:29 AM polprog: I tried that but it reports unknown flash part for me
11:29 AM polprog: I found a bitstream that does it and oocd config that goes along with it
11:30 AM rue_mohr: want to try what I have?
11:31 AM rue_mohr: I just have to sort out 20+ notifications and have breakfast
11:31 AM rue_mohr: :)
11:33 AM rue_mohr: oof, looks like i have a lot of new followers, I didn't think anyone would grab on that one
11:45 AM rue_mohr: polprog, give me a shout if you want to try what I have
11:49 AM polprog: yeah i got 3 new followrs out of the blue
11:49 AM polprog: hmm. You can post the bitfile and openocd setup, i have to take a break now
11:49 AM polprog: ill get to it later today
11:50 AM rue_mohr: k, well I'll show you what I have after shower and breakfast then
11:50 AM rue_mohr: I need to work out what to do with the weekend
11:50 AM polprog: ok
11:50 AM polprog: This linear CCD is tricky
11:51 AM polprog: I dont have the datasheet for the exact part
11:51 AM polprog: I have to find a working level shifter ( Wings got me some ideas)
11:51 AM polprog: and i have no idea if it will even work
11:52 AM polprog: TCD103D
11:52 AM polprog: there is one on chipdocs.com, but they want $25
11:52 AM polprog: screw them
11:54 AM polprog: oh no
11:54 AM polprog: looks like 103 has a different pinout
11:54 AM Wings: What is the duration of the pulse you are having trouble shifting anyhow?
11:57 AM polprog: it says the frequency is 1mhz and the duty cycle is unspecified, it's about 10% in my case
11:57 AM polprog: so 10MHz
11:57 AM Wings: How are you arriving at 10MHz?
11:57 AM polprog: what do you mean
11:57 AM rue_mohr: I think you have to shift those fast, before the caps discharge
11:58 AM rue_mohr: I'm sure it shouldn't be 12V levels tho
11:58 AM Wings: You say the frequency is 1MHz, duty 10%.
11:58 AM rue_mohr: doesn't mean I'm right
11:58 AM Wings: There is no "therefore 10MHz"
11:59 AM polprog: The pulse width is like for a 10MHz square wave
11:59 AM Wings: But I assume you mean pulse length of 100ns
11:59 AM polprog: yes
11:59 AM polprog: thats what i meant
11:59 AM polprog: :)
11:59 AM Wings: okay
11:59 AM Wings: Just have to be clear.
12:00 PM Wings: So you need about 1V/ns slew
12:00 PM Wings: minimum
12:00 PM Wings: Do you know the impedance looking into that pin on the CCD chip?
12:00 PM Wings: (complex impedance, that is)
12:01 PM Wings: I expect it's almost entirely capacitance.
12:01 PM Wings: But, in a nutshell, this might simply require a symmetrical driver of some sort rather than a resistor pullup on an open collector.
12:01 PM Wings: Unless you want to dissipate some heat.
12:02 PM polprog: All i can say is from the 102D datasheet so Im not sure if the input capacitances there are correct
12:02 PM Wings: So what I would use is a CMOS IC.
12:02 PM polprog: But i would rather use a push pull driver too.
12:02 PM Wings: Forget TTL altogether.
12:02 PM polprog: I tried the CD4009, but it didnt work with the voltages
12:02 PM Wings: CD4009 is good up to 20V
12:02 PM Wings: But I guess your input logic wasn't enough
12:03 PM polprog: Yes
12:03 PM Wings: You will either have to find a completely different chip that does it all in one, or put a couple together to get there.
12:05 PM Wings: Hold on a moment though, can't you run Vdd on the 4009 at the lower voltage so your logic works on it?
12:05 PM Wings: Its whole point is level shifting.
12:05 PM Wings: Oh never mind, it only works as Vdd>Vi>Vcc
12:06 PM Wings: Not the other way around...
12:06 PM Wings: CD40109
12:06 PM Wings: That does the inverse.
12:07 PM Wings: And that version is immune to the order of application of power supply rails :)
12:10 PM polprog: I did that and it would not reach 0V on the output at all
12:10 PM Wings: How low was it going?
12:11 PM polprog: Barely, the closer i was to 12V on Vdd the less, i think it would do like 10V at most
12:11 PM Wings: Are you sure your CCD device is not bad?
12:11 PM polprog: CD40109 looks ok, I wonder if i have one
12:11 PM Wings: Wait, you said you already did that?
12:12 PM polprog: i tried 4009
12:12 PM Wings: ...
12:12 PM Wings: I said 40109 though, before that :P
12:12 PM Wings: And you said "I did that"
12:12 PM polprog: sorry, i got lagged :)
12:12 PM Wings: 4009 cannot do low-to-hi
12:12 PM Wings: It shouldn't be able to go to ground :)
12:14 PM Wings: Also, I was being a fool, I meant 7406, not 7405.
12:14 PM Wings: 7406 is the high voltage TTL inverter.
12:14 PM polprog: let me check if i have the CD40109
12:14 PM Wings: If not, see if you have a 7406 of some sort.
12:14 PM Wings: That will also do the voltages.
12:14 PM Wings: But you'll need low resistance to get the speed.
12:20 PM polprog: I found 40109
12:20 PM Wings: :)
12:20 PM Wings: That should do it
12:33 PM rue_mohr: aha
12:33 PM rue_mohr: polprog,
12:33 PM rue_mohr: https://github.com/ruenahcmohr/spartin6-fpga/blob/master/QMTECH.txt
12:33 PM rue_mohr: "flash write"
12:34 PM rue_mohr: forgot I posted all that
12:34 PM polprog: Thanks, ill test it
12:45 PM rue_shop3: I think I'm gonna try to put an i2c memory on that cpld
12:46 PM rue_shop3: I should be building a steam engine
12:57 PM Tom_L: thought you did that already
01:01 PM rue_shop3: a bigger one
01:01 PM rue_shop3: more from scratch
01:01 PM rue_shop3: Tom_L, did you see the dead bug?
01:01 PM Tom_L: yessir
01:02 PM Tom_L: make it a kit led string that changes colors as it chases
01:02 PM rue_shop3: the 1mm pitch isn't really that bad
01:04 PM Tom_L: i did one on the parport tester thingy way back
01:04 PM Tom_L: but it didn't change color
01:04 PM Tom_L: and wasn't on a cpld
01:17 PM Tom_L: your guy see those parts yet?
01:37 PM polprog: looks like 103d has a different pinout
01:37 PM polprog: or rather
01:37 PM polprog: more pins
01:42 PM rue_mohr: Tom_L, his days off are tues and wed
01:50 PM polprog: There is no datasheet on the net for that sensor
01:50 PM polprog: and the only schematic shows more connections than TCD102D
01:51 PM polprog: Ill put this off for when i get some replies
01:51 PM polprog: TCD103D
01:51 PM polprog: is what I have
01:51 PM polprog: But the level shifter works great
02:01 PM Wings: Which one?
02:01 PM Wings: The 40109?
02:03 PM polprog: yes
02:03 PM Wings: Eee! I helped for once.
02:03 PM polprog: The second one worked, the first one I picked was ded :^)
02:04 PM Wings: heh
02:04 PM Wings: It would be neat to also see how a 7406 does.
02:04 PM Wings: But as I said, you're going single-ended with resistive load, so sucky risetime.
02:04 PM Wings: Although you can do things to handle that, but that involves more cunning.
02:04 PM polprog: I dont have any 7406, maybe I should order some
02:05 PM polprog: Anyway, project stalled until I can get more info
02:05 PM Wings: They're handy for driving high voltage loads.
02:05 PM Wings: heh
02:05 PM Wings: I have a few CCD strips around, never tried to use one though.
02:05 PM Wings: Maybe I should one day.
02:05 PM Wings: Could be sueful.
02:05 PM Wings: useful*
02:05 PM polprog: There is a Linear CCD Databook by toshiba but its not on the net, I have to check out my uni library
02:05 PM Wings: Only sueful if I infringe on a patent.
02:05 PM polprog: But i doubt that they will have it
02:05 PM Wings: Do you know the title? Or, better yet, something like an ISBN for it?
02:07 PM polprog: Corporation, Toshiba. CCD linear image sensor: Data book. Tokyo: Toshiba Corporation, 1995.
02:07 PM polprog: one of the results in a citiation database. but not available as pdf
02:08 PM polprog: https://books.google.pl/books/about/CCD_linear_image_sensor_data_book.html?id=GKk5vwEACAAJ&redir_esc=y
02:09 PM polprog: not in the usual book places
02:09 PM Wings: Oh, I can't go buy it at my local Indigo and have a Starbucks coffee while I read it?
02:09 PM Wings: Damn.
02:20 PM Wings: I was able to find books that reference the Toshiba CCD book.
02:35 PM rue_mohr: oo polprog if you get that book, give me a chance to look for linear ccds I have
02:35 PM polprog: sure
03:20 PM Wings: Yes I would like info as well.
03:20 PM Wings: I looked around to find if I could source the book anywhere... no avail.
03:20 PM Wings: Might come down to getting on eevblog or the newsgroups and talking to engineers who've worked on those things.
03:20 PM Wings: To find the book you want.
04:12 PM Tom_L: https://books.google.com/books/about/CCD_linear_image_sensor_data_book.html?id=GKk5vwEACAAJ
04:14 PM Wings: polprog just shared that
04:15 PM Wings: Well, the Polish link to the same, hehe.
04:15 PM Tom_L: http://www.chipdocs.com/pndecoder/datasheets/TOS/TCD103D.html
04:15 PM Wings: mmmm, paywall
04:15 PM Tom_L: ooohhhhh you want it free!
04:16 PM Wings: polprog mentioned the chipdocs earlier, and mentioned they wanted $25 :P
04:17 PM Tom_L: https://www.alldatasheet.com/view.jsp?Searchword=TCD103D&mo
04:17 PM Tom_L: dunno if they're any good or not
04:17 PM Wings: Looked there already too. Different variants, but not the exact one.
04:18 PM * Tom_L stops looking and goes back to his weekend nap
04:18 PM Wings: hehe
04:18 PM Wings: I even dug through archive.com
04:18 PM Wings: And a bunch of other weird arcane places.
04:18 PM Wings: We are all trying very hard to help polprog :P
04:18 PM Tom_L: not me
04:18 PM Tom_L: i gave up
04:18 PM Wings: But you tried.
04:24 PM polprog: No hits for databooks in my uni library
04:26 PM rue_shop2: hmm, I wonder how I frame a data type for serial
04:26 PM rue_shop2: can I not do 1.5 start bits? or was it stop bits...
04:28 PM rue_shop2: oh, I'm the only person in the world that was doing that, ok
04:29 PM rue_shop2: so, {startbit = 0, 8_data_bits, stopbit = 1}
04:30 PM Wings: Stops are flexible, starts more rigid.
04:31 PM Wings: Due of course only to mechanical systems.... But we are living with their leftovers even in the electronic world, haha.
05:04 PM rue_shop3: yea there's a lot of them out there
05:05 PM rue_shop3: but back in my day, you used 1.5 ... stop? bits
05:05 PM rue_shop3: 19.15mm
05:05 PM rue_shop3: wait, I can make a pulley faster on my lathe than the 3d printer
05:06 PM Tom_L: yessir, pulleys _are_ generally round
05:13 PM rue_shop3: the 3d printers dont like the overhangs either
05:22 PM Tom_L: pulley for what?
05:31 PM rue_shop3: for the steam engine
05:32 PM rue_shop3: to drive the generator for running the cpld
05:32 PM rue_shop3: }:]
05:36 PM polprog: coal powered logic device
06:07 PM rue_mohr: coal?
06:07 PM rue_mohr: you cant get that anymore
06:07 PM rue_mohr: wood
06:33 PM Tom_L: ok hopefully all the chip guard on the mill is fixed now
06:33 PM Tom_L: some still have some stress cracks at the screw heads but i'm not putting pressure there now
06:33 PM Tom_L: had 2 more brackets to make and now that's done
06:34 PM rue_mohr: good show
06:34 PM Tom_L: hotter than hades out there
06:35 PM rue_mohr: I got silly
06:35 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Final/Vise_parts/Table_side_brackets3.jpg
06:35 PM rue_mohr: https://twitter.com/RueNahcMohr/status/1416538554319003649
06:35 PM Tom_L: made some shorter ones for each end
06:35 PM Tom_L: and near the vise
06:35 PM rue_mohr: nice
06:36 PM Tom_L: :)
06:36 PM Tom_L: kit's kinda sluggish today
06:36 PM rue_mohr: I should put the pull cord back on it, I'm gonna hurt myself
06:37 PM Tom_L: i can see that coming
06:37 PM Tom_L: so
06:37 PM Tom_L: exactly how efficient is that?
06:37 PM rue_mohr: I'm sure its over 5%
06:38 PM rue_mohr: :) I should have taken a photo of the cylinder, its AWEFULL
06:38 PM rue_mohr: dunno who did what to it, looks like it was run out of oil and fed sand
06:38 PM Tom_L: ran without a filter
06:39 PM rue_mohr: maybe, not sure
06:40 PM Tom_L: but can you program the cpld under that power?
06:52 PM rue_mohr: sure, its 15->24V stepped down to 5V then 3.3V
08:42 PM rue_shop2: http://www.stanleymotorcarriage.com/SteamEngine/SteamEngineGeneral.htm
08:42 PM rue_shop2: ok ok, steam is better, I hear you, ugh
08:45 PM Tom_L: you're really gettin into this steam engine thing :)
08:46 PM rue_shop2: see if I can convert that truck
08:46 PM rue_shop2: I just tripped across what I been wanting to read about
08:46 PM rue_shop2: the stanley steam car
08:46 PM Tom_L: if it wasn't stuffed up in the attic, i'd take a pic of the one my grandfather made for me
08:47 PM Tom_L: i dunno where the one i made is, it may be in the box with it for all i know
08:47 PM rue_shop2: nom 600psi
08:47 PM rue_shop2: :S
08:47 PM Tom_L: wonder how many boilers they had blow
08:47 PM rue_shop2: my goal is 200
08:48 PM rue_shop2: which I think is high
08:48 PM Tom_L: more than the average home compressor
08:50 PM rue_shop2: ... graphite packing!?
08:50 PM Tom_L: ever see any?
08:50 PM rue_shop2: no...
08:51 PM Tom_L: like a rope of graphite
08:51 PM rue_shop2: can still get it?
08:51 PM Tom_L: doubtful
08:51 PM rue_shop2: its the cylinder rod packing that I'm intereted in
08:51 PM Tom_L: https://www.amazon.com/graphite-packing/s?k=graphite+packing
08:52 PM rue_shop2: PTFE with graphite
08:52 PM rue_shop2: so its better now
08:52 PM Tom_L: undoubtedly
08:53 PM rue_shop2: oh, I think its the same stuff as head gasket
08:53 PM Tom_L: surprised you've never seen any around your plumber friends
08:54 PM rue_shop2: so, 4" dia pistons, 600psi
08:55 PM Tom_L: head gasket is harder
08:55 PM rue_shop2: 7.5 tonnes
08:55 PM rue_shop2: oof
08:56 PM Tom_L: careful, you may blow something across the back yard and it could be you
08:56 PM rue_shop2: (the stanley car)
08:56 PM rue_shop2: I'm going for 200psi on ~2.5" for 500lbs ea or so
09:01 PM rue_shop2: so I need to make some cylinders
09:01 PM rue_shop2: and the rod packing is the hardest part I think
09:06 PM Tom_L: what about rings?
09:15 PM rue_mohr: lots of solutions
09:17 PM rue_mohr: https://twitter.com/fhk2014_kk/status/1406854394650505218
09:17 PM rue_mohr: ^^ ooo!!
09:22 PM Tom_L: reminds me of the bar machines i used to run in a way
09:22 PM Tom_L: all cam driven
09:23 PM Tom_L: wonder what it's making
09:23 PM rue_mohr: yea
09:23 PM rue_mohr: looks like springs
09:24 PM Tom_L: i can't tell
09:25 PM rue_mohr: so
09:25 PM rue_mohr: I cant get a cylinder hone till later in the week
09:26 PM rue_mohr: the lathe is not equipped to face the pipe I want to use for a cylinder
09:26 PM Tom_L: why not?
09:26 PM rue_mohr: I need a steady for it, and mines too small
09:27 PM rue_mohr: I could cast a larger one
09:27 PM Tom_L: oh
10:42 PM rue_shop2: dia 24.4mm
10:43 PM rue_shop2: inner 11mm safe
10:44 PM Tom_L: little over 1/4" wall
10:44 PM Tom_L: yes
10:45 PM Tom_L: depending on the material that is
10:45 PM rue_mohr: I have to make a thing to hold a bearing...
10:46 PM Tom_L: i thought we did that already
10:51 PM rue_mohr: its for the plasma cutter
10:51 PM rue_mohr: https://twitter.com/RueNahcMohr/status/1416602199514238979
10:51 PM rue_mohr: in those holes, I need to mount a 608 to run on the rail
10:53 PM Tom_L: do they fit in the holes?
10:57 PM rue_shop2: holes are 24.4, bearings are 22
10:59 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Cube/Hole_test2.jpg
10:59 PM Tom_L: test cuts
11:00 PM rue_shop2: hah oooo
11:00 PM rue_shop2: 608s eh?
11:00 PM Tom_L: probably yes
11:00 PM Tom_L: skate bearings
11:01 PM rue_shop2: thats 608
11:03 PM rue_shop2: ok, cpld serial stream
11:04 PM Tom_L: you can't cut the bearing holders on the mill?
11:33 PM rue_shop3: I have to put them perpendicular thru the plate
11:34 PM rue_shop3: I think I achived serial
11:38 PM rue_shop3: wtf
11:38 PM rue_shop3: when I use hex view, it says 0x30
11:39 PM rue_shop3: but in text mode it just says unknown character
11:39 PM rue_shop3: 0x30 should be '0'
11:39 PM rue_shop3: ah
11:39 PM rue_shop3: sync gets screwed up and it otherwise thinks its 0x82
11:51 PM rue_shop2: ok! two stop bits might have fixed it
11:52 PM rue_shop2: no it needs more
11:53 PM rue_shop2: its not helping that I'm just sending the same character repeatedly