#robotics Logs

Jun 06 2017

12:01 AM rue_house: real neon pilot lamps
12:04 AM rue_house: ok that should work, if I load the servo beyond 25%, it'll turn off
12:04 AM rue_house: try it tommorow I think
12:05 AM Anniepoo_: ne-2 neon tubes
12:05 AM Anniepoo_: 8cD
12:06 AM rue_house: hmm
12:06 AM Anniepoo_: you can make a cosmic ray detector
12:06 AM rue_house: dosn't that need a photon multiplier?
12:06 AM Anniepoo_: nope
12:06 AM rue_house: oh
12:07 AM rue_house: hows that different from a geiger tube?
12:07 AM rue_house: or even a gieger tube
12:07 AM Anniepoo_: wire tubes in parallel, use them as the dump gate on a relaxation oscillator
12:08 AM rue_house: these are just pilots, I ahve raw neon bulbs comming
12:08 AM rue_house: ... might have been 1000, I dont recall
12:08 AM Anniepoo_: probably ne2 and resistor inside
12:09 AM Anniepoo_: i discovered this as a kid
12:09 AM Anniepoo_: big moment in my life
12:22 AM rue_house: heh
12:22 AM rue_house: its a neon bulb, no idea what isotope
12:37 AM Anniepoo_: dunno radiochem of neon, not important here
12:48 AM mrdata: they will all be stable isotopes
12:51 AM mrdata: so probably natural abundance
01:00 AM Anniepoo_: also, nobody going to change it
01:09 AM rue_house: your saying I'm not to turn the neon in the tube to nitrogen?
01:36 AM rue_house: I wonder if I can set up a small temporal fold so I can talk to myself for a while
03:15 AM Anniepoo_: no, probably cant transmutate neon
03:16 AM Anniepoo_: sorry rue, middle of night
03:20 AM Anniepoo_: you need to find some japanese roboticists who speak english, so you have people around middle of nite
03:33 AM Jak_o_Shadows: Yeah, rue, if you have a handy folder online, I would like to see tem
03:33 AM Jak_o_Shadows: I agree, the examples online are either too siimple, or too complex
05:33 AM juanonymous is now known as kristhian
07:00 AM kristhian is now known as juanonymous
07:05 AM juanonymous is now known as kristhian
12:50 PM dStruct: hey all, I'm wondering if anyone here happens to be familiar with the Intel 80c52 series of microcontrollers for motion control?
12:51 PM SpeedEvil: I am slightly familar with them - why that micro
12:53 PM dStruct: SpeedEvil: so I have a legacy system I'm trying to upgrade, and it runs an 80c52 controller which does first order equations, mixed with a 10v IO board to control 2 channels of hydraulic actuators. What I'm trying to find is a modern digital unit that I can run the assembly code on to upgrade it
12:53 PM SpeedEvil: There are lots of 8052 clones available
12:54 PM dStruct: SpeedEvil: the idea is I want to integrate the 10vdc digital and analog IO portions into a single microcontroller and get away from using an analog board for offset and gain stuff
12:54 PM SpeedEvil: seems reasonable.
12:54 PM dStruct: SpeedEvil: yeah but trying to find one that does 10vdc IO has been tricky, I've been searching for months now
12:54 PM SpeedEvil: There are ~0 micros that will do 10V out.
12:54 PM SpeedEvil: (some car focussed ones probably will)
12:54 PM SpeedEvil: Do you just need 0-10V output?
12:55 PM dStruct: SpeedEvil: yup, it's not differential so +0-10v
12:55 PM SpeedEvil: A simple opamp would do it, and a micro.
12:55 PM dStruct: SpeedEvil: standard for most hydraulic servos and load cells
12:55 PM SpeedEvil: Just 2* gain, PWM, RC filter.
12:55 PM SpeedEvil: PWM pin -> RCRC filter to kill HF, opamp.
12:56 PM SpeedEvil: Or you could fins a micro with a DAC, and use only the opamp
12:57 PM dStruct: SpeedEvil: I wish I had time to sit down and design something but unfortunately I do not. I was hoping to find something to drop straight in. The legacy system is a Technology 80, 80c52 board with a Robotrol SBX388 both long unsupported
12:59 PM dStruct: SpeedEvil: actually what I said was incorrect this particular system IS a differential -10 to +10vdc, the other system is single-ended 0-10vdc
01:01 PM SpeedEvil: Do you have the asm?
01:02 PM dStruct: SpeedEvil: I did find the Delta Motion RMC75 which looks like it should work, and the engineers there seem to think they can make it happen. But I was hoping for a better option of some kind
01:02 PM dStruct: SpeedEvil: yes
01:02 PM SpeedEvil: Do you actually need to run the code, or would simply ripping off the algorithms also be an option
01:02 PM dStruct: SpeedEvil: the RMC75 will not take my native assembly so I'll have to do some re-engineering of the code
01:03 PM dStruct: SpeedEvil: well the easiest option would be running the asm itself, but it's looking more like I'm going to have to rip off the equations and re-engineer something
01:03 PM SpeedEvil: For example, much modern stuff would be quite capable with a very dumb C implementation of vastly overperforming.
01:03 PM dStruct: SpeedEvil: yes, such as the RMC75, it has the capability but not the compatibility I'm looking for
01:05 PM SpeedEvil: If you can edit the output routines, and the input routines and the timing, you can use one of the many 8052 clones with more features so you can just ditch basically all of teh board except for the 8052 and a dual opamp, and a couple of passives.
01:05 PM SpeedEvil: For example, the 8052 had no IO, but modern ones have lots of PWM and fancy stuff which would work fine for hydraulics
01:06 PM SpeedEvil: Maybe also ask in ##embedded and ##electronics In principle, an arduino class processor (not perhaps with the arduino stack) would be fine for this too.
01:06 PM dStruct: SpeedEvil: exactly, I'm trying to locate something that has an 8052 MCU yet also can do +/- 10vdc analog IO and about 4 DIO
01:07 PM dStruct: SpeedEvil: I'll do that, thanks for the input :D
01:07 PM SpeedEvil: the +-10VDC is easy, and needs literally about four resistors per channel and an opamp
01:07 PM SpeedEvil: (to go from 0-5V)
01:08 PM dStruct: SpeedEvil: the main issue with the system is the gain and offset values are handled by a less then reliable analog board
01:08 PM dStruct: SpeedEvil: that and the fact that if the 8052 or any components go tits up parts are impossible to find
01:09 PM SpeedEvil: Making the parts reliably available is not that hard - DIP8 opamps rail-rail are going to be around a while yet. (though decreasing in number)
01:10 PM SpeedEvil: (for a simple redesign)
01:10 PM SpeedEvil: I'd probably go with an arduino board (not using the arduino stack) for that reason.
01:12 PM dStruct: SpeedEvil: yeah that would work and that was my first consideration, however if I'm going that route I may just end up with something like an RMC75 with a warranty and support
01:13 PM SpeedEvil: ah
01:14 PM SpeedEvil: I think pretty much any stock 8051 board is not going to have the stuff, and is going to need you to make a board of some form.
01:16 PM dStruct: SpeedEvil: yeah I don't expect the 8052 itself to handle the IO portion but I'm hoping I can find something that is 8052 based (or emulated) that has that capability
01:17 PM SpeedEvil: A sllly option would be something like a raspberry pi with a 8052 emulator
01:30 PM dStruct: SpeedEvil: would be nice, but I need hard real-time determinism, which the RPi is async
02:10 PM robotustra_: hi
07:24 PM rue_house: pwf, whats with 8052 now!?