Back
[00:50:05] <Triffid_Hunter> rue_shop: sounds like a wonderful voltage to fire converter
[01:13:31] <rue_house> well, its designed to be hooked up to a controller that dosn't let that come up
[01:13:39] <rue_house> I didn't have one connected yet
[01:14:07] <rue_house> the motor driver has pullups on the N channels fets that do the high and low side
[01:15:26] <rue_house> Triffid_Hunter, I'm making an 8 channel force feedback controller
[01:16:11] <Triffid_Hunter> rue_house: yikes, the gate driver I used in my RC car has built-in shoot through prevention and dead time
[01:16:28] <rue_house> yea, I had to make my own
[01:16:45] <Triffid_Hunter> rue_house: why? can't order hip4082 from your favourite supplier?
[01:16:46] <rue_house> there are no, nice, cheap, compact motor drivers that have less than .5V drop at 1A
[01:17:15] <rue_house> I wanted a dip8 that would do it all
[01:17:18] <rue_house> they dont exist
[01:17:28] <Triffid_Hunter> yeah that's because all the interesting chips are surface mount
[01:17:47] <rue_house> nobody seems to have made a low RDSon fet based, all in one driver
[01:17:59] <rue_house> I couldn't even find any in soic8
[01:18:12] <rue_house> not for less than .5V drop at 1A
[01:21:07] <Triffid_Hunter> rue_house: what about tb6612?
[01:21:30] <Triffid_Hunter> it's ssop24, needs a big package for heatsinking
[01:22:00] <Triffid_Hunter> rue_house: or VNH2SP30 and friends?
[01:22:39] <rue_house> reading...
[01:24:54] <rue_house> tb6612voutsat is 0.5V
[01:25:08] <rue_house> so that'll drop 1V in the whole bridge
[01:25:31] <rue_house> I think
[01:25:58] <Triffid_Hunter> rue_house: well the thing is, the silicon process for logic and for really good mosfets is quite different, can't put both on the same chip, so your choices are either high Rds(on) or a gate driver and external fets
[01:26:31] <rue_house> yup, not that we dont have the tech to put multiple dies in a carrier...
[01:27:03] <Triffid_Hunter> rue_house: we've the tech but that sounds expensive in process, esp after confirming yield
[01:27:30] <rue_house> 30 pINS!?
[01:27:45] <rue_house> (30A is also insane)
[01:28:10] <Triffid_Hunter> rue_house: just a couple of examples, hit element14 or digikey or whatever and go look in the motor driver section ;)
[01:28:23] <rue_house> no no no, I need cheaper than that
[01:28:28] <Triffid_Hunter> parametric search is your friend :P
[01:28:56] <rue_house> yes, but when you try to look for good chips on aliexpress they come with prices for good chips :)
[01:30:02] <rue_house> see, a , even 5A version of that last chip VNH2SP30 would be perfect
[01:30:11] <Triffid_Hunter> rue_house: oh I'm not suggesting you necessarily buy them from the site, just use their parametric search to find part numbers for ali
[01:30:25] <rue_house> I did, I tried everything
[01:30:52] <rue_house> the VNH2SP30 must be cheap, board+ from china is $3 ea
[01:31:15] <rue_house> ah $2 ea/chip...
[01:31:27] <rue_house> still way over budget for this project, but, nice chip..
[01:31:59] <rue_house> if they had current control, one could make an awesome stepper driver
[01:34:52] <rue_house> SI2318 is what I ended up using, x4/motor
[01:35:41] <rue_house> SOT23, 3A continious rated, 36mR
[01:36:18] <rue_house> so, if you short a supply against them, they fuse quickly.
[01:36:39] <rue_house> I have a boostrap on the top one
[01:37:10] <rue_house> a TLC4950 generates the low side signal which is also fed into a ULN2803 for the inverted high side
[01:37:56] <rue_house> beacuse the boot strapped high side comes up to 19V, which is above the rating of the tlc4950 outputs
[01:38:12] <rue_house> (12V gate drive with a rated 7V motor supply)
[01:39:15] <rue_house> the other thing that dosn't exist is a good 500mA rated, common .. cathode? SIP diode array
[01:39:28] <rue_house> like $1-$5ea
[01:39:53] <rue_house> given the price of 1N914....
[01:41:23] <rue_house> I have a half driver on a 6 pin SIP, thats 0.7" x 0.5" two transistors, flyback diodes, pullup resistors, and bootstrap cap & diode
[01:41:40] <rue_house> yay smt :)
[01:43:11] <Triffid_Hunter> that sounds like way more chips and things than using something like hip4082 and some external fets
[01:43:27] <rue_house> not by much
[01:43:40] <rue_house> I didn't want to, the price fit
[01:44:07] <Triffid_Hunter> rue_house:
http://imgur.com/a/IDQ5u <-- seen this? my R/C brushed ESC project
[01:44:31] <rue_house> but it occured to me yesterday that if I made it a dual rail supply and changed the driver circuit, I could ahve done it with half bridge drivers and saved 2x the transistors
[01:45:02] <rue_house> that explains why you know the good chips off the top of your head
[01:45:27] <Triffid_Hunter> heh :P
[01:45:34] <rue_house> irf1404 or soemthing with less gate capacitance?
[01:46:03] <Triffid_Hunter> rue_house: I used IRLB8743
[01:46:18] <rue_house> gate capacitance killing switching time has hounded a number of my projects
[01:46:41] <Triffid_Hunter> Qg typ 36 max 54 nC
[01:47:11] <Triffid_Hunter> rue_house: if you want low Qg in a power mosfet, check IRLL024.. can carry about 1A and the Qg is tiny
[01:47:31] <rue_house> logic level...?
[01:47:38] <Triffid_Hunter> rue_house: Qg of zxmn4A06 is pretty low, I use that up to 2-3A with 3.3v gate drive
[01:47:46] <Triffid_Hunter> rue_house: yeah I like my logic fets even when I don't need 'em
[01:48:01] <rue_house> whats your switching freq/time?
[01:48:19] <Triffid_Hunter> rue_house: on the ESC? umm 15khz from memory
[01:48:48] <Triffid_Hunter> rue_house: IRLL024 has Qg of 7-11nC, zxmn4a06 is ~18nC
[01:49:22] <Triffid_Hunter> IRLL024 is good with 5v gate drive, zx goes down to 3.3 (as does AOT240L which is a complete beast but Qg is of course higher)
[01:50:52] <rue_house> I have a 25Mhz scope, so about 2us makes me happy
[01:51:16] <Triffid_Hunter> rue_house: all these parts I find with parametric search fwiw.. I've got a project here with FDMC86102 and FDMC86139 which are N and P channel mosfets in a fairly attractive LFPAK33 which is like power DFN8 but I think a bit smaller
[01:51:33] <rue_house> your 14V or so design motor supply?
[01:52:02] <Triffid_Hunter> rue_house: ESC is rated 9-15v or so, upper limit is Vgs(max) and lower limit is the UVLO in the hip4082
[01:52:14] * rue_house nods
[01:52:16] <Triffid_Hunter> it'd be ok at 20 as long as you don't decelerate hard
[01:52:25] <rue_house> heh
[01:52:38] <rue_house> I'v blown up two industrial brushless that way
[01:52:54] <rue_house> not enough cap on the supply to eat the regen
[01:53:25] <Triffid_Hunter> rue_house: thing I like about the hip4082 is I can get all the motor drive modes with two GPIO and two PWM.. forward, reverse, brake (short), coast (open), drag (forward/brake), locked antiphase (forward/reverse)
[01:53:28] <rue_house> :( that was a nice motor (X2)
[01:53:34] <Triffid_Hunter> rue_house: and no crowbar? heh
[01:54:14] <rue_house> they assume they can dump more into the supply than what I hooked up
[01:54:43] <rue_house> the 4082 is nice
[01:55:02] <Triffid_Hunter> yeah, it's basically everything except the fets
[01:55:35] <Triffid_Hunter> and due to the internal logic with high/low drive and built-in deadtime generator you basically can't shoot-through if you tried
[01:55:39] <rue_house> I messed around with using LM386 and LM555 as fet drivers, they dont do that bad a job, ok for slow switching
[01:55:48] <rue_house> (level convert and drive)
[01:55:58] <Triffid_Hunter> I hear the TLC555 is quite lovely
[01:56:47] <rue_house> :( 4082 is still like $2 ea
[01:57:18] <Triffid_Hunter> but, there's proper gate drivers now with some fairly ridiculous specs these days, on smoothie we use TC4427, rated at 1.5A and I've seen some rated up to 12A
[01:57:40] <rue_house> I have some dip8 drivers I got as small load drivers
[01:58:13] <rue_house> IR4427 iirc
[01:58:28] <rue_house> AOP605
[02:00:25] <Triffid_Hunter> rue_house: curious device
[02:00:39] <Triffid_Hunter> be good for sync switchers if it came in a smaller package
[02:00:53] <rue_house> and they are cheap
[02:01:43] <rue_house> gngiht
[02:01:47] <rue_house> oh
[02:01:55] <rue_house> wait a sec
[02:02:36] <rue_house> http://ruemohr.org/~ircjunk/projects/reprap2/p1080521.jpg
[02:03:11] <rue_house> I upsized it a bit, instead of cutting the 3' rod funny lengths, I just cut it in half (it was close)
[02:03:32] <rue_house> redesigned the X carrige, geared down the XY
[02:03:57] <rue_house> :)
[02:03:59] <rue_house> gnight
[02:04:53] <rue_house> ... I2.5...
[10:30:22] <branjb> are entry level job postings looking for more and more unrealistic requirements in reaction to college grads being less competant than ever?
[10:33:30] <z64555> No, they're looking to hire the guys that are retiring
[10:35:18] <z64555> Apply for the ones that you think you can handle, regardless if you didn't meet the age or time in the workforce
[10:36:00] <z64555> I'd stay away from applications that list GPA unless you made a 5.0
[10:36:15] <branjb> http://careers.peopleclick.com/careerscp/client_arm/external/en-us/rssfeed.do?functionName=viewFromLink&jobPostId=28093&localeCode=en-us&rssFeedId=1504630 is an example of one I am looking at. I'm familiar with most of the things that it requires, but only at the very fundamental level. Maybe my undergrad program just sucks, but we only really touch on each topic listed there for no more
[10:36:15] <branjb> than a month in a given course
[10:36:17] <branjb> hahaha
[10:38:52] <branjb> "Excellent software engineering skills including understanding of data structures and algorithms in C/C++ and Java, and graphical user interfaces" I've been programming in various languages for ~10 years and I'd say my coding abilities have only recently gone above the "not embarrassed to have other people see it" level
[10:41:55] <branjb> I just wonder how much of some job postings they are actually serious about vs "it sure would be nice to hire a senior engineering at an entry level salary"
[10:45:08] <z64555> "Willing to take risks to create revolutionary solutions" is a red flag to me
[10:45:16] <branjb> hahaha
[10:45:40] <branjb> here's another one, another entry level position: "The candidate must be familiar with all facets of embedded firmware design and development; from bare metal board bring up to verification of various filters implement in assembly."
[10:46:12] <branjb> are there really undergrad EE / CpE programs that teach full microcontroller implementation from the PCB level to working design?
[10:46:24] <z64555> avoid that one, they're trying to hire a department on a one-man salary
[10:47:05] <z64555> undergrad programs? no, not unless you've scored an internship, or are returning to school
[10:47:21] <branjb> it seems like most of the indeed/linkdin postings are equally poor. The postings on individual companies websites seem much more realistic and less corporate/HR fluff in the requirements sections
[10:48:29] <z64555> Yeah, unless the HR firm they hired knows what the position actually is, the posting will likely be sketchy at best
[10:49:51] <branjb> "Job posted by xxxxx xxxxxx
[10:49:51] <branjb> Global Staffing Manager"
[11:05:01] <branjb> what do you think of companies that have "engineering mentorship" or "Design Rotation Program" type programs? targeted towards fresh grads, no specific job posting, 18 month positions that rotate you around different departments to see where your skills fit in or what you are interested in?
[11:06:10] <branjb> an excerpt from the posting, "Work from 18 months to three years to learn our products, customer needs, and markets. During this time, you interface between National Instruments and our customers to find creative and timely solutions to customers' technical issues as well as work on a variety of technical projects. Through this technical support role and project work, you gain the critical
[11:06:10] <branjb> depth and breadth in your technical skills required for you to grow your career successfully elsewhere within NI. "
[11:10:28] <z64555> yes
[11:10:46] <z64555> those are top-of-pile applications
[11:12:28] <z64555> If hired, you'll get /usuable/ experiance that's applicable to similar fields
[11:13:23] <z64555> And gets you connections within the company, which would help further your career within the company and with others
[11:13:27] <branjb> yeah I'm interested in the companies anyway so those sound like pretty good deals
[11:13:47] <z64555> way better than the hat raffle, at least
[11:14:45] <branjb> only really starting the job search now since I'm closer to graduating, and finding a lot of things that raise red flags (ie, startup companies that don't really seem to know what they want)
[11:16:42] <branjb> targeting the austin, tx area and there seems to be a lot of reputable/established companies already
[22:16:57] <justanotheruser> If I have a stepper servo motor going at 2000RPM and I am sending pulses to it and suddenly stop, what happens, it just halts?
[22:18:50] <Tom_L> it should
[22:19:13] <Tom_L> it will decelerate
[22:19:25] <Tom_L> may miss a few steps stopping
[22:49:20] <z64555> Yeah, the load and no-load inertias determine how many miss-steps it might have
[22:54:05] <Anniepoo> 8cD
[22:55:09] <z64555> Hi annie, justanotheruser is playing with steppers
[22:56:20] <Anniepoo> fun
[22:56:48] <Anniepoo> i'm wondering if mbrumlows around
[22:57:09] <Anniepoo> trying to get more info about his bot
