@PlasmaHH +1

The only valid answer would probably be "I am a metrology nut" but then again, asking here...

I can quite easily build him something that surpasses the requirements.... But he won't want to pay for it

cheap bastard...

oh btw. anyone has a spectrum analyzer for free?

@PlasmaHH If you find one, see if they've got a spare. I would like to have one, too.

@PlasmaHH Can the shipping costs be €4500?

@Asmyldof Only if you have a "shipping for free only this week" coupon... ^^

@PlasmaHH I do, but you have to do about 2 months of free billable work to get one

@Asmyldof I kinda prefer "working" here for the money (read: surf the interwebs and talk with weird people online)

@PlasmaHH Eh

I hope I return from Aussie to a few assignments, so... let me know if you change your mind :-P

@Asmyldof Are you going to Australia?

I hope so

Else those tickets were a HUGE waste

Work or vacation?

vacation? him?

A true hero of neuroengineering, DA Robinson, died yesterday. jap.physiology.org/content/21/3/1068.short shows an early and important work that birthed decades of research.

@PlasmaHH lol

It's weird how well you guys seem to know me through only online interaction

@JRE Going to visit my Brother and his Fam In Law, because I need to waste 4 weeks worth of time, but I am bringing my Altium-Capable Laptop, so... very likely going to be no better than 50/50

Haven't seen my brother in about 7 years now, so...

@ScottSeidman 's not loading. What's it centre around?

@Asmyldof I know how that goes. Germany to US isn't as far as NL to AUS, but you still don't get to visit often.

Yesterday I found my "Windows Plugin Pack" again... basically a Windows 98 explorer I programmed to wrap MS Explorer, which let you type any character you wanted in file and folder names (and some other new features) and it translated them to combinations of legal characters on the disk

@JRE Had I not been busy with a thousand things all the time I'd have probably visited two times already, and for him the same goes... It's in the family, I suppose, this "Do nothing, get bored, do something" drive

@Asmyldof It loaded for me. Seems to be about implanting detectors in (or around) the eyes of monkeys to see which way they are looking during certain types of experiments.

From 1965.

@JRE Aaaah. The great days before Computer Vision was invented and Viciously Maiming Animals was abolished.

These days if you want to viciously maim animals you have to call yourself a hunter with slightly bad aim

(Though I dare not claim the research was not useful, just to be clear)

I really hate it when on a private ebay auction you get the item damaged during shipping

Anyone got advice on trying to drive a small ( < 1 W) BLDC motor to 60Hz with extreme accuracy (like 0.001% error or less)? Any integrated drivers with really high resolution PWM/feedback capabilities? I have a Texas Instruments DRV11873 but it appears to convert my 12 bit PWM down into some shitty 7 or 8 bit PWM control over the motor

There are significant "steps" on the output drive control in terms of resolution

What now, 0.001% or extreme accuracy? ^^ also did you calculate how many bits you actually need?

Atmel and an H bridge

@KyranF You shouldn't need extreme resolution of your PWM. To achieve that kind of accuracy you're going to need to use feedback and adjust the PWM occasionally to keep the frequency matched. The adjustments to the drive strength will average over time to a much more precise value than the resolution of your PWM at any given moment.

the integrated driver has huge aliasing on the output resolution of its own internal PWM output, I need sub-hz steps on the "frequency generator" feedback which is 6 pulses per rotation of the BLDC motor. From some testing it appears that just analogue control of the motor driver chip motor VCC input voltage gives the kind of resolution/continuous adjustments that i want. Right now, a "step" of PWM gives like 6-10 Hz on the FG feedback.

so the plan is to provide a vaguely close PWM to get with 5-10% of the setpoint and then go over to a VCC adjustable regulation circuit to do small excursions around the usual 5V to get the fine speed control I want

for example I find that 61% PWM duty cycle and then adjusting VCC from 5V to 4.93V gives very good results

so this mornings tests have been productive! now for a digitally controllable VCC-adjust circuit... any ideas for that? I'm going for a LM317 + DAC for adjustable VCC into the motor driver for now

@PlasmaHH if the native PWM output of the BLDC driver chip gives obvious steps of 10Hz on the FG feedback as the high res PWM input is slowly increased, and I want more like 0.5Hz or better steps on FG, then I need x20 more step-count than whatever the IC has built into it. Let's say it's 8-bit (even though it appears to be more like 7-bit), 255 steps would need to be 5100+ to get the resolution i'm going for. So thats more than 12 bit.

@ThePhoton I don't have external feedback, only feedback generated by the driver chip, and the driver chip is down-converting the PWM input to a much lower resolution PWM output

Anybody ever used one of these beasties?

ti.com/lit/ds/symlink/lm3900.pdf

I'm guessing that the fact this is the first time I've run across this amplifier design means it never found any huge number of applications.

@ThePhoton Perhaps not in our day and age:

> SLOS059 – JULY 1979 – REVISED SEPTEMBER 1990

@NickAlexeev Yeah I saw that...but somebody must be buying a few of them to keep them in stock.

@ThePhoton They were reasonably decently used, as their near-to-RtR output capabilities were very good for the day.

It is not unlikely people still use them

Though half the argument for production may be Legacy Maintenance

Wouldn't know for sure

Haven't designed with one since I stopped having time for serious Analogue Hobbying

Mind you, I'm more used to using the N.I. variant of this one: ti.com/lit/ds/symlink/lm359.pdf

@Asmyldof Hm, DigiKey has got a thousand in stock. Commercial and industrial temperature ranges. The datasheet doesn't mention anything about MIL-spec. These might still be used for some disposable medical device designed back in the day.

Norton / Current Differencing amplifiers are incredibly useful, though the LM?900 types are OLD. But we all know people still get taught with and then end up using the IRF(Z)44 and such

So...

Anyone's best guess what the purpose of their continued existence is

I'd genuinely not be surprised if some group of universities ends up being the other half of the annual sales

@NickAlexeev here's the 1972 version.

@ThePhoton Ah! The good old N.I. I remember from all the databooks now lost in time. With complete internal schematic. Something TI would never allow

@Asmyldof National Semi, not National Instruments.

Right, Of course

I too have my moments of not moments

> p.s. Either way, let's call this thing Zanus-Alexeev device. Oct 19

Oh dear, I may have inadvertently proposed to a girl via EE.SE. There goes my diamond.

@NickAlexeev proof or it didn't happen

i found it

lol

well, maybe you'll both get rich and live happily ever after ?

@KyranF IP dispute may get resolved by dynastic entanglement.