Hi Buddy,  - Thankyou, this circuit is working really well for me now, i have made up 3of them on a single board for use with a few projects that require really fast switching....   - -  the rise and fall time is really fast on this circuit diagram simple due too the fact that - the photoisolaters are designed too be pinned too the power rails with 2-3k resistance, (in the datasheet). which delivers a 8v signal, - however in the datasheet, that also recommend a 100ohm biasing resistor, which my circuit is based on.  -   ...   similar too the new signal generator i received from ali express today, the rise and full times look like the learning tower of piza, - ,,, 

but mine is super fast, - basically because it is burning lots of waste energy too tighten up the signal.  .  .. i really like doing this - it was not part of my teachings when i studied electronics, and i never see it done in practice,  - but im convinced its an important ascect of free energy, - because the efficiency of the spikes is directly related too the fall time, -

also please add "10volt zener" too the diode .   - as this is used too bring down a large positive bias , that the opto isolater is not able too,

thanks for this draw up by the way, - its critical infrastructure,  even though my will likely require a little ground balancing untill the 12v regulator is accurate.

hehe i smoked a few 1500v trannies tonight with my fall times, - but making lovely resonant waves,

I found this hex schmitt-trigger buffer in the ~6ns transition time, of course its low voltage (~6V) and needs a driver for higher current loads.
https://assets.nexperia.com/documents/data-sheet/74HC7014_Q100.pdf

Regarding the $3 0-150KHz signal generator I bought from AliExpress and posted about, I just looked at rise and fall times on it.
From 5 to 20V input(same as output), rise time was 42ns worse case (36ns typ) and fall times was 21ns, not bad for a couple bucks.

If you guys ever wanted a good green laser, this is one and only $7 with battery and charger. I have bought about 15 of these sofar, all test from 70 to 95mw, they are pretty strong and low cost https://www.banggood.com/Burning-Laser-303-Green-Laser-PointerLight-Star-Cap-532nm-5mw-p-955849.html?rmmds=flashdeals&ID=228&cur_warehouse=CN

Houston, we have a problem, lol. 40A in 1us. :wtf:

I tried a few standard snubbers, RC snubber, RCD (polarized), and straight brute force schottky diode.They do help the flyback voltage, but really only the ring portion, and the first avalanche is not changed much, the straight diode clamps the excessive voltage that is damaging my fets. (technically it's the energy in the avalanche event that heats the fet repeatedly, then finally she blows). The problem is the most effective flyback voltage clamping single diode severely limits my rpm, usually under 1 mil, anyone know why?

FYI:The 37 year old IRFZ44 fet is not ruggedized (taking repetitive avalanche events), like the newer fets IRFZ46 do.

I Forgot to mention, in another project, I was messing with a brushless hard drive motor with a single drive fet and signal generator. Why, to see if driving this awesome zero play motor can be used with a very low cost controller. This has zero feedback, so acceleration is limited and so is the output power. I am simply driving a phase with the fet. A little playing around and i set to 60Hz, and give it a little spin to kick start the motor., then I slowly ramp up the freq to increase rpm. This is a seagate hard disk drive that I robbed the magnets out of. Its is a 7200 rpm drive, but I used a single simple fet to drive a single phase (really two phases) with zero feedback, I got it to 25,000 rpm, pretty happy with that, ~3.5 times faster than normal 3 phase with brushless motor controller, it was limited mainly by my 20V power supply.I will switch to higher supply voltage next week.

Quote from Lynx on June 4th, 2019, 08:25 AM

I'm also experiencing this very problem, but fortunately my fellow staff members are über savvy when it comes to handling these kinds of problems, so they'll be checking this out.

No the thread is just fine, just keep posting business as usual :-)

Is there any resolution to this issue, I am concerned, friends won't come, please let me know a time frame, thanks.

Should read OK now,.
If not, try closing your browser, google Open Source Energy and open the link duly found, that should give you a "correct" certificate and the site should load just fine :-)

25,000 RPM hard disk drive motor, driven by simple single FET and function generator, with limitations.

https://youtu.be/SRk651ThM0k

That would be quite the control systems engineering challenge to build a micro controller that can detect the state of the motor purely from the current jitter and precisely step up the frequency.  That resonant looking jitter is very interesting.  It's obviously the difference between the mechanical phase of the motor and the electrical phase of the signal generator--I think they call it the slip angle or something similar.  Really cool how it's not what I would have expected though, that it seems to resonate.

I been flying quadcopters for about 7 years, the first brushless electronic speed control (ESC) where large, heavy, kind of expensive and slow response. they were made for planes where response was not a requirement, as your thro finger is only so fast (turns out its about 60ms with the short throw (45 degree) we use). Also the receiver only output at 60Hz rate.

Then a guy named Simon kirby (SimonK) rewrote the Atmel based assembly code for response for quads. night and day. The main reason is the flight control was running 300Hz PID loop, but the old esc was based on 60Hz, so the SimonK firmware was night and day as it could be updated upto 490Hz (something our fingers cant do, but the flight controller could). Then SK added a huge improvement, called comp_pwm. For those that dont know about Comp_pwm, instead of the CEMF current running through the FET body diode and heating the FET, the FET would be on to conduct the CEMF and this did two things, lower the fet temp but hugely, allows the esc to brake or slow down the motor rpm, before this the esc only had control of accelerating the motor, if it wanted to slow the rpm down it basically let aero drag slow the rpm. Again this was huge for stability of the quad.

Then BLHeli FW came out that was better, then BLHeli_S firmware came out that had a better hardware requirement, this was another night and day, and it accepted 32KHz updates, that our flight controller now output (32KHz gyro read/process/output in less than 32us, times 4 motors). Quad motors got better too, lighter (less than 1/3rd), smaller, more power, faster response and less cost ($10~$14), my quad motors can go from idle rpm (zero thrust setting) to 90% rpm in 110ms (around 32000 rpm change with 5 inch three blade props, full thro my quad draws 1800W 120A). These improvements and flight control hardware, crummy Atmel (8bit 20MHz) to stm32 F1 (32bit 70MHz), to now stm32 F4 (32bit 240MHz) processors and its firmware. The response and stability as freaking awesome now and this all happened in the last ~4 years.

BlackBox recording has revolutionized how we analyze the quads preformance, instead of saying "it feels better/worse" BB logging shows us what is really going on, pretty amazing.

Thanks for the tech brief Oz :thumbsup:

I got my first levitate today, it is finicky, i think I need better everything, lol!
The circuit is ultra simple, I need a better gate drive and much better coil and magnet. I used two little magnets with a tail (lead cut from a capacitor), its job is to stabilize the magnet at the top from flipping over. the current probe really gives great feedback to where the sweet spot is (distance from magnet/hall/coil/voltage applied).
The levitating magnet is circled in red, real crummy picture, trying to hold the hall sensor, get the phone pointed and triggered, and then blur, lol.
Why, this is a baby step to something bigger, yep, talk is cheap. :-D

More MagLev today, I was tossing out a dishwasher and I robbed a few items from it. The main water inlet valve looked nice, plus a few other motors/pumps/etc. My first attempt I got it to work, but the coil and fet would get hot, I added a 45A schottky and it was blazing hot. With nothing more than changing the coil, it runs on 21mA. Its more stable, but still not as stable as I would like.
The new coil is maybe an 1" by 1" and is 1.2H and 1KΩ. There is a few things to balance this system, the weight, the magnet strength, coil voltage, sensor position to name a few. Another baby step, lol.

If this small coil has 1.2H, I cant imagine what the newman has. The online calculator I linked works pretty well if you know a few parameters. Maybe you can give it a quick try to estimate the inductance in that huge coil. I was measuring some inductance of my quad motors, if you leave the magnet bell on, it really screws with the inductance numbers.
https://www.eeweb.com/tools/coil-inductance

Anyone that has built a maglev, probably knows there is a few things to help stability. The first is adding a tail to the magnet, gravity just pulls is down and this stabilizes the magnet from rotating (x & y axis, not z), I did the same thing (add smaller magnet or some weight). I played with it again and i got it more stable, enough to have a single ball levitate (but sometimes it does start to oscillate and falls, but it is better. I really like I can leave it running for hours with no baking the coil/fet/diode, that was a big improvement.

The ESC that I use in my quads are awesome and only 10~12 bucks each. 30A on 4S (up to 18V input) and they never fail with the hell i put them through flying. I think its the gate driver is one of the main reasons. So i took a spare esc, removed the SiLabs micro and brought out the 6 input signals (its three half bridge driver), made a configurable interface to my hall sensor, almost done. One of the main benefits I hope is the bidirectional driving will help clamp the BEMF, might just be wishful thinking, lol.

I forgot to attach the datasheet for the 3 phase brushless motor driver, its converted from chinese.

The 3 Laws of Science...
1. If it smalls bad, it's chemistry.
2. If it's mushy, it's biology.
3. If it doesn't work, it's physics

A bus station is where a bus stops...
A train station is where a train stops...
On my desk I have a work station...

Finding 2mm magnet balls is a bitch, lol. Many on alibaba claim to have them, but when trying to purchase them, they dont have them. Any suggestions?

I was going for the low hanging fruit (2mm ball) before venturing into plan B. Plan B has many more technical challenges, many baby steps, so results are going to take longer.

For some insight to my background, I been doing RC cars/planes since late 70's. Here is a 130+mph plane I been flying, at this speed and less than 10 feet away it has its nerves, lol

https://www.youtube.com/watch?v=HTL52e6OMj4

Quote from --Oz-- on June 20th, 2019, 01:25 PM

A bus station is where a bus stops...
A train station is where a train stops...
On my desk I have a work station...

:rofl2:

Maybe you need a play station instead.   :-D

Quote from --Oz-- on July 7th, 2019, 09:30 PM

Here is a 130+mph plane I been flying, at this speed and less than 10 feet away it has its nerves, lol

No way man.  My poor eyesight would guarantee how that story ends.  I would need to enclose myself in some sort of translucent cargo net.

From watching your quad videos, those things don't exactly crawl either.

I can sure see how you got hooked on this hobby.  Very impressive.

Thanks. I got hooked on racing rc on-road cars, love the building/tuning/racing/competition for 21 years. I liked planes because of the 3d and you get to fly for hours (slope soaring gliders or combat gliders). When quads came around, they were difficult to fly, but once you learn, they can do about anything you can think of, awesome 3d playground. Picture of some of the planes I fly.

Any updates on the newman motor?