Electric Motor Secrets

[Mark]

Yea what the F? What he said, we can't wait forever. Is the force with you or not!!! WE NEED AN UP DATE!!!

[uusedman]

I have to agree with these guys. Jetis you got to post something.

Right

Right???

HOW long shall we wait still?

WHAT'S happening?

Don't You have the least compassion to our unbearable impatience regarding Your so far excellent project?

You could at least submit an empty post? just for chewing on?

::almost dead of waiting::



I hope You ok?

[Jetijs]

Sorry guys for the delay . I was SO busy in my day job and my other project to earn money (crisis you know), that I hardly found time to work on my motors. Also I found a problem in the circuit that needed to be fixed, that means getting the circuit out of the box and messing with all the wires again/ That took time. Also I did not want to make any posts without a video of the motor running as I felt that you are tired of my delays already

Anyway. Today I put some life in the motor. I only tested one motor this time and everything ran just great in the first try. I can switch each phase ON and OFF and the "on the fly" timing adjustment works just great. The motor just flew away. I did not measure any RPM's or any other data, just wanted to get this motor running for now. The ON time on the timing wheel is about 50 degree of an ark. For now I just ran this motor on about 12v and it is already so fast (and loud) that I am scared to step the voltage up Also the cooling fan is not attached yet so I did just some small tests. For the recovery on this test I used 1n5408 diodes which are not the best to use for this purpose, I will get some fast switching diodes soon. The vibration of the motor at those speeds is very low, so the balancing is more or less correct.
Here is the video:

YouTube - Lindemann Rotary attraction motor V3.0

Anyway, I am very pleased with the first results

[Mark]

LOOKING GOOOOOOD! Jetijs was that 5.4 amps it was using? I thought the 1n5408 were pretty fast diodes, what are you going to use? Have you got any ideas what kind of horse power that will put out. Looks sweet though, excellent work once again.

Off subject, Jetijs have you played around with a Tesla pancake bifilar coil at all? I was wondering if you could run a motor off one.

[Jetijs]

Mark, I could not find any data about the 1n5408 diode speeds and I looked in data sheets of several manufacturers. I don't think that they are good for fast switching. Sure, they will work, but for this purpose we need the fastest diodes we can find. I need something that can handle up to 1000v and is very fast. Unfortunately the fastest diodes can handle lower voltages. I was thinking of using something like MUR4100 (SI-D 1000V 4A/70Ap 75ns), UF4007 (SI-D 1000V 1A/50Ap 75ns) or UF5408 (SI-D 1000V 3A/150Ap 75ns). These are the fastest 1000V diodes I can get in my electronic shop. Any suggestions are welcomed here.

2.8A is what one pair of coils consume (one phase) and that is with the bearings not warmed up yet. If I ran it for a longer time, the speed would go up and the current down. Also this is with not perfectly adjusted timing yet. You can see in the video, that at this one time the motor starts spinning much faster, this is when I turned the second pair of coils ON so the current consumption doubled. So far I am not concerned about the current consumption, we will deal with that later. I will try to measure the HP when both motors will have their axis coupled and the ON times will be chopped to the correct frequency for best recovery.

I have not yet experimented with Tesla pancake coils

[StevanC]

pursuit for excellence

"The jedi council must know of this one..."



I can only say one thing, and that's two words:

"I congratulate You"

now back to work and do some benchmarks,
and make that on the double ("all in sixteen" would we say around here )

:threatening 'hands on hips' position:

[gyula]

Mark, I could not find any data about the 1n5408 diode speeds and I looked in data sheets of several manufacturers. I don't think that they are good for fast switching. Sure, they will work, but for this purpose we need the fastest diodes we can find. I need something that can handle up to 1000v and is very fast. Unfortunately the fastest diodes can handle lower voltages. I was thinking of using something like MUR4100 (SI-D 1000V 4A/70Ap 75ns), UF4007 (SI-D 1000V 1A/50Ap 75ns) or UF5408 (SI-D 1000V 3A/150Ap 75ns). These are the fastest 1000V diodes I can get in my electronic shop. Any suggestions are welcomed here.

Hi

The 1N5408 is a general purpose rectifier, if it were a fast diode its switching speed would have been included in its data sheet.
Better to use fast or ultrafast diodes for catching the flyback pulse. Here are some further types as an addition to your types mentioned above:
STTH5R06 600V 5A 40ns
RHRP860 600V 8A 35ns
STTH1210 1000V 12A 90ns
STTH1212 1200V 12A 90ns
STTH512 1200V 5A 48ns
STTH812 1200V 8A 50ns

rgds, Gyula

[erfinder]

I like this one.....

30A 1200V 40ns

http://ixdev.ixys.com/DataSheet/L189.pdf

[Peter Lindemann]

As Usual....

Jetijs,

Excellent, as usual. Watching you, methodically, go through the process of solving each problem and moving on the the next, has been an inspiration. You have not only built a beautiful motor, but you have produced a grand tutorial on how much work, skill, time, money, and PERSISTENCE it takes to advance this work.

Congratulations!

I have every confidence that your model, when fully operational, will demonstrate all of the characteristics this motor is capable of. I definitely owe you a beer. To bad all I can afford is an "e-beer".

Peter

[Jetijs]

Thanks Peter
I love "e-beers"
Thanks also to gyula and erfinder for the diode suggestions

[StevanC]

Jetijs, i saw the vid...

i had a talk to a friend of mine who's a EE from semiconductor background (riz of ex YU) and when i asked regarding diodes (1N vs UF) he expressed the following:

A diode in this configuration (the back-shunt on SSG) is by far more fast than a BJT - it's regarded that a 1N4007 is lots faster than a 2N3055 emitter to base - and the times You presented:
95 nsec is by far (twice) faster than 150nsec (0.15usec) the BJT operates at HARD SPEED.

Normally, there seem to be very little devices capable to switch as much current as a 2n3055 that fast and not require a lot more tension:

I learned the other day that to switch a IRFP260N really fast, one has to supply 10V minimum - the closer to 20V the better. And to saturate it to 10V during less than 100nsec one can easily require 10A to flow - now to supply as much current for similar time again a BJT is needed - the FETs need simply too much energy to be switched hard ("who's older chick or egg") so in the end we need a really hefty 40V 20A SOA BJT thats FAST.

Now I ask:
Why not drive it with a such one already @12V?

and IMHO that's EXACTLY what Bedini figured: It CAN be driven FASTER but the most efficient way is trough a single BJT rough ENOUGH and fast ENOUGH.

Please, do all tests with the FETs ans especially pay attention to efficiency, recovery and performance, for i 'have a feeling' that the most sweet lies in the garden of "BJT" and i mean above 60hFE while over 32A SOA...

I look forward to news from You

commercials:


-in case of abduction call our hotline-
RENT-A-RUG limited.

[lighty]

I learned the other day that to switch a IRFP260N really fast, one has to supply 10V minimum - the closer to 20V the better. And to saturate it to 10V during less than 100nsec one can easily require 10A to flow - now to supply as much current for similar time again a BJT is needed - the FETs need simply too much energy to be switched hard ("who's older chick or egg") so in the end we need a really hefty 40V 20A SOA BJT thats FAST.

You're not correct. MOSFET can be switched On and Off extremely fast as well as the IGBT (the other voltage driven device). Yes, one needs quite fast voltage impulse by source who can also provide short high current impulse. Hence dedicated MOSFET drivers Jetijs is using- check out their datasheet.

[StevanC]

No TABOO?

You're not correct. MOSFET can be switched On and Off extremely fast as well as the IGBT (the other voltage driven device). Yes, one needs quite fast voltage impulse by source who can also provide short high current impulse. Hence dedicated MOSFET drivers Jetijs is using- check out their datasheet.

I see - it's commonly accepted to think so. Now what do You think, provides the actual switch ON in those driver circuit ? Midget made relays?


It's BJTs that do the MAKE - nothing is faster than a hard driven BJT. AFAIK
)

The MOSFETs only provide a price advantage and size advantage for a given power to handle.


You tell me why they make hFE1000 rated 1000+V 300A BJTs for military?
*Switching 1:1000 ratio with a 0.6V tipover*


And i dint claim it pays better out with BJT, only it's possible to make a more efficient circuit?
the drive current can add to the power current - no loss.


What You think?

[theremart]

Looking at the Jetjis design..

I was thinking today about Jetijs setup of two motors. It reminded me of Bob Teals design which resembled a cam shaft in an engine. I was thinking that one could stack stators on the same shaft and stack the coils as Bob teal did...

But Teals design made alot more sense to me today when I looked at Jetijs new version of his setup.

[lighty]

@StevanC

The truth is that MOSFETs and other voltage driven devices are much easier to set up and it's possible to drive them in floating configuration like with pulse transformers or optocouplers.

Modern MOSFET's easily go into high MHz range although I don't see what that has to do with this particular use in attraction motors.

As for the 1000V/300A devices- they are now commonly replaced by IGBTs (now that the price dropped) especially in switching applications. And guess what IGBTs are also voltage driven.

In case of BJT drive current adds to the power current? Not exactly- some of the bias current will be a portion (a rather small portion I might add) of the total current through coils. Main driving current portion must be provided by driver circuit so no gain there.

However, I suspect this discussion don't have much to do with this thread. If you want we can take it to the greatest EDA forum on the net EDAboard.com although I don't have much time to engage into such discussions I'm sure a number of engineers will have something to say about this topic. However, you might be surprised that preffered switching device (especially heavy duty) are either IGBTs, MOSFETs or (where possible) thyristors.