STEORN DEMO LIVE & STREAM in Dublin, December 15th, 10 AM

[Omnibus]

Thanks again

I know having a higher MHz scope would be better - but would  a 20 to 60 MHz scope be useful for these motors or do you need higher ?

Bill

I think you need higher. Indeed, you can do some crude measurements with these in this particular experiment but given the controversy you really need to present high quality data.

[markzpeiverson]

Thanks again

I know having a higher MHz scope would be better - but would  a 20 to 60 MHz scope be useful for these motors or do you need higher ?

Bill

Bill:

I don't think I've seen anything that indicates you need to measure risetimes that are faster than several 10s of nanosecs, so I think any of the scopes with a BW of between 20 and 60 MHz are adequate.  Here are the specs from Saelig...

Specification
Model          PDS5022S   PDS6042S    PDS6062S/T    PDS7102T
Bandwidth    25MHz        40MHz        60MHz            100MHz
Rise time      ≤14ns        ≤8.75ns       â‰¤5.8ns           â‰¤3.5ns

Even the slowest can measure risetimes of close to 10ns.  Again, so far I can't remember any mention of needing that kind of performance on any of the components.  Well, wait a sec, someone did mention the turn-on time for one of the components and having to change to something else with a much slower turn-on time and he went from 100V+ to 3V... let me find out what that was.

-Mark

UPDATE:

Back at Reply#973 was the following post from @GroundLoop (with typos corrected by me)...

"The difference in output was dramatically reduced with a slower transistor. The hexfet I did use had a switch on time of approx. 54 nano second. The transistor switch on at maximum 600 nano second. The output voltage was reduced from 180 VAC to 3VAC."

The 54ns hexfet turn-on time is the only mention of sub-microsecond timings that I'm aware of, and its still 4 times longer than the risetime of the slowest scope above.  So I think any of those would do... if possible, I'd tend toward the 40 or 60 MHz models...

[markzpeiverson]

The difference in output was dramatically reduced with a slower transistor. The hexfet I did use had a switch on time of approx. 54 nano second. The transistor switch on at maximum 600 nano second. The output voltage was reduced from 180 VAC to 3VAC.

I think this is a VERY IMPORTANT piece of info regarding the sensitivity that different components can have on performance.

GL, can you please VERIFY this observation. Go back to using the hexfet and measuring the output V; then use the transistor, then back to the hexfet, then back to transistor.  If the output V changes consistently from ~180VAC (for hexfet) to 3VAC (for transistor), then you may have identified a critical design element of this system.


The only thing that makes sense as to why the turn-on time of the hexfet/transistor makes so much difference is the magnetic viscosity hypothesis... which, from experiments I've looked at on utube, is likely in the 10s of nano-seconds.  The turn-on time of the transistor is 10 to 40 times LONGER than the delay due to magnetic viscosity!!!! Thus, any effect from MV is destroyed by the fact that the transistor takes way too long to ramp the voltage up on the coil.

I know that GL is experimenting with a solid state version, so things are not rotating, but consider this in the mechanical version of Orbo.  Here we are, dealing with a motor that is running at maybe 1000rpm so far, which means that the rotor makes one revolution in 1 millisecond.  Even if one takes into consideration 4 coils instead of 1, then the coils are being pulsed every 0.25 millisecs.  This is ~ 10,000 times slower than the delay from MV.  All other timings on this thing are nowhere near nanosecs...

-Mark

[Groundloop]

@markzpeiverson,

I did blow my first hexfet, IRF PF50, so I tried a NTE2354 transistor. Almost all
the high voltage was gone from my generator coil. I then tried a IRF540 witch
is almost as fast as the PF50. Then I got my high voltage back again. So I can
confirm it. I will not switch to a new hexfet until I get more of the PF50 because
it is a lot of work un-soldering and then solder in a new hexfet. I'm also out of
solder wick.

Alex.

[markzpeiverson]

@markzpeiverson,

I did blow my first hexfet, IRF PF50, so I tried a NTE2354 transistor. Almost all
the high voltage was gone from my generator coil. I then tried a IRF540 witch
is almost as fast as the PF50. Then I got my high voltage back again. So I can
confirm it. I will not switch to a new hexfet until I get more of the PF50 because
it is a lot of work un-soldering and then solder in a new hexfet. I'm also out of
solder wick.

Alex.

Hi Alex!
So you have at least some consistent evidence that turn-on time of the switching component (whatever it is), will affect Vout... additional trials would certainly answer that question conclusively... sorry, but I'm a scientist as well as an engineer and the scientist needs statistically significant data!   


There's one thing that bothers me, and correct me if I'm wrong...
You are blowing devices that are spec'd at way more than 180VAC, and I would guess that you're no where near the current limitations of the devices.?  If so, then perhaps there are voltage/current spikes (from BEMF??) that are way too fast for your scope to resolve??  Experience has taught me that if there's something in your setup that doesn't make sense, something that you don't understand, then its probably wise to first determine what's causing that anomaly.

Thanks for keeping us updated on your excellent work!
-Mark