[Robert Murray-Smith] doesn’t like the price of inverters to convert DC to AC. That led him to build a dynamotor, or what is sometimes called a motor-generator set. These devices are just DC motors driving a generator. Of course, motors can also be used as generators and [Robert] had a stack of brushless motors in the form of PC fans. A two-fan dynamotor was born.
The brushless motors are attractive because, traditionally, the brushes are what usually fail on a dynamotor. The fan that will act as a generator needs some surgery, but it is simple. He scraped off all the control electronics and connected wires to the coils to form a three-phase generator. There’s no need for the fan blades in that configuration, either. If you were using ordinary motors and a generator, getting shafts concentric would be an important task. With the fans, it is simple to just line up the mounting holes and you get perfect alignment for free.
How does it work? [Robert] has a second video showing the output on a scope. You can see both videos below. The dynamotor makes a good-looking sine wave, probably much better than most reasonable-priced solid state inverters. He didn’t mention how much current he could successfully draw, but it probably isn’t much. You’d also need a transformer to replace a commercial inverter that would put out line voltage, so that would be some more loos in the system. On the other hand, if you wanted AC at a lower voltage, you might just replace all the transformers, if you were building a piece of gear yourself.
We’ve looked at how these things work in some detail. There were common in old tube radios, particularly military ones.
Thing is, to drive the (fan) motor, you need that three-phase AC right away!
What to do, what to do?
Functional folks invented the fixpoint combinator for this. Or something.
it’s BLDC, not AC…
If you look after the electronics, ist’s AC. It’s a BLDC if you wan’t to hook it up to DC and put complicated switching electronics in between. If you get rid of these, you can plug AC and see it spin.
But I too don’t see the problem in this exact implementation, as he uses a fan that has an embedded ESC, but to other’s this is important detail.
A BLDC is a three phase AC motor, just for lower then line voltage. The driver circuit is basically a simple VFD.
Well, no. A BLDC lacks a rectifier stage at the input so you can’t run it off of AC, obviously.
Secondly, the driver resembles a variable frequency drive, but where VFDs nearly always run open loop, a BLDC always runs closed loop: via hall sensors, quadrature sensors, encoders, or block commutation, by switching two phases at a time and using the third phase to sense the back-EMF phase angle.
It’s possible to run the electromechanical part of a BLDC motor as a synchronous motor, but it’s difficult to accelerate without losing sync. Proper synchronous motors usually have an induction cage to provide self-starting, so you can just flip the power on and it accelerates itself to synchronous speed.
That just works until you put some charge on it.
To get the same current that in solid state inverters you will need huge motors that are many times the size and weight of a solid inverter. Don’t even think about efficiency.
Besides this system to get the target voltage need to be operating at the same speed no matter the current you need.
Nice experiment but totally useless in the real world unless you are into heavy industrial power conversion.
BTW, the board of the fan is a DC to AC converter and you will get much better efficency and no wear replacing the 2 motors with one transformer, so making it … a inverter.
Guys, yeah. It’s pointless, the BLDC is an inverter. Of course.
But it’s a great, fun, pointless hack! I love it!
The BLDC driver produces 3 phase square waves, not sine waves. The output from the driven motor is 3 phase sine wave, which is the whole point of this hack.
Now think about uses for it… AC is often more useful than DC, so converting DC from a battery charged by a solar or wind turbine to 3 phase AC could be very useful. Not using fan motors, of course, but there are plenty of miniature, high power motors available very cheaply in the hobby RC market that would provide useful output power, possibly at much higher voltages, than a couple little fans.
We now the fan drivers use squared waves but you really dont need a lot more to using the same principle do it with sine waves (and transformer instead motors).
And with motors you still have the problem of the (awfull) efficiency. Because to get the desired voltage you need the motorshave to spin at the same velocity all the time.
The RC motors won’t fix a thing. You still will need it running at full speed to matain the same voltage, no matter how much current you need. And fix that problem with ac/ac or ac/dc converters defeat all intention of no using inverters and you end using inverters AND motors.
Nice and fun experiment but no real application.
This setup is terrible as an inverter, because the frequency and voltage are mutually dependent, and both vary with the load. The output isn’t really a sine wave either but some mangled-up version that depends on the exact geometry of the magnetic circuit. It’s noisy, and powerless, and not actually three-phase but two-phase as most fan motors are.
But there’s one thing that you can do with it, and that’s a table-top model of an electricity distribution system. You could build a couple little “power plants” and distribution transformers and synchronous/induction motors that run off the power, and show how they interact with each other.
Exactly! Who knows where your musings might lead you?
The drive circuit relies on a hall effect sensor as feedback., so you might have to fool it somehow.
(The sensor is in the same IC package these days.)
I get the idea is not so much to produce A/C “work” current, but more to produce a pure “reference” sine wave, as opposed to a solid state “mostly” sine wave.
Might be a bit more intriguing to start with a couple of 10k rpm HDD motors instead of fan motors. Just be damn sure you get them on center to the Nth degree!
If you want a pure sinewave for a reference, you could do much better with a cheap pair of op-amps, or even a simple phase shift oscillator and a low-pass filter! This is absolute crud.
If you want complicated, how about using a rotary phase converter (1PH AC motor driving a 3PH AC generator) to power a 3PH motor, which drives a DC generator, to power a DC motor?
That’s what people with old Monarch 10E series metal lathes do, when they want to keep them original, don’t have 3PH power and don’t want to use a solid state VFD or phase converter.
There’s an article idea. Dig into why a *simple* solid state phase converter that just converts 1PH AC to 3PH AC costs so much more than a digital VFD with 100+ settings, variable speed output, a braking resistor connection and an option to add wired remote control.
Because the VFD output is really dirty PWM square wave at a high frequency and not meant for any longer runs of cable, since it would start to radiate a load of EMI everywhere. A shielded motor cable is always recommended.
The phase converter would include a rather complicated filter in front, which you can also buy for the VDF, called a dv/dt filter which turns the current sinusoidal to get around the EMI issue, and for high power three-phase ones you run up to thousands of dollars again.
That reminds me about this locomotive: https://en.wikipedia.org/wiki/SNCF_CC_14000
http://web.archive.org/web/20070306231356im_/http://belph80001.free.fr/Locoelec_Section/croquisweb_fichiers/image034.gif
It’s a cascade of transformer, single-phase AC motor, 3-phase AC generators/motors and DC motors/generators supplying the traction AC motors instead of just a VFD.
It is basically a diesel locomotive where the engine has been replaced by a fixed-frequency AC motor. There were no solid-state electronics that could handle so much power at the time.
How many loos in the system before you can call in a plumber?
Oh good grief, “pure sinewave”, distorted piece of… and use it as a function generator? Late April’s fools?
Those old dynamotors generated HV dc output, as I recall, for tube plate voltage.
This project reminds me of some other old tech: back in 1902 the NY subway rectified power for third rail using synchronous motors with switching commutators: it was mechanical synchronous rectification!
About 25 years ago I was trying to design a 100khz switching supply, but no diodes were available that would switch the output at that frequency for the voltage I needed.
Later (too late) I remembered those old synchronous rectifiers… with a little control circuitry I could’ve used a mosfet as an synchronous output rectifier.
My point is that silly hacks like this might stimulate creative thought that makes other solutions possible.
Neat, hadn’t thought of using a wee little fan as a generator. Might be an upgrade to the little brushed motor on a hot air engine, after tidying up and removing anything that looks like a fan. Another for the deepest darkest part of the back burner :|
Pretty sure that if the motor and the generator are separate devices connected by a shaft (i.e. you can unbolt and replace one or the other), it’s a motor-generator, but if the two are inseparable (looks like a brush motor with two sets of brushes), it’s called a dynamotor.
One possible niche application would be if you really need a lot of isolation between the two sides, just couple the two together with nonconductive rods and it’s easy to get the maximum isolation voltage well into megavolts by using sufficiently long rods.
This guy’s whole channel comes down to perpetual motion machines and other pseudo-engineering. Don’t waste your time.
Well there is a way of doing this that is so simple I can’t believe people are overlooking it unless they’re just purposely doing so. If you have relays that drop in and out depending on the output current then it is possible to create and entirely perpetual electrical system it all goes down to multiple stage input output using relays and the mechanics obviously block and tackle type gear systems
Cool hack, but pure dine inverters are cheap. Unless you’re doing ham radio stuff or live by someone who is, and need more radio silence than the cheap ones offer…. Solid state. Always solid state. It’s awesome!
If you do need radio silence, I think I’d rather try to DIY an ultraclean inverter with some kind of linear transistor post filter that can remove the noise or something.
Why is everyone mentioning 3 phase AC?
These fan motors only have 2 phases.