Electric Motor Secrets

[Tecstatic]

Jetijs,

Is it possible for you to make a scope shot showing both the current through the FET and the voltage between source and drain of the FET ? This way we can see the losses in the FET.

Is the 24V supply voltage constant ?

What is the DC resistance of your coil ?

"There is another thing that doesn't give my mind a rest. Why is it that when I adjust the dutycycle the current waveform gets elevated?"

You can operate the motor in two modes:

1. Discontinuous, the recovery current falls to zero before the next power pulse. Gives max recovery.

2. Continuous, a new power pulse is started before the recovery current is zero. You see this as an elevated curve.
This gives more torque assuming the core is not saturated.

I do not understand that the diode get so warm, I just checked the datasheet. It looks like it is meant for line frequencies (50-60 Hz). The data sheet does not tell the reverse recovery time. I have never used this diode, maybe it is too slow for this use.

I just saw you have two optical forks, you don't need more than one 40106, two inverters to drive the input for the UCCxxx is sufficient to use one input and two outputs for each channel ( total of the 6 schmitt-trigger inverters).

Eric

[lighty]

@Tecstatic

Ummm, why would he need to introduce Schmitt triggers? His driving signal is nice with fast rise times. You do realize that the more electronic circuitry is introduced the more difficult it will be to troubleshoot it from a distance.

[Tecstatic]

True, but I like the idea of operating IC's within their operating specifications. Your suggestion to use a driver with an internal schmitt-trigger could be the answer.

[lighty]

You probably know that old engineering saying "don't fix it if it works". Also, most of the specs are more of the general guidelines and one very quickly learn when doing EE for living that calculated values and specs often can be taken very liberaly. Also in this case driving signal is almost perfect so obviously everything works just fine.

I wouldn't get into more complicated circuitry if not necessary and judging by the driving signal it's not necessary. I understand your wish to get it as perfect as possible and I'm kind of perfectionist myself when I engineer circuitry. If I was doing it myself I would do more advanced engineering just for the fun of it but then again I know my way around EE and I can troubleshoot it myself. Jetijs is in a different situation- he is not EE professional and he has to be guided through troubleshooting procedures over the internet.

In any case its Jetijs's device and I will help him with advice as much as I can no matter what he decides.

[everwiser]

You can proof circuits in a simple form but as time goes on, efficiency and longevity will likely require some amount of circuit complexity. If Jetijs' circuit is getting warm/hot and it's not the load, then either something is wrong OR the circuit needs more refinement.

Eric's recommendations come from his strong electronics background (based on his posted comments, he DOES have a strong electronics background) and his desire to aid Jetijs AND see a viable, efficient driver circuit that will be applicable and stable for all going down this road.

In the end, it is Jetijs circuit and his decision which direction to go with it. Why not encourage the brainstorming that Eric is bringing to the subject and be inclusive in your comments?

We're all in this together; even lurkers, such as myself.

[lighty]

Eric's recommendations are all correct. One can surely engineer more sophisticated circuitry. I do it for living and I wrote that if I was engineering similar stuff for me or other professionals I would surely employ much more complex circuits. However if something goes wrong in more complex circuit then you need more knowledge and experience in order to troubleshoot it. Jetijs is not EE professional and he has to rely on myself or Eric to troubleshoot it on the distance. And it is a major obstacle. Also, Jetijs is living in a location where getting even some basic components is a kind of obstacle for him.

Circuit driver stage at the moment works just fine as can be seen by the gate signal. So, at the moment no additional Schmitt triggers are needed. It's basically a disagreement in apporach between I and Eric on a purely professional level. In my view when something works just fine there is no need to make it more complex because you can only introduce more variables that can cause something to go wrong. And as I already stated, the more complex circuit is, the harder troubleshooting is.

If MOSFETs are only slightly warm to touch (and I understand that is so by Jetijs's description) there is nothing wrong with circuit or driving. The only thing he can do is to try to lower the ON resistance. He can do it by employing better MOSFETs with lower ON Rds or by paralleling several MOSFETs. If paralleling MOSFETs additional problems might occur so for the sake of simplicity I would chose the other way. Eric's suggestion for MOSFET with much lower ON Rds is sound. I would of course suggest something that is more accessible to Jetijs but the basic idea is the same.

The only thing getting hot in Jetijs setup is isolation diode and I already explained what he has to do in order to resolve that obstacle. There is a point of the current levels going through that diode but don't forget that there is a large buffer capacitor basically providing energy to the coils so probably a slower (and larger) diode could be used at that position in circuit.


All being said, Eric gave some good advices but I think it won't help the situation for the reasons I already stated. It could, however complicate things to a level where possible troubleshooting will be harder. I worked in ariforce for a number of years and I worked both on complex sophisticated electronics in US helicopters and on more simpler and more rugged electronics in Russian helicopters. Guess which one failed less often and which one was easier to troubleshoot and repair?

[uusedman]

I cant believe it took me this long to answer.
Let me know what you think. The switch is a magnetic reed switch, 50v 2amp.

[Jetijs]

uusedman, try this circuit:



Your reedswitch will eventually burn out if much current is going through it. Using the circuit above, the reedswitch will only need to switch small current just to turn the transistor ON. and the rest of the current will flow through the transistor.You can use 2n3055 transistor for this, they are cheap. As for the resistors, try different values (both the same) from 1k down to 200 Ohms.
This should improve things a lot

[lighty]

@uusedman

You cannot connect diodes in parallel. I already explained why in my posts to Jetijs.

[uusedman]

@uusedman

You cannot connect diodes in parallel. I already explained why in my posts to Jetijs.

I glanced over it, tonight, I will try a big diode.

[theremart]

cheap powerful reed switch.

Reed Relays

scroll down on this link...

-----------
(KO) DRR-DTH-445
The ideal Reed switch and here's why: 1. It has form C, SPDT contacts. 2. Contacts rated 250v, 5 amp, 30 watt. 3. Small size, 0.201"D x 2.25"L. The glass body is 1.437" long. The N.O. lead is 0.562" long. The common and N.C. lead are 0.25" long. 4. Highest quality by Hamlin. 5. Sealed tube is weather proof. 6. We have 18,000 in stock so save big bucks off the $6 OEM price. Trip reed switch with permanent magnet as in alarm system, or wrap tube with magnet wire for very low current switch.

$3.00 each...

[uusedman]

@Jetis

I had 2 new 3055 transistor from my bedini motor project. And, both burned Here is how:

I start the motor with the new circuit you provide and it was running well. So I i wanted to see the motor run under the condition of NOT collecting the back EMF, the motor got a little heavy RPM increase to about 510. Then, when i close the circuit to collect the back EMF while the motor is running, the transistor gets burned. Also, the transistor was getting really hot, is that normal?


The first transistor was burned having 470 ohms on base and 470 on to emitter. Second transistor was burned with 940 ohms on base and 940 on to emitter.

The motor did look running smother while collecting back EMF however, the RPM did decrease about 10%. Maybe in order to get the rotary attraction moving, it should be above a certain RPM (which is to be announced).

I am still using the previous inductors of 22 gauge, first inductor being 110 windings and second being 135 windings. Going to use 16 gauge on the next experiment, once i get more or better transistors.

[uusedman]

@Jetis

I had 2 new 3055 transistor from my bedini motor project. And, both burned Here is how:

I start the motor with the new circuit you provide and it was running well. So I i wanted to see the motor run under the condition of NOT collecting the back EMF, the motor got a little heavy RPM increase to about 510. Then, when i close the circuit to collect the back EMF while the motor is running, the transistor gets burned. Also, the transistor was getting really hot, is that normal?


The first transistor was burned having 470 ohms on base and 470 on to emitter. Second transistor was burned with 940 ohms on base and 940 on to emitter.

The motor did look running smother while collecting back EMF however, the RPM did decrease about 10%. Maybe in order to get the rotary attraction moving, it should be above a certain RPM (which is to be announced).

I am still using the previous inductors of 22 gauge, first inductor being 110 windings and second being 135 windings. Going to use 16 gauge on the next experiment, once i get more or better transistors.

@Lightly

I used 5407 diodes, is that too big?

@theremart

thanks for sharing that info

[Eric]

re: burned trannies

hi uusedman!

just to let you know transistors are very easy to burn out when not used properly so make sure to buy a bunch. the mjl21194g by ON-semiconductor is a very good npn bipolar transistor for building the lindemann circuit. i have had a lot of success with this transistor. when used right, it never heats up in my curcuits. next it helps when ever you post a question regaurding the circuit you are working on it helps to post a drawing or reference link to a previous circuit in a past post. try to be specific in the drawing with regaurds to how you hooked each component up I.E. polarity of the pathways whether its an NPN or a PNP trans. and how you hooked up the collector, emitter, and base. chances are, if you are blowing the trans. that quickly, its hooked up in a way its not supposed to be.

hope that helps!
Eric

@Jetis

I had 2 new 3055 transistor from my bedini motor project. And, both burned Here is how:

I start the motor with the new circuit you provide and it was running well. So I i wanted to see the motor run under the condition of NOT collecting the back EMF, the motor got a little heavy RPM increase to about 510. Then, when i close the circuit to collect the back EMF while the motor is running, the transistor gets burned. Also, the transistor was getting really hot, is that normal?


The first transistor was burned having 470 ohms on base and 470 on to emitter. Second transistor was burned with 940 ohms on base and 940 on to emitter.

The motor did look running smother while collecting back EMF however, the RPM did decrease about 10%. Maybe in order to get the rotary attraction moving, it should be above a certain RPM (which is to be announced).

I am still using the previous inductors of 22 gauge, first inductor being 110 windings and second being 135 windings. Going to use 16 gauge on the next experiment, once i get more or better transistors.

@Lightly

I used 5407 diodes, is that too big?

@theremart

thanks for sharing that info

[uusedman]

@Eric

This is the circuit i used.
I will make a detailed video once I get more transistors.

First experiment,
Resistor 1 470 ohms
Resistor 2 470 ohms

Second experiment,
R 1 940
R 2 940


@Jetis

Any reason why the RPM decrease when collecting the back EMF?