I was pricing oscilloscope today and they are very expensive.

No problem. You can slow down the 555 frequency and will be able to measure the voltages at the output of the 4017 with a multimeter.

My only concern is the proper switching of the IRFs, a scope would be better... But is not absolutely necessary.

 

I am in the process of drawing up an actual layout of what components go where on the vero board. Talk about tedious work.

Actually not. Even though this power supply has a short circuit protection I would be starting with a lower voltage for the power part.

If this is not possible I would use a separated 12V power supply (as you suggested) for the IRF/magnet part and put a resistor of a few kOhm in series with "+", not connecting the magnets/LEDs. Then you could measure the voltage pulses at the points where you will connect the magnets too.

If there is any cross conduction between the FETs or wrong wiring, a voltage will appear over the series resistor, if everything is ok, no voltage will be present (since there is no current)

If I understand correctly I will have a separate power supply right where it is blocked out in black? Right?

And when I check for 0v, I will set the voltmeter at these points?? Ohh I just realized that the resistors are 100 ohm I meant to put 3k ohm.

No problem. You can slow down the 555 frequency and will be able to measure the voltages at the output of the 4017 with a multimeter.

My only concern is the proper switching of the IRFs, a scope would be better... But is not absolutely necessary.

When you say "output of the 4017" do you mean all of them? O0-O6?

If so what should be my reading? It will vary right?

 

If I understand correctly I will have a separate power supply right where it is blocked out in black? Right?

yes. on the other MOSFETs/diodes too.

And when I check for 0v, I will set the voltmeter at these points?? Ohh I just realized that the resistors are 100 ohm I meant to put 3k ohm.

you can do that. but I would actually measure one ofter another. then you also don't need the resistors since the input resistance of the voltmeter is very high.

When you say "output of the 4017" do you mean all of them? O0-O6?

If so what should be my reading? It will vary right?

Yes the six outputs. If you lower the frequency of the 555 (increase the timing capacitor) you will be able to see the outputs turn on and off, one after another. The go from zero to almost 12V.

 

I want to make sure I have this right, Can you tell me if the gate, source and drain is right on my drawing?

I'm not 100% sure about the 9540.

thanks

 

I want to make sure I have this right, Can you tell me if the gate, source and drain is right on my drawing?

The arrow for the 9540 is inverted. You can also refer to the datasheets.

I was thinking again about the power supplies and I see that it wouldn't even work with a lower voltage for the power part. The 9540 is switched on and off by the voltage difference between gate and source. If the two power supplies have the same ground reference and the one for the magnets is lower it could not turn on the PMOS FET.

However. I think now that after testing the circuit it might be better to use only one power supply for the whole circuit. We'll see.

 

I don't want to sound a like a kill-joy, but going back to an earlier suggestion; You could use a steel bar as the armature (not a bar magnet) and then the entire circuit becomes a 4017 and 3 transistors to drive the 3 coil pairs in sequence. And a single power supply.

Is it really worth the enormous increase in circuit complexity and work to make that little bar magnet go round and round rather than a little steel bar go round and round?

Either way you can put some LEDs on the coils, so the students can still see the coils activated in sequence 1-2-3.

 

and 3 transistors to drive the 3 coil pairs in sequence. And a single power supply.

Either way you can put some LEDs on the coils, so the students can still see the coils activated in sequence 1-2-3.

Yes,would be much easier. But I have the impression the the OP wants to show the students the attraction and repulsion behaviour of magnetic poles, not the attraction between a non-magnetized ferromagnetic material and an electromagnet.

He couldn`t identify one side of the iron bar as being north or south.

 

I don't want to sound a like a kill-joy, but going back to an earlier suggestion; You could use a steel bar as the armature (not a bar magnet) and then the entire circuit becomes a 4017 and 3 transistors to drive the 3 coil pairs in sequence. And a single power supply.

Is it really worth the enormous increase in circuit complexity and work to make that little bar magnet go round and round rather than a little steel bar go round and round?

Either way you can put some LEDs on the coils, so the students can still see the coils activated in sequence 1-2-3.

I do get the complexity and I realize that I'm way over my head, but that's the whole reason I'm attempting this, it's all still new to me and the novelty hasn't wore off yet. I can definitely see that the circuit is way overboard, that's one of the reasons why I'm so excited to do it. If/when this is a success, In my eyes, it's going to be like landing on the moon for the first time (I will probably be the only one sharing that feeling) but it will be such an accomplishment for me.

Thanks to everyone's help on this project (especially praondevou), it has sparked a thirst for knowledge in this hobby, I have been reading a through the volumes and learning so much.

Is it really worth the enormous increase in circuit complexity and work to make that little bar magnet go round and round rather than a little steel bar go round and round?

Before I started this thread, I thought that I would've had to compromise and do something similar to what you suggested, I was willing to do so. I never thought that everyone would go out of their way like they have to help make this possible.

I am in still working the layout of the physical components on the circuit board, I'm being extremely meticulous with everything, making sure that I didn't put a jumper wire where it wasn't supposed to go and ensuring that everything is connected correctly. It seems that the package will be here on Monday, which gives me plenty of time to perfect the layout.

Yes,would be much easier. But I have the impression the the OP wants to show the students the attraction and repulsion behaviour of magnetic poles, not the attraction between a non-magnetized ferromagnetic material and an electromagnet.

He couldn`t identify one side of the iron bar as being north or south.

The students have a hard time grasping this part of the lesson, most of them that come through are visual learners and need to see it in action to wrap the minds around the concept. The next time I teach that particular block of instruction will be in 2 weeks, I'm hoping that it will be done and I'm sure this is going to help.

 

Looking at the first post again, did you put a layer of tape over the threads before you started wrapping the magnet wire on them?

Also, how many turns of what gauge magnet wire is on each screw?

 

Looking at the first post again, did you put a layer of tape over the threads before you started wrapping the magnet wire on them?

Yes, I used black electrical tape.

Also, how many turns of what gauge magnet wire is on each screw?

Off the top of my head I can't remember the gauge, and if I recall, I did 180 turns on each bolt, (which was tough because my kids were distracting me while I was doing it...)

 

I don't want to sound a like a kill-joy, but going back to an earlier suggestion; You could use a steel bar as the armature (not a bar magnet) and then the entire circuit becomes a 4017 and 3 transistors to drive the 3 coil pairs in sequence. And a single power supply.

Is it really worth the enormous increase in circuit complexity and work to make that little bar magnet go round and round rather than a little steel bar go round and round?

Either way you can put some LEDs on the coils, so the students can still see the coils activated in sequence 1-2-3.

But would the steel bar still rotate? The magnet gives the attraction/repulsion needed to spin. I'm pretty sure just a ferrous armature would just sit there unless the switching rate was very slow. When this is done, maybe JCOX could try with both types of rotor?

 

But would the steel bar still rotate? The magnet gives the attraction/repulsion needed to spin. I'm pretty sure just a ferrous armature would just sit there unless the switching rate was very slow. When this is done, maybe JCOX could try with both types of rotor?

It will rotate. There is attraction of the ferromagnetic material to the magnet.

If the where only 4 coils at 90 degrees each you wouldn't know which direction it goes, though. Since we have a coil each 60 degrees the steel bar will follow the magnetized coils, it will always try to align with them.

 

Quick question, do I need to use heat sinks on the FET's?

 

Quick question, do I need to use heat sinks on the FET's?

If you really have only 1A through the coils then you will not need a heatsink.
You said however you measured 0.7Ohm at 13.8V. That's much more than 1A.



EDIT: I was seeing this as an almost static system. At higher rpm you may need a small heat sink, yes.

 

can I connect all of them to the same heat sink? or do they need to be separate?

 

I do get the complexity and I realize that I'm way over my head, but that's the whole reason I'm attempting this, it's all still new to me and the novelty hasn't wore off yet. I can definitely see that the circuit is way overboard, that's one of the reasons why I'm so excited to do it. If/when this is a success, In my eyes, it's going to be like landing on the moon for the first time (I will probably be the only one sharing that feeling) but it will be such an accomplishment for me.

Thanks to everyone's help on this project (especially praondevou), it has sparked a thirst for knowledge in this hobby, I have been reading a through the volumes and learning so much.

I absolutely respect what you have said there and I must say it is wonderful to see a great attitude towards learning like that from someone who teaches others.

It would be great if all tech subject teachers had your attitude!

And sorry for derailing your thread!

 

can I connect all of them to the same heat sink? or do they need to be separate?

The drains of all FETs are also connected to the mounting flange (is this the right word? I mean the thing with the hole in it). Since the drains of a 9540 and a 540 pair are already connected together they can be mounted on the same heatsink.

The other pairs cannot be mounted on the same heatsink unless you use isolation pads which I do not recommend for TO220.

IF a heatsink will be necessary then only at high RPM for longer periods, if the current through the coils is really only 1A. And then only a small one.

 

if the current through the coils is really only 1A

That was constant amperage. When you mention the short circuit protection it got me thinking that maybe the power supply is lowering the voltage because it thinks there is a short. So I went down and tested it and sure enough when I hook up the electro magnet the voltage drops to 0.66v. I perceive that to be a big problem, do you guys agree? Do I have to get a power supply that doesn't have short circuit protection. Which one would you recommend?

EDIT:

what if I were to increase the resistance in the coils by using thinner gauge wire and wrap more of them? It seams like that will be like chasing your tail, but you guys have the knowledge...

 

I absolutely respect what you have said there and I must say it is wonderful to see a great attitude towards learning like that from someone who teaches others.

It would be great if all tech subject teachers had your attitude!

thanks!

And sorry for derailing your thread!

No need for apologies, I understood where you were coming from.

 

I perceive that to be a big problem, do you guys agree?

Yes, indeed.

Do I have to get a power supply that doesn't have short circuit protection. Which one would you recommend?

No. To maintain 12V at 600mΩ coil resistance you would need a much more powerful power supply. Your wire would probably not support this current neither.

what if I were to increase the resistance in the coils by using thinner gauge wire and wrap more of them? It seams like that will be like chasing your tail, but you guys have the knowledge...

That's the way to go. 12V relays for example have several hundreds of Ohms usually.

According to this calculator you would need a lot of turns though to get 10Ω (1.2A at 12V). It tells me 50m of 28AWG wire... How much wire did you use and which gauge?

The other option would be to lower the power supply voltage but then we get to a limit very soon (probably at 5 to 6V) because of the p-channel FET (the 9540). If we lower the voltage at it's source to much we won't be able to drive it anymore since the gate needs to be low enough to turn it on. Lowering the power supply voltage also means you would need a different power supply or a voltage regulator.

The same problem doesn't exist a fullbridge with high side n-channel MOSFETs. Since you bought 8 n-channel MOSFETs there is maybe a possibility to do something about it without the need to rewind all the coils with much more turns... 0.6Ω however is really not enough I guess. So you will need a power supply with less voltage anyways, the best would be one with an adjustable output (from zero to max) or a lower voltage fixed output power supply.

Don't worry you get this working. You can always start with the logic part, see if it works and then build the rest. When we figured out how to do it.

Edit: something else I just saw. The current for the gate charge of the MOSFETs comes from the CMOS gate outputs. They cannot provide very much current which will slow down the FET gate charge, it will possibly also cause an increased power dissipation in the CMOS gate output stages, although I cannot say to what degree. In other words, this also puts a limit to the maximum rotational speed.