Need help with high power inductive/electromagnetic load

ionhunter

Hi all


I need a schematic of following for a BS2: (I have the BOE model)

Power source: ~28 Volts, 30 Amps, in that range
Drive at least 2 (ideally 4 or more) inductive loads (electromagnet, about size of fist), at constant
voltage (~26 to 30 V) and incremental Amps from 0-30. Each electromagnet must be independently controlled in milliseconds.

Output can be in PWM, I would love to, depending on cost, to have the control output in sine, sawtooth or other configurations, or approximations thereof.

Also, I am thinking at a later date to add feedback control to the electromagnets, so keep this in mind to the design.

I am aware combination of op-amp, diodes, transistors etc are required, I have a very basic understanding of micro-electronics, so I need this as simply as possible, including the code to control all this.

Any takers?

Philldapill

Are you looking for Hardware or code? If it's hardware you're after, I've designed a few medium power PWM drivers. If it's code.... a Propeller would make short work of what you want to do...

BTW, we need some more info. What's your objective?

Phil Pilgrim (PhiPi)

Oh nuts! I just answered your duplicate post in another forum. (Duplicate posts are not permitted, BTW.) After I answer here, I will delete my post there so you can do the same. Here's what I wrote:

Just to clarify what I think your post means:

1. The DC resistance of each magnet is about 1 ohm.

2. The current you speak of is average current.

3. By "millisecond" control you're referring to is the PWM you mention later.

Correct?

I'm not sure why you'd want anything more exotic than squarewave PWM, though, since the coil's inductance will average it out anyway. Can you explain?

-Phil

Chris Savage

Duplicate post removed.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Engineering

ionhunter

Thanks for the input guys,and removing the double post Chris.

Philldapill, yes I would be very interested in any medium power PWM drivers you have designed, as for the Propeller chip, would it be needed to control 2 electromagnets independently?. s

Phil Pilgrim: Thanks for your input. To clarify I need to control two electromagnets independently of each other for now. Later on I will tackle multiple units later.

For each electromagnet I need the following: Constant voltage of about 25-30V, variable current from about 0-26 amps.
I need to switch on/off the electromagnets from about 1/5 seconds on to constantly on/off.
Each electromagnet controlled independently.

So again, 2 variables to control: Time on/off via PWC, and strength of magnetic field by varying current from 0-26 amps in steps of one unit, precice decimal control is not needed.

What am I building? Well, at least this phase is an electromagnet controlled braking system for a much more adventurous project, trust me its gonna be awesome...

Phil Pilgrim (PhiPi)

You could use the MoBoStamp-pe to provide the PWM output. It's a BASIC Stamp 2pe with two onboard coprocessors that each provide two independent 37.5KHz 8-bit PWM channels. These can be used to drive the gate of a MOSFET which, in turn, can drive your electromagnets.

-Phil

ionhunter

Phil Pilgrim:

Right now I have the BS2, as for the components you mentioned MOSFETs, can you explain you reasoning?

I was thinking of interfacing NPN Darlington array to BS2, then to this a couple of NPN high power transistors with protective diode across the electromagnet. Controlling the Base current of the Darlington to control timing and field strength. The only problem I think I have is to have a feedback loop? Possibly an op-amp?? How would I manage this?

What do you think, is this workable, any input would be appreciated.

BTW, here are the specs for the electromagnets:
Each electromagnet measures 46mm in diameter and 36mm high with two 12mm protruding 4mm x .70 threaded connection terminals. Polymer body encapsulates the laminated steel wound core and includes four steel sleeved 3mm mounting holes in a rectangular configuration on 19mm by 33mm centers. The tripolar ground surface faces are additionally protected by an bonded circumferential steel band 1.75mm thick by 6.5mm deep. The cross-sectional view photo below exhibits the hermetically sealed core windings which are comprised of 132 turns of 24ga solid copper wire having a DC resistance of roughly 1.5 ohms. Zero air gap flat plate loading of one of these electromagnets will generate approximately 14lb of holding force per volt of applied EMF. For example, supplying 6VDC to one of these will exert 84 lb of lifting force. Perfect for student labs or experimentation. Higher duty cycle usage may require a drop resistor.

Phil Pilgrim (PhiPi)

Without a good, continuous PWM source (the BS2 uses a duty mode output for its "PWM"), the MOSFET idea is not so hot. You can use PWMOUT to charge a cap to a predictable voltage, though, which can then control your cascade of bipolar transistors (which will need massive heatsinks, BTW, since they're operating in their linear region). However, I would recommend using an op amp to regulate the current, for two reasons:

1. Its input impedance is quite high, so it will not discharge the holding cap quite as quickly as a bipolar transistor would, and
2. You will not be relying on the (unreliable) gain of your transistor cascade for current regulation.

Due to the wasted energy in the form of heat that you will be experiencing, a true switchmode PWM solution with a MOSFET would be much better, but the BS2 just doesn't have the capability to drive such a system without a mess of external components.

-Phil

Addendum: Oops! I nearly forgot about the PWMPal. This would give you the continuous PWM output needed to drive a MOSFET.

Post Edited (Phil Pilgrim (PhiPi)) : 4/29/2009 10:21:44 PM GMT

ionhunter

Ah, I see your reasoning.

So basically, I am in between three camps here: BS2 module seated on a PWMPal , which in turn drives up to 4 MOSFETs via PWM, which in turn drives either 4 electromagnets within one current range (0-20)OR 2 electromagnets with two current ranges (0-10 and 11-20)?

OR

BS2 module s , which in turn drives op-amp(how many) via duty cycle, which in turn drives either 4 electromagnets within one current range (0-20)OR 2 electromagnets with two current ranges (0-10 and 11-20), or different current ranges??

OR

Just having a range of say 10 different resistors in parallel with incremental range of 2 amps each total, and just having a 10 relay switches control it in 2 amp increments. This would then create other problems, in that ONE electromagnet would need 10 inputs, how viable is this?

At the end I am looking for the easiest method, even to "cheat" by using straight components with minimal programming.

Keep these great responses going!