Will this bipolar design work???

[666]

Hey all...

I´m building a direct drive device for a machine using a powerful stepper

Bipolar, rated 8.4 amps/Phase, Torque 9.2NM fullstep bipolar drive.

I allready designed/engineered a bunch of unipolar drivers (e.g. "Alexander´s 3 axis pic" at open source boards/Pminmo.com), but all for unipolar steppers.

Having tried these om the motor (8 wire) i just discovered that the fet´s blew due to inductive kickback from the coils. Tried to make a workaround, but unsuccessfull.

Can´t use them for bipolar steppers, so i have to buils something much more powerful... A dual H-bridge design!

I made a diagram attaches to this post (PDF).

I have no test/build option at the moment, being tied to bed after a back operation last week.

But i made this diagram.

The step/dir X-lator itself is nothing new, but the bridge-design is made using only N-channel fet´s i have in stock allready.

So my questions are regarding the dual h-bridge design.

Will the fet´s (FZ´s) turn on as expected? I use a 9 volt supply with optocupler circuitry to get from 5V ttl level to 9V cmos inputs.

Using FZ and NOT LZ the required gate voltage (RdsON) must be quite higher, but i consider the demands met regarding the datasheet info.

Note: The Cmos S/D circuit will be replaced with my own PIC design if the H-bridge circuit is delivering as expected.

Sorry for the poor schematic... Quickly clamped together from my bed!

Regards from Denmark!

666

[666]

I guess iĺl need some "High side voltage" compensation??? In order to turn the high side on i should need a pulse higher than Vcc?

[mmcp42]

Originally Posted by 666 I guess iĺl need some "High side voltage" compensation??? In order to turn the high side on i should need a pulse higher than Vcc?

indeed - what some call the "charge pump"
alternative is to use opposite polarity FETs, like MTP2955, for the high side

[666]

I allready made H-bridges with P-channel FETś at the high side.

They work fine with smaller motors, , but not with motors as big as the ones i have in stock.

Problem is that RdsON is too high. Typical 0.2-0.5 ohms.

The motors will fraw 8 amps+ / phase, causing the amount of watts deflected as heat in the P-channel FETś to produce heat unless cooling is applied.

The design i use doesn't allow heatsinks because of lacking space.

Other drawback is the linearity of the P-channel FET's at higher rpm's. The motor will get a "rugged" running at high speeds.

So i guess i'll have to make a second psu to provide the rised VCC voltage to turn the N-Channel FET's on.

Or maybe as you mention the "charge pump" design using a diode/cap. clamper fo the job. Problem is that the clamping circuit only will work as long as the cap is charged.

In a braking situation i guess the motor will react pretty funny if only the low side FET is on.

Thanks for your reply... 666/alexander

[george4657]

Take a look at this thread
http://www.cnczone.com/forums/open_s...-tinycpld.html
Even if you don't use a cpld the bridge circuit is about as simple as you can get.
Two things you have to watch out for is shoot through caused by both upper and lower fets being on at the same time due to on off times being different.
The other thing is your circuit has no current control. There is nothing to stop the current from going above spec and burning out fets and/or motor. Most use PWM and comparators to control the current.

George

[666]

Hi George, and thanks for your reply...

The current limitation/chopping wont be a problem.

Allready made a working chopper circuit.

The linked thread looks interesting.

Regards 666/Alexander