NerdKits - electronics education for a digital generation

You are not logged in. [log in]

NEW: Learning electronics? Ask your questions on the new Electronics Questions & Answers site hosted by CircuitLab.

Basic Electronics » Power supply questions

October 31, 2013
by sask55
sask55's Avatar

I am thinking about simplifying a setup that I have by possibly eliminating some components.

I have a motor controller setup that has three Toshiba TB6560AHQ chips which are connected with multiple data connections to three ATmega 168 micros. Apparently these Toshiba motor controller chips are quite sensitive to a number of different possible circumstances that will destroy them. I have inadvertently destroyed two of them by making simple mistakes or assumptions in my test setup. Toshiba makes it very clear on the chip data sheets and usage considerations sheets what steps should be taken to protect the chip. There are numerous reports on related web sites of these chips going up in smoke.

toshiba block diagram

The Toshiba chip block diagram clearly indicates three voltage supplies. First: The power supply to the MCU which is communicating with the controller chip. Second; a separate 5V supply for the controller chips internal logic. Third: a 24V high current supply used to power the stepper motor through the chip. Power-on and power-off sequences are important for these chips; therefore it is necessary to have independent control of the 24Volt and 5 volt voltage sources to these chips.

In a couple of posts on the web, I have seen a number of people strongly suggest that the controller chips and the micro use separated power supplies. I don’t really understand what is meant by that, and it is not explicitly stated in the Toshiba documents. In the Toshiba block diagram it appears to me that all of these supplies are using a common ground but do have individual voltage sources.

One question is. For this application what is a separate power supply? All of power supplied to this setup will originate from the same ac wall outlet source. It is just a question of where does the separation begin. For my test setup I am powering the MCUs and the logic level power to the controller chips with a small wall wart power adapted type supply. I am currently using two separate 7805 voltage regulators. One 7805 supplies the logic level voltage for the Toshiba chips. The power to those chips is controlled through a relay in order to facilitate sequenced start-up and shut down. The other 7805 is for the Micros and supporting hardware it is constantly on when the system is powered up. I am wandering if I actually require two voltage regulators? Could I power both the switched controllers and the micros from one voltage regulator and not have any issues?

I am powering the motor coils through Toshiba chips with a 24 Volt 15 amp max switching power supply. I think the concern is that the motor controllers will be constantly and rapidly changing the current flow to the motor coils. The amperage draw on the 24 Volt supply will change very rapidly from near 0 amps to a possible theoretical level well above 15 amps. It is highly unlikely that the amperage draw will ever actually be above 15 amps as that would require more the four motor phases to be outputting at 100% at the same time. Each of the three motor have two coils (phases) that are limited by the controller chips to 3.5 amps /phase at 100% torque setting and that should never happen. All the switching activity on 24 volt supply may cause voltage spikes, drops and transients in the voltage level. There are a number of capacitors of various values and locations in the power supply circuits that are intended to reduce or eliminate any fluctuations in the voltage level.

I would prefer to use the 24Volt switching supply to handel the DC power to everything. I think that would be defined as a single power supply even with the voltage regulators and capacitors in the five volt circuits, therfore not a good option. I could use that supply on the high voltage side of both of the 7805 voltage regulators? Does anyone have any experience with this kind thing? It will be difficult to test this under actual working load conditions before I build a printed circuit board version of the circuit. It would be difficult to change my basic power supply design after a board is produced.

Will the power supply capacitors and voltage regualtors likely handle the possible voltage fluctuations? In that case I can design with just the 24 volt supply? Or! Should I play it safe and continue to supply the logic level power from a completely different AC to DC source? There is a possible inherent danger to that approach. If for some reason the logic level voltage was lost to the Toshiba chips while the 24 volt supply remained I believe it would destroy all of the controller chips, an unfortunate reality I found out the hard way. The relay on the 24 volt supply to the contollor chips should help to mitigate most of that risk.

Anyway! If anyone has any experience along these lines I would be very interested in your thoughts.

October 31, 2013
by BobaMosfet
BobaMosfet's Avatar

sask55-

Yes, done this lots. Use just one 24VDC source. It can serve everything. Power your LM7805 with it, and power both your MCU and the TB6560AHQ logic-side power from the same LM7805.

Tie all grounds together. Your 24VDC is your chassis ground. Your LM7805 ground should tie to the chassis ground. Your logic grounds to the LM7805, your LM7805 grounds to the 24VDC. Your motor grounds to the 24VDC, and the power-side (non-logic) side of the TB6560AHQ grounds to the 24VDC.

That keeps potentials separate, but relative, and all grounds proper.

You should not have voltage fluctuations. These need to be eliminated with snubbers, diodes, and fuses. The fuse is there to prevent too much current draw, if the motor stalls. If it does, it will pull the whole thing down, because it's dead short. Fuse will prevent it from damaging itself or anything else, or causing a fire.

But if you use a fuse, you also need to learn how to properly choose the right one, because you must deal with in-rush current, overall current draw, and how long it must last before tripping.

BM

October 31, 2013
by sask55
sask55's Avatar

BobaMosfet.

Thanks for the advice.

I was hoping to be able to do it that way. Simplifies the layout reduces the number of components.

These Toshiba chips are current sensing. The voltage level of 24 V produced by the supply is over driving the motors, by a lot. The stepper motor specs are 2.1 Ohms resistance /phase. The max rated current 3.0 Amps/phase. Without the PWM from the controller effectively reducing the current flow to the coils the 24 volt supply would develop 11.4 amps in a coil that was holding. When the motor is holding a position, which is a perfectly legitimate condition for a stepper motor, The coil driving voltage is DC. I=V/R would apply. 11.4 amps would heat up and burn out a motor coil very quickly. I have had these motors stopped (holding a position) with 100% torque level selected on my setup many times, the motor do get a little warm.

The max current flow is pre-set by design, by carefully selecting the value of the current sensing resistors connected to the controller chips. If the controller is working it will never deliver more current then is selected. The advantage of overdriving a motor is to “push “a selected level of current, thru the coil even at higher motor speeds. In my project the max motor speed will require a controller clock rate of about 5kz. The Excitation mode setting to the chip will determine the output form of the controller. The 2-phase excitation that I am using effectivly cause the voltage levels applied to the motor coils to become a AC square wave, the frequency of that AC current determines the motor speed. The added impedance resulting from this AC current applied to the inductive load of the coils is compensated for by the controller chip to maintain the desired current level. The controller chips are constantly adjusting the duty cycle of the PWM to the coils to match the speed and maintain the highest level of torque possible without going above the selected current level.

Toshiba specifically recommends using only fast acting fusses selected to match the total amperage of both coils operating at 100% of the pre-selected max current level.If the current draw of the chip ever exceeds that level even very briefly something is wrong the controller is not limiting the current as designed and hopfully the fuse will blow before something else does.

As I understand it a motor that is not turning because it is incapable of producing sufficient torque to move will still be driven with the same pre-set current level. Without addition input to inform the MCU that the motor is not turning, the MCU will just carry on sending input signals to the controller chips and the controller chip will be applying current to the motor coils as if the motor was turning as expected.

Darryl

October 31, 2013
by sask55
sask55's Avatar

I was incorrect about a holding my steppers at 100% of its maximum rated current level.

I had forgotten that I designed the MCU software to automatically reduce the torque level (current level) to 50% if the period between controller clock pulses exceeded a certain time interval. Ie the system reduces the current in the coils at very low turning speeds and when holding. This is a common technique to reduce heat build up. I don’t know how long the motor coil may last holding in one position with its rated max 3.0 amp current applied to it.

Sorry that was misleading. I did not want anyone to try holding a stepper in a stationary position with 100% of its rated maximum current to the coils. That is not recommended by the manufactures of the motors do to heat build up in the coil.

Darryl

Post a Reply

Please log in to post a reply.

Did you know that talking to the microcontroller over the USB/Serial link is easy under Windows, Linux, and OS X? Learn more...