dc motor control

Matt Battle

Hey all-

I found out the dc motors that drive the roomba needs a current spike over 3A for them to turn over. I was going to use two SN754410NE to control the motors but that is only 2A. Does anybody know any good driver ICs that would supply the 3A+ needed for the motors? Thanks guys and girls.

-Matt

Post Edited (Matt Battle) : 2/17/2006 6:00:07 PM GMT

Matt Battle

Update: After some more research I found out that in cleaning mode, the Roomba wheel motors draw 5.67 A of current and possess
13 V of voltage. That might change somethings.

Jon Williams

Have you seen out new HB-25 motor controller? [noparse];)[/noparse]

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax

Matt Battle

Jon-

Yea, I have looked at the hb-25 but I think it would be an over kill for what I want to do. I am trying to keep it simple and save some money. I don't want to buy a 50 dollar piece of eqiupment and damage it or break it haha. Some simplr h-bridges is all I am looking for

-Matt

PJAllen

matt battle saith...
·3A+ needed... wheel motors draw 5.67 A... the hb-25... would be an over kill for what I want to do... keep it simple and save some money.·Some simplr h-bridges is all I am looking for

· "Simple h-bridges" & good for >6A, he says.· Tis mind-boggling.

· Anyway, since this Roomba presumably moves about, or at least once did --·then why can't you actuate its motors, control-wise, from where it does/did?· It must have already what you're re-trying to re-do.· Yes?

· So far as saving money goes -- wrong hobby.

Matt Battle

PJ Allen said...

matt battle saith...
·3A+ needed... wheel motors draw 5.67 A... the hb-25... would be an over kill for what I want to do... keep it simple and save some money.·Some simplr h-bridges is all I am looking for

· "Simple h-bridges" & good for >6A, he says.· Tis mind-boggling.

· Anyway, since this Roomba presumably moves about, or at least once did --·then why can't you actuate its motors, control-wise, from where it does/did?· It must have already what you're re-trying to re-do.· Yes?

· So far as saving money goes -- wrong hobby.

PJ Allen-

I am kind of lost to what you are suggesting.· Are you saying·to take what I need off the PCB from the Roomba?· That is my last option.· Something like this is what I am looking for:http://www.solutions-cubed.com/solutions%20cubed/SB1_2005.htm.· Just wished I could find out what the chip is that they are using.· Can always go with a LMD18200 but that might not be enough.

-Matt

John R.

Why is using what was there the last option?

PJ isn't suggesting using the "smarts", just suggesting that there was uProcessor running the thing. It sent a signal to "something" (likely an H bridge or "smart motor controller". Why not use this "something" as opposed to buying anything? Basically, cut the trace from the former uProcessor, and connnect your output pin(s) to the "working" end of the cut traces.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
John R.

8 + 8 = 10

Matt Battle

John R. said...
Why is using what was there the last option?

PJ isn't suggesting using the "smarts", just suggesting that there was uProcessor running the thing. It sent a signal to "something" (likely an H bridge or "smart motor controller". Why not use this "something" as opposed to buying anything? Basically, cut the trace from the former uProcessor, and connnect your output pin(s) to the "working" end of the cut traces.

John R.-

I really don't want to do that because I wont learn why it does what it does.· I want to build almost all of the circuits from scratch and understand what each component does and why I need it.· I think what I should do first is study the PCB that the Roomba has and see where they go and what they do.· That should give me some what of an understanding and a ground to start from.

-Matt

bobledoux

My robotics experiences suggest that these projects have enough challenges. You may be creating an additional challenge (the H-driver) and cost, by trying to recreate what is already there. Effective H drivers are not always a trivial design project. This is especially true when pulse width motor speed control is added to the mix.

Why not study the design already in the unit and figure the best way to make use of it? The original engineers must have gone through a number motor controller variations before settling on the current controller. Do you really want to reinvent this wheel?

LoopyByteloose

I guess you are more of a hardware guy, than a software guy [noparse][[/noparse]me too.]

After two years, I have a room full of parts but very little really built.· It is like sailing, some people build boats and never sail; others sail and never build.

There are MOSfets that can build an H-bridge, but one thing that I have experienced is my own projects become larger than what I could put together with already fabricated technology.· Parallax's new H-bridge is an investment, but considering that it will handle so much, so well - you may find it migrating to big and better robots.

Still, I wonder why destroy a Roomba?·

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Matt Battle

bobledoux said...
My robotics experiences suggest that these projects
have enough challenges. ·You may be creating an additional challenge
(the H-driver) and cost, by trying to recreate what is already there.
Effective H drivers are not always a trivial design project. ·This is
especially true when pulse width motor speed control is added to the
mix.

Why not study the design already in the unit and figure the best way to
make use of it? ·The original engineers must have gone through a number
motor controller variations before settling on the current controller.
Do you really want to reinvent this wheel?

bobledoux-

I wish I could study the design and figure out how to make use of it but first I need to figure out what componets they used in the h-bridge.· Then I would·rebuild it on a breadboard.·I don't want to reinvent the wheel just understand it.

-Matt

Kramer said...
I guess you are more of a hardware guy, than a software
guy [noparse][[/noparse]me too.]

After two years, I have a room full of parts but very little really
built. ·It is like sailing, some people build boats and never sail;
others sail and never build.

There are MOSfets that can build an H-bridge, but one thing that I have
experienced is my own projects become larger than what I could put
together with already fabricated technology. ·Parallax's new H-bridge
is an investment, but considering that it will handle so much, so well
- you may find it migrating to big and better robots.

Still, I wonder why destroy a Roomba?

Kramer-

I was told that the roomba wasn't in·working condition·so I figured that I should take it apart and understand how it works.· The new HB-25 might be more then I need.· It could be useful down the road though.

-Matt



Post Edited (Matt Battle) : 2/19/2006 3:31:09 PM GMT

Matt Battle

I think I figure out which H-bridge the Roomba uses and I want to recreate it on a breadboard but my problem is trying to figure out what transistors it is using. From what I can read on them one says D882 and the other one says B772? I was thinking that I could use TIP41 or TIP31 but I don't understand what Vcoe and Hfe means if someone can help me on that. Thanks

PJAllen

V_ceo is, basically, the maximum voltage allowable from collector to emitter.

h_fe, also known as "beta", is, basically, the current gain factor of the transistor.· So, if it has a "beta" of 100 and there is 1 mA of base current (I_B), then you have the ability to draw 100 mA I_C or I_E (rough approximation, I_C = I_E.)

That makes it all a little simpler than it is --·but they write textbooks about this sort of thing.

Matt Battle

PJ Allen said...
V_ceo is, basically, the maximum voltage allowable from collector to emitter.

h_fe, also known as "beta", is, basically, the current gain factor of the transistor. So, if it has a "beta" of 100 and there is 1 mA of base current (I_B), then you have the ability to draw 100 mA I_C or I_E (rough approximation, I_C = I_E.)

That makes it all a little simpler than it is -- but they write textbooks about this sort of thing.

PJ Allen-

With that in mind, which one would be the best to get? TIP31 ot TIP41? I was going to get the TIP41 just to be on the safe side by I am thinking that it would draw more power right? I guess I should get a electronic handbook.

-Matt

PJAllen

· Attached is a schematic for a very buildable H-bridge.· It uses transistors that you should able to obtain from any of the usual suspects.
· While you're at it, get one of those (sucky) little Mabuchi motors as your test subject.
· If you can't get this going, then you'll need to... take a different tack.

Beau Schwabe

Pj Allen,

I might be concerned if the inputs to your schematic became Hi-Z, then ALL of the transistors would
have a tendency to turn on.

Attached is a schematic that should work.

Also, refer to the link below for choosing the proper current limiting resistor values, based on your
motor demands.

http://forums.parallax.com/attachment.php?attachmentid=37701

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Matt Battle

This is kind of·the H-bridge that·the Roomba uses.· I am going to build it and see how it works.· Then my next step is to connect a·BS2 to it and run some code.

Post Edited (Matt Battle) : 2/21/2006 6:10:58 AM GMT

LoopyByteloose

If the Roomba was broken, I truly hope you have tested the motors first. [noparse][[/noparse]it sounds like you have.]

What Beau is trying to point out that in the 'All logic on' condition some H-bridges become smoke and sparks.
You have to be extra careful to not write software that has such a bug in it. You simply create a 'short circuit'.

Other H-bridges are known at 'smokeless' designs due to extra protective logic, internal fusing, and so on.

Personally, I think that understanding the contol of the H-bridge, the basic kinds of stepper motors drivers, the PWM concepts of servos, and the 3-phase Ac drivers for brushless motors in model airplanes are extremely good for a full understanding of motor control.

If Parallax ever adds a complete motor control course, it will likely be popular. But I think that they would have to use the SX-28 in order to get 3-phase AC.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Matt Battle

Kramer said...
If the Roomba was broken, I truly hope you have tested the motors first. [noparse][[/noparse]it sounds like you have.]

What Beau is trying to point out that in the 'All logic on' condition some H-bridges become smoke and sparks.
You have to be extra careful to not write software that has such a bug in it. You simply create a 'short circuit'.

Other H-bridges are known at 'smokeless' designs due to extra protective logic, internal fusing, and so on.

Personally, I think that understanding the contol of the H-bridge, the basic kinds of stepper motors drivers, the PWM concepts of servos, and the 3-phase Ac drivers for brushless motors in model airplanes are extremely good for a full understanding of motor control.

If Parallax ever adds a complete motor control course, it will likely be popular. But I think that they would have to use the SX-28 in order to get 3-phase AC.

Kramer-

I thought when you have an "All logic on" condition, it will cause the motor to "brake".· I thought that the·diodes protect·the circuit from shorting out?· I might be wrong on both parts.· I need to know if I am good to go on my h-bridge choice before I buy the parts for it or do I need to rethink my choice?

-Matt

Beau Schwabe

Matt Battle said...
I thought when you have an "All logic on" condition, it will cause the motor to "brake".· I thought that the·diodes protect·the circuit from shorting out?

There are certainly some·H-Bridge controllers that will "brake" in an "All logic on"·condition, but as a blanket statement for all H-Bridge controllers
this notion is incorrect.· Some H-Bridges will self destruct in an "All logic on"·condition.
·
The diodes are to protect the transistors from large voltage spikes generated by a rapid·collapse of the magnetic field in the coils within the motor
which happens when· you pulse the motor with PWM or·simply turn the motor on or off.
·

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Matt Battle

Beau Schwabe (Parallax) said...

Matt Battle said...
I thought when you have an "All logic on" condition, it will cause the motor to "brake".· I thought that the·diodes protect·the circuit from shorting out?

There are certainly some·H-Bridge controllers that will "brake" in an "All logic on"·condition, but as a blanket statement for all H-Bridge controllers
this notion is incorrect.· Some H-Bridges will self destruct in an "All logic on"·condition.
·
The diodes are to protect the transistors from large voltage spikes generated by a rapid·collapse of the magnetic field in the coils within the motor
which happens when· you pulse the motor with PWM or·simply turn the motor on or off.
·

Beau-

Wow, collapse of the magnetic field sounds like something from star trek.· I am just lost to how that works but now that gives me something to read about tonight.· Between my H-bridge and yours what is the difference in where you have the diodes placed? Am I good to go on buying the parts for my h-bridge?· It seems I can build both of our h-bridges with the same parts.· Which is good.

-Matt

Beau Schwabe

Matt,

"Between my H-bridge and yours what is the difference in where you have the diodes placed?"

There is no difference, I just discovered a while ago that an off the shelf bridge rectifier can be
connected in the proper orientation so that it can be used·as back EMF protection.



"Am I good to go on buying the parts for my h-bridge?· It seems I can build both of our h-bridges
with the same parts."

Look at the design a little closer.· I am using PNP's as well as NPN's.· You have all NPN's with uneven
loading on the inputs between the top (Q1 and Q2) and the bottom ( Q3 and Q4 ) transistors.

EXAMPLE:
The bottom transistors(Q3 and Q4) "see" this ...

I/O >----/\/\----->|------GND
 
 
 
while the top transistors (Q1 and Q2) "see" this ...

I/O >----/\/\----->|----CCCC---->|----GND

 
 
 
CCCC = Motor Coil (Inductor)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Matt Battle

So something like this then?

pjv

Hi Matt;

Take a closer look at Beau's circuit; he's got it right.....(been using that for years). If you're tricky and reliable with your timing you could also dispense with the protection diodes, but I do not reccomned that for anyone but the most experienced.

The one you have drawn here will smoke the first time you power it up while the two inputs are not clamped to either rail.

Cheers,

Peter (pjv)

Beau Schwabe

Matt,



Sort of.... PNP's require a negative signal of .6V across the B-E junction, where as NPN's require a positive signal

of .6V across the B-E junction. With your latest configuration, you will either have a "brake" condition or a "smoke"

condition. (See Fig1.jpg)



A derivative of your schematic would be to use what was posted by PJ Allen. This circuit will work as long as you

do not let the inputs float. (See Fig2.jpg)



Another derivative is the circuit that I provided. The requirement here is that the inputs should never both go HIGH.

http://forums.parallax.com/attachment.php?attachmentid=40549

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 2/22/2006 1:12:57 AM GMT

pjv

Hi Beau;

I would beg to differ.

In case number 3, with both inputs open or floating the bottom (NPN) transistors are both OFF, so there is no place for the upper (PNP) transistors' base current to flow, and hence they will be OFF.

Cheers,

Peter (pjv)

Beau Schwabe

Ok....Agreed.

I will make the changes... It's been awhile since I used that circuit myself.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Matt Battle

Beau-

Wow, and that H-bridge is in a book. I wonder how many people have smoked their circuit? Well, I need to use some TIP41 becuase of the Amps they can handle but I don't know what its counterpart is????

Update:· I found out TIP41 counterpart is TIP42.

-Matt

Post Edited (Matt Battle) : 2/22/2006 10:32:00 AM GMT

LoopyByteloose

Wow! Only an IC layout engineer would notice a Bridge rectifier could be inserted into an H-bridge.

After you get all that layout done, you have really only controlled direction and on/off.
One of the finer points of H-bridge is to have the speed controlled with PWM too.

The basic concept remains the H that transfers direction, but there are a lot of refinements.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Matt Battle

Kramer said...
Wow! Only an IC layout engineer would notice a Bridge rectifier could be inserted into an H-bridge.

After you get all that layout done, you have really only controlled direction and on/off.
One of the finer points of H-bridge is to have the speed controlled with PWM too.

The basic concept remains the H that transfers direction, but there are a lot of refinements.

Kramer-

Yea, I am just picking at the tip of the iceberg with H-bridges. There is alot more to that. Once I get this h-bridges working I want to start working on controlling the speed and implementing touch sensors (bumper).

·I have few more question before I place my order today. I notice PJ Allen has 750 ohms for the resistors? How did he come up with that number? Do I need heat sinks for the TIP41 and 42? What is the tool used to insert wire into connectors? How do I tell the exact size of a connector? That is it for now. I know once I get my parts I will have more questions

-Matt

Matt Battle

Hi all-

I have been playing around with a simple h-bridge for about 2 days and been trying to figure out how to make it better then I thought about using some IC h-bridges instead of the many different components in my h-bridge which include 4 transistors, 4 resistors, and 4 diodes. I would like to hear your guys pros and cons between the two.

-Matt