go_away

Author Topic: H bridge {loading problem } N and P fet bridge [schematic included ]  (Read 1931 times)

0 Members and 1 Guest are viewing this topic.

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
hello guys im new here
i dont really know if its the right place for the post

as i have been working on h bridge for awhile its my 1st design as you see below

using it to control 14amp 24 volt (250 watt) motor

controlling using Arduino (500hz frequency )

(1-0) move in 1 direction
(0-1) other direction
(1-1) Break

the upper side is P channel IRF9540 or 9540N i used both
http://www.irf.com/product-info/data...a/irf9540n.pdf

the lower side is N channel IRF540 or IRF540N also
http://www.irf.com/product-info/data...ta/irf540n.pdf

for level shifting ( 5 to 12v)
im using BC547 NPN

http://www.futurlec.com/Datasheet/Transistor/BC547.pdf

to keep the voltage (VGS and VSG ) for P and N transistors
im using 12 v zener

here they name it only by voltage and wattages
im currently using 1 watt zeners

the max available is 5 watt

R zener is 1k in case its not obvious enough !

the driver works fine without loading
but if i add them my sort of cart project
15k load ( current consumed 10-12 amps )

the motors dont seem to move the transistor HEAT AND HEAT then they blow !

can you guide me whats the problem

sorry for the long post ,just hoping for help asap !

ah the schematic 
http://desmond.imageshack.us/Himg690/scaled.php?server=690&filename=hbridge.png&res=landing


i'd like to add
i tried to decrease the load and see whats gonna happen

the robot moved

but too slow the first voltage was about 3 v or something and keeps increasing as the motor moving within a noticable speed

the current drawn is around 5 ampere

any thoughts ? :S
« Last Edit: July 08, 2012, 04:47:40 PM by thekaiser11234 »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Hi,

[...]
ah the schematic 
*url=http://desmond.imageshack.us/Himg690...ng&res=landing*
[...]
any thoughts ? :S

Thought... Always check the links you post... At least if you want other people to follow them.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Hi,

The forum software just ate yet another long post, so here's a quick recap...

Need readable schematic w. description of which inputs goes from where and at what voltages.

No wonder your MOSFETs heat, you're starving their gates.

I don't quite see what you're trying to do with the superfluous zeners, but add gate pull downs to the low side MOSFETs.

You need lots of current to switch efficiently, I just made a controller with switching times of <50ns on and <125ns off, for a 24V/250W motor - it only gets around 800mA of gate current, to keep the price down, but I'm quite satisfied with the resulting switching times.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
hello sir

inputs can be pwm or dc 5v level
coming from Arduino as i mentioned in the topic

the zeners are for getting  the vgs or vsg for p and n fets 12 and lower not greater

so you suggest me lowering the resistance connecting collector of bc547 and gate of PFET ??


about the post your talking about i really cant  find it :S :S


so if you think the design is ok but resistance values aint how can i determine their values ?



Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Hi,

the zeners are for getting  the vgs or vsg for p and n fets 12 and lower not greater
Yes, but they're not needed, as long as the resistors are chosen so that VGS never gets above 20V.


so you suggest me lowering the resistance connecting collector of bc547 and gate of PFET ??
Yes, you need lower resistance, but that circuit will never work - too many flaws, it's a wonder it didn't just went up in smoke at first power up ;)


about the post your talking about i really cant  find it :S :S
I'm not sure what you mean??
I didn't mention any other post (in my last post).


so if you think the design is ok but resistance values aint how can i determine their values ?
Your design is NOT OK (I wouldn't even call it a design, as is). From the schematic you posted, I guessed the resistors were 1k and 10k exclusively,so I made you a schematic (attached) that is somewhat similar, but with correct resistance values and a proper use of the bipolar transistors (note that two of them are now PNP-devices).

I also added protection diodes (no, the inherent diode stretch in a MOSFET won't do for a free wheeling diode) and removed the 1nF cap over the terminals (it should be 100 times larger that and moved to directly over the motor terminals, together with two grounded to the casing (unless one of the motors supply wires are already connected to the motor casing).
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
Thank you very much sir i will at your design and try to learn it thank you again :)

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
but 1 thing

isnot 20v for vgs or vsg is too critical (i  mean is too high ?  as its the maximum value )

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
sir ,
im asking this design is only allowing rotation in 2 direction

and no braking available as shoot through may happen

what about the getting upper side ground of bc547 changed into inputs
??

or im missing something obvious?

thank you sir
« Last Edit: July 09, 2012, 07:02:56 PM by thekaiser11234 »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
but 1 thing

isnot 20v for vgs or vsg is too critical (i  mean is too high ?  as its the maximum value )
20V is the max. limit and to repeat myself "as long as the resistors are chosen so that VGS never gets above 20V." (which it never will in the circuit I drew - run the numbers past a calculator for the exercise).
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Hi,

im asking this design is only allowing rotation in 2 direction
Yes, just like the one you posted and which I patched up did (and all other H-bridges, but that is more a case of motors only being able to rotate in 2 directions).


and no braking available as shoot through may happen
You didn't post a better schematic (or even an explanaton), so I based the one I did on what I imagined you wanted yours to do - your schematic wasn't able to brake either (or move at all).


what about the getting upper side ground of bc547 changed into inputs
??

or im missing something obvious?
The emitter of an NPN transistors goes to ground (a simplification to avoid getting you on a wrong track).
The way you tried to use them in your schematic is wrong (and one of the reasons your circuit didn't work). And I based the new one on correcting the circuit.
If you wanted it to do something else, you should have said so - we're not mind readers and can only go on what info you provide.

If you provide a (correct) truth table that you want it to follow, like eg.
Code: [Select]
In1  In2  M1  M2
 0    0    ?   ?
 1    0    ?   ?
 0    1    ?   ?
 1    1    ?   ?
(where you fill out the question marks, of course), I'll make it follow it.

If you want to have max. control over the bridge, you need 4 outputs, one for each MOSFET and then you can just break the connections between hi and low side and run them independently (remember to add a pull down resistor to the low side then).


As is, it will go in the direction where you provide PWM (or just a "high") and coast when both inputs are at "low". It can use power braking (a timed pulse of reverse direction current) and can provide a very firm control by pulsing the inputs in sequence (remember a dead band of around 2µs, to make sure you don't short your battery), where 50% CW and 50% CCW will be a dead stop, but with full current draw. This is the running method when you want the tightest control over a DC-motor.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
hello again sir ,

(1-0) move in 1 direction ( 1,0)
(0-1) other direction         (0,1)
(1-1) Break                       for me it was low side braking but does it make a difference if its high side or low side braking ?

(0 -0 )would  be all transistors off

thats what i wanted to happen and posted as first said

though i guess the long post was a confusion sorry im new to this :)


also i wana ask


Why is connecting an input to the emitter of BC547 is wrong can you please clarify this ? :)

Thank you again sir

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Hi,

(1-0) move in 1 direction ( 1,0)
(0-1) other direction         (0,1)
(1-1) Break                       for me it was low side braking but does it make a difference if its high side or low side braking ?

(0 -0 )would  be all transistors off
Functionality-vise It doesn't matter whether you shunt the motor high or low, but since N-channel devices have lower RDS-ON in a "complimentary" pair, I prefer shunting low.

The best way to accomplish what you want is to use 4 I/O lines, as you can then control the dead band freely and easily in software, whereas it would take a load of extra hardware to make it shoot-through proof with the best switching times, while in software, you can tailor the dead band to the specific MOSFETs you use.

I have attached a schematic of doing it that way.
R3, R4, R17 and R20 can be discarded if the inputs are driven (high or low) at all times, while the +24V is applied - I wouldn't though.


Why is connecting an input to the emitter of BC547 is wrong can you please clarify this ? :)
A forum post is not the place for a transistor tutorial, so you'll have to Google for either a full tutorial or perhaps just one on transistor switching.
I promise you that the time will be well spent.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline thekaiser11234Topic starter

  • Jr. Member
  • **
  • Posts: 8
  • Helpful? 0
i mean ...

its like connecting BC547 emitter to ground and not connecting it to ground

or is there a difference ?

 


Get Your Ad Here

data_list