A better H bridge...

[audioguru]

That guy is completely wrong about PNP transistors.

 

[ikalogic]

Okay here is what i came up with ..

**broken link removed**

now, the 2 transistors i added Q9 and Q10 are to perform a logic task.

Ron H, you may notice that all transistors are 'active turn off'

i know the 2 extra transistors may cause some micro seconds of delay..

Now you can also notice i didn't fixe the value of the base resistors of the TIPs (Rt) .. i am leaving it open for suggestions.

I still beleive that injecting smothing like 40 mA in the base of the TIPs through 1/5 watt 220 Ohm resistors would be better (Ron H, i just viewed the data sheet. and if my collector current jumps to 5A, i'll need 20 mA of base current to maintain the collector current.., if i multiply this by 2, as a factor of safety i'll get a fare 40 mA.. what do you think?)

 

[Roff]

Okay here is what i came up with ..

**broken link removed**

now, the 2 transistors i added Q9 and Q10 are to perform a logic task.

You need current limiting resistors in series with Q9 and Q10 collectors.

Ron H, you may notice that all transistors are 'active turn off'

i know the 2 extra transistors may cause some micro seconds of delay...

Probably a few tens to hundreds of nanoseconds, depending on what type of transistors you use, how much excess base current you drive them with, and the impedance of the turn-off circuit.

Now you can also notice i didn't fixe the value of the base resistors of the TIPs (Rt) .. i am leaving it open for suggestions.

I still beleive that injecting smothing like 40 mA in the base of the TIPs through 1/5 watt 220 Ohm resistors would be better (Ron H, i just viewed the data sheet. and if my collector current jumps to 5A, i'll need 20 mA of base current to maintain the collector current.., if i multiply this by 2, as a factor of safety i'll get a fare 40 mA.. what do you think?)

Do what you want. I think the safety factor is built in to the datasheet. Excessive base current means additional turn-off time.

I think you'll have to make substantial changes to build in dead time. An ON transistor needs to begin turning off a microsecond or so before the other transistor on the same side begins to turn on.

 

[ikalogic]

Again here are the corrections.. i added the series resistors (I just forgot them) thx again you saved me from lot of trouble and smoke!

**broken link removed**

About the Base resistors for the TIPs (Rt) well i'll go for a 470 ohm giving a 2 mA current.. i think its fare to start testing with this value.

For the Dead time, its noted, i'll take this in account when driving the H-bridge.. unless there is another solution.. (i was thinking about using 2 op-amps to manage this... am i on the right path?)

Reguard less the dead time, are all transistor well saturated in this last circuit? (actually this was my biggest problem from the begining)

thx

 

[audioguru]

A 74LS00 logic high output can supply a minimum output voltage of 2.7V with a 0.4mA load. The 470 ohm resistors in series with some of the transistors are a (2.7V- 0.7V)/470= 4.3mA load which is 10 times higher. I don't think the gate is suitable for driving NPN transistors. An ordinary Cmos CD4011 with a 12V supply would have a much higher output current.

 

[Roff]

A 74LS00 logic high output can supply a minimum output voltage of 2.7V with a 0.4mA load. The 470 ohm resistors in series with some of the transistors are a (2.7V- 0.7V)/470= 4.3mA load which is 10 times higher. I don't think the gate is suitable for driving NPN transistors. An ordinary Cmos CD4011 with a 12V supply would have a much higher output current.

Yep, and he would only have to level shift at the inputs (assuming the inputs come from a 5V logic system), or maybe not at all.

 

[ikalogic]

ok, now how can i test the performance of the h bridge when its done?

with a high power resistor? , with a motor and stalling the rotor?

is there any sites which does this?

 

[Styx]

ok, now how can i test the performance of the h bridge when its done?

with a high power resistor? , with a motor and stalling the rotor?

is there any sites which does this?

what kind of control do you have? current-loop?

 

[ikalogic]

what kind of control do you have? current-loop?

what do you mean by current loop..?? you mean closed loop?

 

[ikalogic]

A 74LS00 logic high output can supply a minimum output voltage of 2.7V with a 0.4mA load. The 470 ohm resistors in series with some of the transistors are a (2.7V- 0.7V)/470= 4.3mA load which is 10 times higher. I don't think the gate is suitable for driving NPN transistors. An ordinary Cmos CD4011 with a 12V supply would have a much higher output current.

I would use 4011, but i need to drive it from TTL logic..

the 74HC00 seems to me like a real problem solver, with its +/- 25mAcurrent sink/source at 5v... right?

 

[audioguru]

the 74HC00 seems to me like a real problem solver, with its +/- 25mAcurrent sink/source at 5v... right?

Yes, but the max supply current is only 50ma so you can't have all 4 gates with 25mA each simultaneously.

Also, the 74HC00 has Cmos-level inputs and needs a pullup resistor to work from TTL signals. The 74HCT00 has TTL-level inputs so works fine with TTL-level signals.

 

[ikalogic]

Yes, but the max supply current is only 50ma so you can't have all 4 gates with 25mA each simultaneously.

since only half of the h-bridge is on at once, i'll need only 2 channels, then, the 74hc00 will power the 2n2222 and the 2n2907 transistors, with a base current of about 8 mA each.. so its really ok i guess..

Also, the 74HC00 has Cmos-level inputs and needs a pullup resistor to work from TTL signals. The 74HCT00 has TTL-level inputs so works fine with TTL-level signals.

I think putting 1K pull up resisotrs to the inputs of my h-bridge will make is usable with ttl or cmos devices.. it won't harm nor take too much place..

 

[Roff]

You don't need 8mA of base current. Accepted practice for a saturated transistor is to set Ib=Ic/10.

 

[ikalogic]

You don't need 8mA of base current. Accepted practice for a saturated transistor is to set Ib=Ic/10.

even better so i'll only need 2 or 3 mA in my case...

 

[Styx]

what do you mean by current loop..?? you mean closed loop?

yes, if you want to do stall-tests you will need something to control the current

 

[ikalogic]

yes, if you want to do stall-tests you will need something to control the current

okay! that's seems interresting, i heared about it before, but allways though doing it with a shaft encoder was easier.. do you know a site or have any paper that talks about this technique?

EDIT: ooops after re-reading what you wrote.. i think i got it wrong.. well i absolutely want more information about this.. for what purpose de we need to control the current while making a stall test? shouldn't the current rises? shouldn't we just measure it? what do you mean by control in that case?

 

[Nigel Goodwin]

EDIT: ooops after re-reading what you wrote.. i think i got it wrong.. well i absolutely want more information about this.. for what purpose de we need to control the current while making a stall test? shouldn't the current rises? shouldn't we just measure it? what do you mean by control in that case?

Not so much 'control' as 'limit' the stall current will be MANY times the normal running current - without any limit it may well destroy things.

 

[Hero999]

To get the stall current just measure the resistance of the motor with a multimeter and use ohms law to calculate the current.

 

[Nigel Goodwin]

To get the stall current just measure the resistance of the motor with a multimeter and use ohms law to calculate the current.

Except multimeters don't read anywhere near that low! - also the stall current is dependent on the rest of the circuit, the wiring, the switching and control circuitry, plus the supply itself as well.

 

[Styx]

Not so much 'control' as 'limit' the stall current will be MANY times the normal running current - without any limit it may well destroy things.

To an extent, but a proporly build current-controller will control the current through the stalled machine from say... 0.5A to .. any limit you want
You control the level. If that level is the same as yr current-limit then they are the same