A better H bridge...

[ikalogic]

I was convinced by advices from hero999, and did some updates to the H-bridge in my projects page, attached is a a schematic i made after some researches on the net..

**broken link removed**

So guys, Do you think this way the transistors will fuly saturate and give better performance?

are the vaules of the resistors ok? (according to my calculations, the 10 K resistors are limiting the base curent to 1mA.. , with an HFE or more than 1000,i think this should work just fine, right?

thx in advance.

 

[Roff]

You have R6 in the wrong place.
Darlingtons will never saturate. They will always have 0.6V - 1V or more across them (depending on collector current), due to the fact that the "input" transistor saturates, providing negative feedback to the "output" transistor.

 

[Hero999]

Was there really a need to create a new thread for this?

Just use MOSFETs anyhow, if you use a higher gate voltage for the high side transistors you can use all n-channel units and save yourself the extra power dissipation.

 

[ikalogic]

sorry it was an error,
i am updating right now with the new schematic

and sorry Hero999, but the other place was not really suitable for this technical stuff..

i want to use TIP transistors.. they are much cheaper here in Egypt.

Also note that the tip122/tip125 have internal protection diodes... do i need to add another diode?

 

[Roff]

The 1Meg pullups will take forever to turn off your PNPs.

 

[ikalogic]

The 1Meg pullups will take forever to turn off your PNPs.

so.. what solution you propose:

- make it 10Meg instead?
- Compleatly remove it..?

 

[Nigel Goodwin]

so.. what solution you propose:

- make it 10Meg instead?
- Compleatly remove it..?

No, make it MUCH smaller!.

 

[ikalogic]

No, make it MUCH smaller!.

ok then,.. a 470k would be suitable ?? 100K?

i don't wanna get to close to the 10K value, cause i am affraid to make a voltage divider effect...

 

[Nigel Goodwin]

ok then,.. a 470k would be suitable ?? 100K?

i don't wanna get to close to the 10K value, cause i am affraid to make a voltage divider effect...

Think about what it's doing! - there's no problem with 10K - or with less than 10K, and make the other resistor lower as well!. 10K is rather high to saturate the transistor anyway!.

 

[ikalogic]

okay.. i am begining to get really confused..

can anyone tell me 2 arbitrary values for R1 and R5 (which are 1M and 10K on the schematic...)

thx

 

[Roff]

The pullup resistor only has about 1.5V to 2V across it when the transistor is on. It therefore only carries Vbe/R amps. Even if you made it 1K, it would draw only around 1.5mA to 2mA. Scale the other resistor to account for this current. Active turn-off (as in your NPNs) is much faster.

 

[ikalogic]

THX A LOT!

i got it!

now i made some calculations, and i'll use a 1k ohm pull up resistor and a base resistor of 470 ohm, resulting in a base current of about 15 mA.

with the high HFE of 1000, i think 15 mA base current will be enough.

now i still got one questions, is it dangerous to rely on the integrated protection diodes? should i add other diodes?

 

[dknguyen]

YOu have no body diodes in your schematic since you are using BJTs and BJTs don't have that parasitic diode that MOSFETs do. Preferably add some Schottky diodes.

But yes, it can be risky to rely on the MOSFET body diodes. It can be done, but you should really look at the spec sheets, your switching scheme, and power levels before you decide.

 

[audioguru]

You looked at current gain, when the transistor is linear with 3V between its collector and its emitter.
Instead, look at its base current when it is saturated, it has much higher base current.

 

[ikalogic]

now i made some calculations, and i'll use a 1k ohm pull up resistor and a base resistor of 470 ohm, resulting in a base current of about 15 mA.

with the high HFE of 1000, i think 15 mA base current will be enough.

Am i right on those points?

Then, i found this schematic...
**broken link removed**
according to what you guys said here, the performance of the 2 transistors Q1 and Q2 is not very good with reguard to the turn-off time.. right?

he should add pull up resistors, right?

 

[ikalogic]

You looked at current gain, when the transistor is linear with 3V between its collector and its emitter.
Instead, look at its base current when it is saturated, it has much higher base current.

ok, then i'll have to raise the base voltage with 10 or 20 more milli amps...

 

[ikalogic]

Thanks everybody,

now here is the final version, after all your comments:

**broken link removed**

Leave the logic part.. i know its fine, please just concentrate on the red shaded part and veryfy its free of errors.

I calculated all the resistances so that the base current in the 4 TIP transistors be around 38 to 43 mA.

thx a lot

 

[Sceadwian]

This is only going to increase efficiency during switching, and not on the order that a mosfet already enjoys by it's very nature. If you just used the proper power fets in the first place, the logic gate's alone are capable of driving the bridge directly... This is why FET's and IGFETS exist.

 

[ikalogic]

This is only going to increase efficiency during switching, and not on the order that a mosfet already enjoys by it's very nature. If you just used the proper power fets in the first place, the logic gate's alone are capable of driving the bridge directly... This is why FET's and IGFETS exist.

Thanks, but Again, i am in egypt, I can easily find TIP transistors, and they are much cheaper. i want to do it with TIPs. and its doable, I guess. and even with my first design which had some errors, i had pretty good results.

so i just wan't to make sure that this is the right way to do with TIPs.

thanks

 

[ikalogic]

ok guys.. as i am not getting anymore responses, i guess this last schematic is allright...

please drop a work if you feel something is wrong.

thx