NPN Transistors as sole H-Bridge (no PWM), just "old-school" brute-force!

[aquamon]

Want to connect the emitter of Q3 back to circuit efficiently instead of dumping to ground (thru M-).

Ideas please (moved)

 

[audioguru]

Why post this second thread about it?
Emitter followers have too much voltage loss to be used as switches. They should be PNP transistors.
Darlington transistors have even more voltage loss. They should be P-channel and N-channel Mosfets.

 

[aquamon]

Why post this second thread about it?
Emitter followers have too much voltage loss to be used as switches. They should be PNP transistors.
Darlington transistors have even more voltage loss. They should be P-channel and N-channel Mosfets.

Second thread, no: That wasn´t my thread to start with.

I did testing as you implied and I only get an additional 0.01 voltage difference between the two configurations; base resistor of 1K2, might change R2 to that (higher and motor speed drops) to save in reverse; then again, maybe not as this is only ¨1/2-a-circuit¨.
Maybe you think tip122 is J-fet? (NOT an amplifier).
However, total current does increase by 2mA (for a motor drawing 35mA straight, insignificant for a redesign).

Please, do you own tests before posting ¨theories¨; also, doesn´t matter if I use C945, C2120 or TIP122: reacts same, thank you!

At this point, please prove your ¨theory¨.

 

[MikeMl]

...
Please, do you own tests before posting ¨theories¨;...At this point, please prove your ¨theory¨.

Aquamon, I suggest you do some theory proving of your own. Your circuit has some serious drawbacks.

First, your circuit is just a motor reversing circuit. I wouldn't call it an H-bridge, because that usually means that it is being controlled by logic inputs, which yours is not. Yours requires that it be controlled by two external toggle switches or relays; one to turn it on/off (motor run/motor stop), and the other to select the motor direction (requires tying Rev to either the battery voltage, or to ground). Begs the question, if you need toggle switches (or relays) to control the circuit, why bother with the circuit; just control the motor directly from the switches?

Better still, just use a center-off, Double-Pole, Double-throw toggle switch. Only one switch, zero power wasted.

Second, a quick simulation shows how inefficient your circuit is.



With Rev tied low, with a 100Ω load, the voltage drop across the upper Darlington is 1.5V and the drop across the bottom one is 0.7V, leaving only 9.8V across the load (motor). With a battery voltage of 12V, that means you wasted 2.2V.

With Rev tied high, with a 100Ω load, the voltage drop across the upper Darlington is 1.9V and the drop across the bottom one is 0.7V, leaving only 9.4V across the load (motor). With a battery voltage of 12V, that means you wasted 2.6V. The difference is due to the additional drop across D1; why is it even there?

As suggested by Audioguru, contrast your motor reversing circuit to the simplicity and performance of what modern PFETS and NFETs can do:



Note that with such a light load (100mA), the voltage drop across the both top and bottom devices is a few mV. This circuit could switch several Amps and not break a sweat. Try that with yours...

 

[audioguru]

A TIP122 is a power darlington transistor. With a motor current of only 35mA then you do not need a power transistor (a huge TIP122 has a maximum allowed current of 8000mA peak!).
Don't you know that a darlington is TWO transistors in series so that their base-emitter voltages add and make a voltage loss of 1.4V to 2.5V? The Japanese 2SC945 and 2SC2120 are ordinary single little transistors with a maximum allowed current of only 100mA and 800mA and a base emitter voltage of 0.7V to 1V.

A motor switching circuit does not use emitter-followers. Instead it uses transistors that have the proper polarity to be on-off switches.

 

[aquamon]

Aquamon, I suggest you do some theory proving of your own. Your circuit has some serious drawbacks.
First, your circuit is just a motor reversing circuit. I wouldn't call it an H-bridge, ...

Well, I did notice how everyone is starting to believe H-Bridge is originally from logic circuits.

Firstly, H-bridge was previously done via driving a relay.

Secondly, logic circuits and micro-controllers are for the ¨brain-dead¨ when it come to the H-Bridge; hence, most are pushing unproven theories.

Third, my circuit is always ¨FORWARD ON¨, when the remote´s LED switch is activated, then REVERSE: ONE SWITCH controlled.

Fourth, this was done with what I have lying around, basic COMMON parts, no special orders.

Fifth: the intention is to go for more power; hence, TIP122 (manufactured by ST) was planned from day 1.

Sixth: I have done my research long before they ever had a ¨chip¨: limited current.

Seventh: Battery Voltage measured 7.28 - Motor Vdrop 5.95 = Circuit loss 1.33V
Which would be a lot more efficient (FUN) using 12V.

Eighth: DRAW your suggestion, can´t figure out how to do it, eh?

You guys are way too young and probably need a degree in electrical engineering before you go pushing buttons for further banning of members.

Like I said, it works beautifully and I AM extremely happy with it.
That is all that really matters in the end.
SO, THERE!!!

 

[Nigel Goodwin]

Like I said, it works beautifully and I AM extremely happy with it.
That is all that really matters in the end.
SO, THERE!!!

To be honest we'd be far happier if you didn't post such useless and badly 'designed' circuits here, it's helping no one - FAR, FAR better to do it properly.

 

[ronv]

Just goes to show you. Beauty is in the eye of the beholder.
You young whippersnappers.

 

[aquamon]

Funny, my diagram is already HERE!
Design 23
Now, which of you belongs to that site?
For 4X60 cents= $2.40 (4 of TIP122)
and other ¨junk¨ lying around, how can this be an ¨EXPENSIVE WASTE¨
when a
L298 =>$4+++???
not calculating all the other ¨logics¨ and PWM devices needed...

 

[audioguru]

The Talking Electronics site shows many ways to drive a motor. Some circuits wrongly use emitter-followers and others wrongly use darlington transistors, design 23 uses both.
The TIP122 was designed by Texas Instruments, not ST Micro. Maybe ST Micro pays a royalty to Texas Instruments to be able to copy them.

 

[aquamon]

Sorry, not my fault most of you ¨most-helpfuls¨ missed the entire point about this.
This is a cheap and workable solution for those that need an H-bridge that works without the headache & EXPENSE of going to relays; or even, PWM and logic circuits which FAIL with NOISE interference.

Given the open option to use J-FETS to those who are seeking better efficiency (above 82%; if they blow-up, then they have you to blame as I have not tested w/ FET´s).
That´s about the only thing that I may agree with right now...
I still don´t agree with your ¨emitter-follower¨ theory and have seen this sugested in the past why a circuit didn´t work; so, I deliberately proved it (¨theory¨) wrong intentionally via this design (to myself at least).
AND, also, so-as-if the base is driven ¨too hard¨ then Ib isn´t wasted as such...excess goes thru motor (NPN: ¨emitter-follower¨).

I suggest you take a NPN ¨anything¨ and hook it up to an impedence load ¨emitter-follower¨ and see for yourself; as stated earlier, I used up to 1K2 (Rb-c) yesterday w/out motor speed loss.

Also, if I remember right, different manufacturers use different internal resistor values; so, no royalty-fees

 

[atferrari]

You guys are way too young and probably....

Hi Aquaman, you made laugh with that! Good point...and I am afraid, totally wrong!

You started a thread that seems to be going to nowhere....



Happy New Year!

 

[aquamon]

Hi Aquaman, you made laugh with that! Good point...and I am afraid, totally wrong!

You started a thread that seems to be going to nowhere....



Happy New Year!

Yeah, so much for asking for help....the time wasted on forum, could´ve done by now, myself.

 

[aquamon]

Aquamon, I suggest you do some theory proving of your own. Your circuit has some serious drawbacks.

First, your circuit is just a motor reversing circuit. I wouldn't call it an H-bridge, because that usually means that it is being controlled by logic inputs, which yours is not. Yours requires that it be controlled by two external toggle switches or relays; one to turn it on/off (motor run/motor stop), and the other to select the motor direction (requires tying Rev to either the battery voltage, or to ground). Begs the question, if you need toggle switches (or relays) to control the circuit, why bother with the circuit; just control the motor directly from the switches?

Better still, just use a center-off, Double-Pole, Double-throw toggle switch. Only one switch, zero power wasted.

Second, a quick simulation shows how inefficient your circuit is.

View attachment 90096 View attachment 90097

With Rev tied low, with a 100Ω load, the voltage drop across the upper Darlington is 1.5V and the drop across the bottom one is 0.7V, leaving only 9.8V across the load (motor). With a battery voltage of 12V, that means you wasted 2.2V.

With Rev tied high, with a 100Ω load, the voltage drop across the upper Darlington is 1.9V and the drop across the bottom one is 0.7V, leaving only 9.4V across the load (motor). With a battery voltage of 12V, that means you wasted 2.6V. The difference is due to the additional drop across D1; why is it even there?

As suggested by Audioguru, contrast your motor reversing circuit to the simplicity and performance of what modern PFETS and NFETs can do:

View attachment 90098

Note that with such a light load (100mA), the voltage drop across the both top and bottom devices is a few mV. This circuit could switch several Amps and not break a sweat. Try that with yours...

Nothing on this site is 100% efficient...your voltage drops are inaccurate and you also need to remember CURRENT flow takes the EASIEST path...into consideration.
Remembering: several members were banned during the lengthy and foolish OHM´s Law debates.

 

[aquamon]

Hi Aquaman, you made laugh with that! Good point...and I am afraid, totally wrong!

You started a thread that seems to be going to nowhere....



Happy New Year!

 

[MrAl]

Hi,

What is D1 doing for your circuit? It appears to be in series with the motor when the left upper transistor turns on. That would drop some voltage when it is switched into that mode.

Maybe you can start with two half bridges without any diodes and work from there. Figure out how to control both half bridges from the input alone. You can then add parallel protection diodes later which wont be in series with anything.

 

[audioguru]

We NEVER talk about "cheap parts" or "salvaged parts". Brand new high quality electronic parts are plentiful and inexpensive.
Why don't you believe that an NPN emitter-follower has an output voltage at its emitter that is 0.7V to 1V LESS than its base voltage? If it is replaced by a PNP transistor then the voltage loss is only about 0.05V which is its saturation voltage drop at low currents. Since a darlington transistor has two emitter-followers in series then its total voltage loss is doubled compared to a single emitter-follower.

A Jfet is a weak little thing. A power Mosfet conducts much better with a very low voltage loss.

 

[MrAl]

Hi,

Audioguru you reminded me that we havent yet talked about dead time either. We need to do that as well.

To the OP:
Dead time is the time between turning off a lower transistor and turning on an upper transistor, and vice versa. This is required so that the two transistors do not form a short circuit right across the DC buss line. Bipolar transistors have significant storage time so they need plenty of time to turn off, during which time the other transistor in that half bridge has to stay off.

 

[MikeMl]

Mr Al,

shoot-through is not much of a consideration because of this important point I made about aquamon's posted circuit:

"First, your circuit is just a motor reversing circuit. I wouldn't call it an H-bridge, because that usually means that it is being controlled by logic inputs, which yours is not. Yours requires that it be controlled by two external toggle switches or relays; one to turn it on/off (motor run/motor stop), and the other to select the motor direction (requires tying Rev to either the battery voltage, or to ground). Begs the question, if you need toggle switches (or relays) to control the circuit, why bother with the circuit; just control the motor directly from the switches?"

Better still, just use a center-off, Double-Pole, Double-throw toggle switch. Only one switch, zero power wasted.

But what the hell do I know?

According to Aquamon, I am "way too young and probably need a degree in electrical engineering before you go pushing buttons for further banning of members."

 

[Pommie]

Yeah, so much for asking for help....the time wasted on forum, could´ve done by now, myself.

You asked for help.
People gave you good advice.
You told them they were wrong.
Why did you ask for advice in the first place?

Mike.