Looks good. I was kind of surprised that you lost 1.5V using MOSFETs. I knew bipolars would drop that much. Is the voltage drop current-related? 2A no-load is a lot...
Those MOSFETs are poorly spec'd for a high current motor controller - 200 milliohms is a very large Ron by modern standards (5 to 20 milliohms devices are common now).
The circuit returns the high-side gate drive to ground, exceeding the Vgs rating of +/- 10V for the device if V+ is 12V. R1 and R3 should go to the source of the relevant device, not ground. The whole gate drive circuitry is way way too high impedance if you are going to use PWM. Those MOSFETs really want 50mA or so of gate drive to switch fast and clean (modern devices need even more). You are currently switching in about 100us or so, those devices can switch 1000 times faster (with 1000 times less switching losses).
I've got some Micrel MIC4426 MOSFET drivers, they solve a lot of problems. I've used them with FQP7N10 MOSFETs to drive a large stepper motor. I designed a PCB which I made at home, if anyone is interested.
Mark_T,
V_gs is the voltage from gate to source. With regard to the high-side transistors, it's still the voltage gate to source (it's not gate to ground, that's "V_g".)
With 9V supply, Q1(Q3) "V_g" is 14.19V, but V_gs is 5.22V
With 12V supply, Q1(Q3) "V_g" is 19.11, but V_gs is 7.25V.
"The whole gate drive circuitry is way way too high impedance if you are going to use PWM."
-- Hold on to your hat. --
Leon,
Cool. I'm all for that. Start a subject, "Leon Presents..."
My point is that when you are not driving the gate the resistor brings the gate to 0V while the source might be +12V because some other FET is still on (or the reverse diode is conducting - in which case the source will be 12V + 1V forward drop.... Then the Vgs is -12 or -13V...
Oh, erco --
With a nominal 0.2-0.3Ω perfect world on resistance, call it 0.6Ω total; times 2A = 1.2V, and the absence of heatsinks, it's all expected. I think it's mostly imperfect world junction heating with some little losses here & there adding up.
The work goes on.
You're doing the Lord's work, PJ! Pardon my MOSFET ignorance, you know I'm a relay man! I bought some big bags of IRF510s and 511s, just ain't had the time to play.
PJ,
I have been following with interest and am about to try your H-bridge with some 12 volt motors that I have with parts I have in the spares box. Two questions come to mind. One, where is the base of Q1/Q2 connected? Two are you powering the 7555 from the 12V source or some lower voltge?
Jim
Great work & blog! Gotta love that hand-drawn blueprint (OK, blackprint) schematic. Very tidy and neatly labeled. You might just be the next Forrest Mims...
Great work: PJ Allen
'
I have a question about paralleling MOSFETs to increase current drive for bigger motors,using your design.
'
Would the increased gate capacitance kill this idea?
'
I have a working PWM H-B driver circuit that can Handel 40amps. But its not very efficient.It also uses alot more parts than your simple well thought out design does.
'
Any thoughts would be greatly appreciated.
Great work & blog! Gotta love that hand-drawn blueprint (OK, blackprint) schematic. Very tidy and neatly labeled. You might just be the next Forrest Mims...
Comments
Thanks for posting H-Bridge Development and related info.
The circuit returns the high-side gate drive to ground, exceeding the Vgs rating of +/- 10V for the device if V+ is 12V. R1 and R3 should go to the source of the relevant device, not ground. The whole gate drive circuitry is way way too high impedance if you are going to use PWM. Those MOSFETs really want 50mA or so of gate drive to switch fast and clean (modern devices need even more). You are currently switching in about 100us or so, those devices can switch 1000 times faster (with 1000 times less switching losses).
My point is that when you are not driving the gate the resistor brings the gate to 0V while the source might be +12V because some other FET is still on (or the reverse diode is conducting - in which case the source will be 12V + 1V forward drop.... Then the Vgs is -12 or -13V...
You're doing the Lord's work, PJ! Pardon my MOSFET ignorance, you know I'm a relay man! I bought some big bags of IRF510s and 511s, just ain't had the time to play.
I have been following with interest and am about to try your H-bridge with some 12 volt motors that I have with parts I have in the spares box. Two questions come to mind. One, where is the base of Q1/Q2 connected? Two are you powering the 7555 from the 12V source or some lower voltge?
Jim
'Nicely done and presented!
With the MOSFETs always either off or in saturation, and with your revised sequencing algo, how much of a heatsink do you think you really need?
-Phil
With the low side Mosfets being driven by 5 volts are they going into saturation when turned on?
Jim
'
I have a question about paralleling MOSFETs to increase current drive for bigger motors,using your design.
'
Would the increased gate capacitance kill this idea?
'
I have a working PWM H-B driver circuit that can Handel 40amps. But its not very efficient.It also uses alot more parts than your simple well thought out design does.
'
Any thoughts would be greatly appreciated.
Or the next Rod Smart...
I agree though. Thanks for documenting this P.J.
'
I'll give it a try.