Best way to enable an H-Bridge

Thread Starter

akistler

Joined Jan 8, 2018
9
GopherT,
I'm not sure about in a production setting, but since I only have the discrete components, that's what I'm using :D

ScottWang,
I built the 1n4148 version (I don't have any schottky diodes handy) and it acts strangely. With the EN off, the voltage across the motor leads is 10.7V. The EN signal is the output of a 555 timer, I tried directly connecting the EN signal as well as running it through a buffer and an inverter. With the buffer, the voltage remains 10.7V; with the buffer, the voltage sits at 1.3V and the circuit audibly whines.
 

ScottWang

Joined Aug 23, 2012
6,827
@akistler.
Which circuit that you tested? (#22 or the second circuit in #24)

How is the resistance of motor and its draw current?
(if you used the 2N3904 and 2N3906 to drive the motor then I can imagine that the current is quite small and it as the motor of toy, but the small toy just use the 3V motor)

What kinds of buffer and inverter?
 

Thread Starter

akistler

Joined Jan 8, 2018
9
Circuit #24

My real load is a solenoid, not a motor. The measured voltages in the previous post were with a DMM as the load. The measured resistance of the solenoid as 4 ohms, which at face value should require 3 amps at 12V, however it consistently actuates with a relay sourced by a 12V 200mA supply.

I've attached the buffer and inverter I used previously, as well as the current circuit. It's not working at the moment, and whines when the board is powered, I'm going to swap out the transistors, in case I have bad ones/damaged them. I'm also considering reverting to my previously working circuit, as there won't be the voltage drop due to the moving of the EN from the power side of the bridge to the base of the NPN.
 

Attachments

crutschow

Joined Mar 14, 2008
23,362
solenoid as 4 ohms, which at face value should require 3 amps at 12V
The 2n3904/6 are rated at 0.2A max, so they obviously won't work. :eek:
Suggest you go to power MOSFETs for that value of current.
Another advantage of MOSFETs is that you don't need the external protection diodes, as their internal substrate diodes can provide that function.
 

ScottWang

Joined Aug 23, 2012
6,827
Do you have the datasheet or links of the solenoid?
Probably you can buy MR751 Rectifier Diode, 6a, 100v Motorola as the emf diode for solenoid, that is choose 4 times of voltage as Vemf = 12V * 4 = 48V, some emf could raising up to 10 times of the provides voltage.

For a 200mA current that the bjt should use at least 3 times of its rated current as 600mA, choose 1A is better as 2SC1384 and 2SA684.

For a 3A current that the bjt should use at least 3 times of its rated current as 9A, choose 15A is better as 2N3055(npn) and 2N2955(pnp), but heat is the problem for the power transistor.

With the condition of solenoid maybe you can use the MOSFET is the best way for the heat issue, the Vds at least with 1.5 times of 12V that it is 18V, but use 24V is better, choose the Rds(sat) <=20mΩ, choose the Ids with 3 times of rated current as I = 3A * 3 = 9A, but choose the Ids as 12A or 15A is better.

Here are some of MOSFET of N+P pairs for your reference, you can also choose some independently N and P type MOSFETs.

H bridge-1-18-18_akistler_crutschow_ScottWang-06_4MOSFETs.png
 
As title says 'best way to enable', you are missing the action of enable on the upper FETs (yes not matters much in practice, but you want the best way) see L298 block diagram.
 

crutschow

Joined Mar 14, 2008
23,362
As title says 'best way to enable', you are missing the action of enable on the upper FETs (yes not matters much in practice, but you want the best way) see L298 block diagram.
Why you think that's better than what has been posted?
I don't see the enabling or disabling the upper FETs is any better than controlling the lower FETs.
 
in an H bridge design like l298 the enable of the upper FETs is to also automatically include a safety feature, so you cannot shot through if your uC gets lost and send some abnormal turn on signals. Im not saying the posted design would not do the job (because the voltage is low), but if there is a small chance in theory that something would go wrong, in practice it will go wrong :). The user is controlling a solenoid, i think it should use N fets for upper too, I think its more safe than to have the gate of a pfet in the flow circuit of a device generating collapsing field spikes.
 

ScottWang

Joined Aug 23, 2012
6,827
@Sirbu Constantin.
I'm worry about the L298 maybe will damaged not for long when you use it to drive a 3V solenoid, so could you draw the circuit and describe it why it can handle a 3A load has no any problem?

The specification of L298:
Total DC current up to 4 A
O Peak Output Current (each Channel)
–Non Repetitive (t = 100μs) ............................. 3A
–Repetitive (80% on –20% off; ton = 10ms) ... 2.5A
–DC Operation .................................................. 2A
 

ScottWang

Joined Aug 23, 2012
6,827
The user is controlling a solenoid, i think it should use N fets for upper too, I think its more safe than to have the gate of a pfet in the flow circuit of a device generating collapsing field spikes.
Since the power is +12V, when you using the N type for the upper driver and then the Vgs = 12V+3V = 15V, where can you get 15V, please suggest a method?
 

crutschow

Joined Mar 14, 2008
23,362
I think its more safe than to have the gate of a pfet in the flow circuit of a device generating collapsing field spikes.
The collapsing field just tends to keep the current flowing in the solenoid.
That current is absorbed by the substrate diodes when the MOSFETs turn-off so I don't expect any significant spikes.
To minimize and effects of wire inductance, the gate resistors should be connected to the source/drain terminals of the opposite side transistors.
If you are not too concerned about switching speed you could also using larger gate resistors on the P-MOSFETs so they can act as an RC filter with the gate capacitance to roll-off any fast spikes.
 
Top