The calculation of R22(Q7), R23(Q6) below:
Ib_Q7 = 25mA/10 = 2.5mA
Ib_Q6 = 25mA/10 = 2.5mA

R22 = (V_ESP32-Vbe_Q7)/Ib_Q7
= (5V-0.7V)/2.5mA
= 4.3V/2.5mA
= 1.72K (Try 1.8K)

When the bjt using as switch then the hFE = 10 (some others maybe = 20)
So Ic_Q1 = 250mA
Ic_Q7 = Ib_Q1 = Ic_Q1/10 = 250mA/10 = 25mA

Ib_Q7 = Ic_Q7/10 = 25mA/10 = 2.5mA

With your help I am taking this circuit on part-by part

So we now have an NPN-PNP combination switching the stepper and its driver.



The schematic is a bit upside down but do you otherwise agree with this?

 

You know you can still use a spare channel of the ULN2003 and eliminate Q7 and R22.

 

You know you can still use a spare channel of the ULN2003 and eliminate Q7 and R22.

Cutting out two components will help greatly with routing and space on the PCB

Please double check me:

 

 

You the man!!! It all fits on the PCB now!

I am in awe of the generosity demonstrate in this forum.

 

You know you can still use a spare channel of the ULN2003 and eliminate Q7 and R22.

Nice, although I have had thought about that, but I don't have time to write down yesterday, when you using ULN2003 to replace npn, you need to reconsidering the values of R23 in #43, the Vce of npn as 0.2V, so I ignored it, but Vce of ULN2003 as 0.9V ...

 

Nice, although I have had thought about that, but I don't have time to write down yesterday, when you using ULN2003 to replace npn, you need to reconsidering the values of R23 in #43, the Vce of npn as 0.2V, so I ignored it, but Vce of ULN2003 as 0.9V ...

If I understand correctly: The output of the Darlington array within the ULN2003 does not return as close to zero as would an NPN transistor --- thus a larger R23 resistor (as in #43) would be needed to prevent the current flow into the base of Q7...

Sadly I have not been able to figure out how to calculate this confidently...

 

No, just the opposite, you would want a lower value resistor to make up the voltage drop across the 2003.

I also wanted to offer an alternative to switching off the steppers on the high side…that would be simply switching off the 2003 using a N-Channel MOSFET connected to the ground line.

This avoids the current usage for the PNP, and the voltage drop across it. (and 2 resistors, assuming the control for the MOSFET is push/pull)

 

No, just the opposite, you would want a lower value resistor to make up the voltage drop across the 2003.

I also wanted to offer an alternative to switching off the steppers on the high side…that would be simply switching off the 2003 using a N-Channel MOSFET connected to the ground line.

This avoids the current usage for the PNP, and the voltage drop across it. (and 2 resistors, assuming the control for the MOSFET is push/pull)

I really like this idea... does this reflect what you are saying?

 

Yup.

 

Hi there Spidey

I have finally received the parts and built the circuit. The problem, however, is that when the MOSFET is switched off by the ESP32, the stepper motor is still turning (with little torque; easily stalled). When turned on the motor is much stronger.

What would explain this "leakage"?

Thanks.

 

Please measure the voltages of following points when they turn on and off:
1. Two control pins of ESP32.
2. Two pin 8 of ULN2003.
3. The 8 input pins of ULN2003.

 

To stop a stepper from moving you must also stop the command states from changing. And quite possibly the way that the enable pin is driven is not correct. I see a pull down FET but I see nothing to pull it up.
But the most obvious thing is to stop delivering step commands.

 

In some applications, it is required to hold a stepper motor steady at rest using its built in holding torque, to do this the rated DC current should be applied to hold it stationary.
Max.

 

I see a pull down FET but I see nothing to pull it up.

He said this:
is that when the MOSFET is switched off by the ESP32,

 

Please measure the voltages of following points when they turn on and off:
1. Two control pins of ESP32.
2. Two pin 8 of ULN2003.
3. The 8 input pins of ULN2003.

Thanks Scott

The circuit is really only partly constructed. But I will answer what I can.

1) From the ESP32 the signal is either 5V or 0V
2) Pin 8 of the ULN2003 is at 2.1V relative to ground ---- this is in keeping with the low torque behavior of the motor when its driver is supposed to be "disabled"
3) The input pins of the ULN2003 are cycled by the ESP32 to 0V and 5V continuously

I believe point #2 above is the reason for this...

Even if I remove the FET entirely this continues.

 

Thanks Scott

The circuit is really only partly constructed. But I will answer what I can.

1) From the ESP32 the signal is either 5V or 0V
2) Pin 8 of the ULN2003 is at 2.1V relative to ground ---- this is in keeping with the low torque behavior of the motor when its driver is supposed to be "disabled"
3) The input pins of the ULN2003 are cycled by the ESP32 to 0V and 5V continuously

I believe point #2 above is the reason for this...

Even if I remove the FET entirely this continues.

View attachment 198756

Is there not an error in the way the ULN is being used to drive the stepper?

VCC is 5.0V
Inputs to ULN are either 5V or 0V

The ULN is physically disconnected from GND and the motor is still turning with low torque.

Puzzled..

 

The enable/disable- line needs to be either high or low, inbetween values will lead to undefined operation. Read the specs on it.

 

The enable/disable- line needs to be either high or low, inbetween values will lead to undefined operation. Read the specs on it.

Please explain fully.... The inputs from the ESP32 are either high or low. Nothing between.

 

Even if I remove the FET entirely this continues.

So the leakage must be through the ULN2003 and then presumably back to the MCU. There is nowhere else it can go.
Back to the high side PNP?