Haptic Circuit Issue

[waymond91]

Hello All!
I am working on a little project similar that involves a lot of haptic feedback.
So I needed a way to drive about 12+ ERM vibration motors with a few cables (i.e. NOT 12 pwm lines running to each motor). Each motor has its own PCB.
To this end, I designed a little circuit using 2 d-type flip flops that essentially make a 12bit shift register.
The datasheet for the flip-flops is here: http://www.ti.com/lit/ds/symlink/sn74hc74.pdf
On the output of each of the boards there is a PWM line comming out, this connects to a TI DRV2603, a simple NPN transistor in case the TI motor driver didn't work, and lastly an LED.
Datasheet for the TI motor driver: http://www.ti.com/lit/ds/symlink/drv2603.pdf
Transistor datasheet: http://www.mccsemi.com/up_pdf/MMBT3904(SOT-23).pdf

This all works more or less OK. I can drive the motors using the transistor if the LED is not attached. I assumed this was because the flip-flop can only source/sink +/- 25mA. However the LEDs should only draw maybe 5 mA,  and the transistor gain SHOULD be high enough that it shouldn't need it all...

Anyway, here is what I am really struggling with. Whenever I solder in the TI ERM driver in what I believe is the correct orientation, the circuit instantly shorts when I put it in my breadboard ( I have tried this several times). Trying the other orientation does not instantly short the circuit, but it does not drive the ERM either. Even though all of the shift registers conditions are being meant (i.e. I can drop in a copy of the board that uses the transistor and it works perfectly), the PWM output line seems to just float at about 2V (3.3v system).

Any ideas? Just wanted to make sure I was not missing something obvious.

Here is the schematic and couple pictures of the PCB.

Thanks in advance!!

[rstofer]

I didn't read through the entire problem but I will say that the '74 device is intended to source or sink about 4 mA, not 25 mA.  25 mA is an Absolute Maximum rating - that the point where the magic smoke is getting staged for an escape.

Look at the specs for Voh and Vol when the supply voltage is 4V and you will see that the device is intended to source/sink 4 mA,  You might get away with a bit more than that but 25 mA is not the way to design with this gadget.

Your 47 Ohm base resistor seems a bit low.  I would think that 4 mA and 3V drop (3.8V output - 0.8 Vbe) would be the design condition so 750 Ohms would be more like it.  I didn't check to see if Hfe * 4 mA would control your load...

You might think about slacking off a bit on the '74 device.

[jmelson]

R1 sources about 91 mA to the transistor base, assuming a 5V Vcc.  That's WAY outside the ratings of a 74HC part.  The LED current is just a bit less, but still well above ratings.  How did you (mis)calculate these resistor values?

Jon

[waymond91]

Thank you for the response!
Looks like I was pretty off. To be honest, for the LED I used digikeys "LED Series Resistor Calculator" here:
https://www.digikey.com/en/resources/conversion-calculators/conversion-calculator-led-series-resistor
Although now that I look at it again, its telling me I am indeed way off. Especially considering we're targeting 4mA max output for the flip flop.

As for the transistor current biasing, this was a left over value from a previous project. Where I did in fact (mis)calculate the resistor at about 1K, but the motor refused to fire until I put in this much smaller one...

But that certainly explains some of the behavior I have been seeing! I will make the appropriate changes to the transistor and the LED circuit and report back.
I sometimes work from home and some of these projects drag out so long I often forget what design work I've done and what I haven't!

As for the DRV2603, the board I am testing it on does not have the transistor nor the LED populated. So the current supply should not be slammed trying to source those components.
Looking at the data sheet for the DRV2603 it says it only needs a couple of micro-amps to trigger the LED input. So I am not sure what I am doing wrong.

The datasheet is also pretty lax on describing the device orientation (index pin). I did find a small dot on the package and assumed that was the pin 1 index. However, doing this a couple times shorted the circuit each time. Its a small footprint and maybe I just soldered it wrong a couple time in a row? But there definitely weren't any solder jumps between pads...

Looking at the component pins with a jewelers loop reveals all of the pins have about the same shape, so thats not really helpful.

Thanks again and let me know what you think!
Cheers!
Brett

[waymond91]

I snapped a picture with my phone through the jewelers loop to try and show the soldering is passable...