Using PWM with the L293D to get around issue posed by minimum supply voltage

My samples of the L293D motor controller arrived today. After looking around the house to find a small motor and failing, I found a few cheap 3 volt motors on Ebay. Today I looked at the datasheet and noticed that the minimum supply voltage input for powering the motors was 4.5 volts (facepalms self) but I figured I would just sent a PWM signal to the Enable pin to get the right average voltage and hooked my multimeter to the outputs and adjusted the PWM value to obtain an average voltage of around 3 volts. Are there any potential issues with this solution? Just wanted to make sure.

What power source do you use ?
Did you read and understand the datasheet for the L293d ?
There's important information in there that relates directly to your question.
Do you understand the consequences of driving a motor by some transistor (you are actually using a pair of transistors inside the L293d) ?

You found a remark about a minimum of 4.5 Volts, so you figured you'd supply an average of 3 Volts ?

I can't follow your thoughts, so please tell a bit more.

For my power supply I'm using a battery connected to a buck converter.

I'm new to motor drivers, so hopefully I interpreted the datasheet correctly.
Feel free to explain any consequences of driving motors with an IC that uses transistors.
I got the value of a minimum motor supply voltage for the pin that supplies power to the voltage as 4.5 volts from the Vs minimum value (it pointed to Vss which listed 4.5 volts as a minimum) I should also mention that since the Arduino UNO (what I'm using) uses 5 volts for the main VCC pin, my minimum Vs value is actually 5 volts if I interpreted that correctly, so that's what I've set the buck converter to.

I should clarify that I"m providing the minimum required voltage on the voltage supply pin for driving motors, I'm supplying a PWM signal to the Enable pin to get the average voltage to around 3 volts on the output pins that connect to the leads of the motor.

Ok, the way i read it, you got some motors, and after that you got their datasheet.
But you meant the datasheet of the L293d.
Be clear when asking a question, read your question before posting it (and maybe again after doing so).

Transistors will generate a voltage drop (they "eat" voltage).
You have at least 2 transistors in your setup, so you will have 2 times such drop.
That's the consequence of using transistors for this job.
So now it's up to you to figure out how much of the voltage will be lost inside the L293d.
And of course whether or not you need to reduce voltage, and if so then by what value.

I f you need or like to control speed, you should also keep in mind that you can't go to 100%

And last but by no means least:
PWM doesn't control voltage.
It controls current, and any measured voltage is a function of that controlled current unless filtered.
Try to measure a PWM output that isn't connected to anything and be surprised by the result.

Sorry about that, it sounded great when I wrote the OP, but after stepping back and after seeing your reply, I saw that it wasn’t clear enough.

From what I have been reading it’s the Vce value that’s the voltage drop and there is two Vce values listed on the datasheet 1.4 volts and 1.2 volts which would make the total voltage drop 2.6 volts. so 5-2.6= 2.4 volts. I was confused when I took a voltage reading and got 4.89 volts but I noticed that the Vce value mentions the drop when collector current is .6A so I’m guessing it will only come close to that voltage when I put an actual load on the circuit. So I actually need to supply 5.6 volts on the voltage supply pin (5.6-2.6=3 volts).

A voltage drop of 2.6 Volts is a lot.
So have another look to find out if they already added the 2 transistors’ drop.
I would expect a 0.7 volt drop per transistor but i haven’t actually looked at the datasheet recently.

On an older thread somebody mentioned a total of 2.6 volts drop, not sure what it would be listed as if they added the drop.

You can safely power a 3V motor with the L293D driver using a 5V power supply. As previously noted the L293D driver consumes a minimum of 1.8 V, which rises by 1-2 volts as the output current increases. Consult the L293D datasheet for more details. With a 5V power supply, you can be reasonably sure that somewhere between 2.5-3.5 V will get to the motor. Be aware that you can't expect to get more than about 1 ampere per channel out of the L293D, even if it is on a heatsink.

Your thinking is sound, you can lower the effective voltage to a motor using PWM. The real voltage is still a pulse waveform but for a motor this does not matter.
However it is unlikely that a multimeter will give you anything sensible when measuring a PWM signal.
Measure the maximum voltage you get across the motor and then calculate the PWM value you need to drop this to the 3V the motor needs.