3vdc motor PWM problem

hi all. i posted awhile back about my time fountain project, everything is up and running! it works beautifully (for a few seconds, read on) with my new submersible dc motor pump. if you dont remember the project, heres the deal:

I need a constant drippage of water. I do this with a 3v dc motor, which pumps through a tiny valve then out the tube. its PWM with the following: 555 timer, diodes separate up and down time. 1uf cap - 1KOhm up, 1KOhm down

The timer is isolated from the circuit with a diode and a 1000uf cap, so the motor doesn't interfere with the timer. thanks for the help with this. The timer works perfectly, as with a piezo it gives a steady tone.

The problem is that the motor will rev up to the correct rpm, which yields the perfect flow of water. Then, after a few seconds, its RPMs will drop lower and keep bouncing up and down (i can tell this because there is an audible difference in the tone from the motor, and the water flow fluctuates)

Any idea what is causing the motor to not stay at the steady rate? Thanks a ton!

-sam

Reply to
randomname
Loading thread data ...

one more piece of information: the power supply is 6 volts.

Reply to
randomname

Back pressure? How do you relieve the pressure?

Reply to
Homer J Simpson

I don't... I need the pressure so that the flow rate is lessened.

Reply to
randomname

The problem does seem to be the backpressure... but its strange.

When I first turn it on it will work for a good minute or so, then it will start going into "struggle mode" for a few seconds, then it will go back to normal. It will alternate between the two states for awhile and then it will just settle at "struggle mode".

Struggle mode is when its RPMs lower and it doesn't pump enough (bad).

How can I fix this?

Thanks a lot

-sam

Reply to
randomname

Update:

even without the valve, the motor still struggles. I dont get it.

Reply to
randomname

--
Sounds like you might have pin 5 floating.  If so, try running it to
ground through a 0.1µF capacitor.
Reply to
John Fields

There is already a .1uf cap from pin 5 to ground... and I don't think it's a timer issue, because when I hook up a piezo in parralel to the motor, it gives a steady tone.

Any other ideas?

Reply to
randomname

Your motor / pump may not like running at the rpms and pressure you have. Are you trying to build "The Time Fountain"?

formatting link

If you deliver less than the needed pressure your system won't work. If you deliver more the motor / pump won't like it. That's why you need a relief of some sort - a spring loaded valve or similar.

Reply to
Homer J Simpson

hmm... sounds like the timer is floating around on its bias and alternating your duty cycle.. maybe you mite want to clamp the input with a .1 or so cap for the bias control. its been a while but i think it mite be #5 pin? could be wrong there. also you mite want to make sure your motor isn't getting hot and make sure your timer isn't getting hot.

--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5
Reply to
Jamie

Sounds like something is overheating and shutting down. I don't think the 555 has thermal shutdown, but I bet any voltage regulator you're using does. A finger should be an adequate instrument to test for overheating in this instance, provided you bear in mind what you're about to touch might be at 150 C or so (initially just brush it with your finger, don't grab it till you know it's not too hot to handle).

Tim

Reply to
Tim Auton

Yes, that's what I've made.

I do think it's a motor issue, but I don't understand how the motor can affect the 555 timer. The timer slows its frequency too (I can hear the change in the piezo hooked up to it).

How do I solve this? Where could I find a suitable valve?

I'm going to experiment with different PWM setups (different duty cycles, frequencies) tonight to see if the problem persists.

Thanks,

-sam

Reply to
randomname

It sounds like you've built an oscillator.

The PWM comes up, the motor pumps, and you reach a head of pressure. At this point, you need to regulate your PWM by, maybe pressure or flow rate, not just constant-current. The motor fills up the vessel, the pressure reaches the rated value, and the motor stalls. That relieves the pressure, and the motor picks up speed again, until the pressure builds up again and the motor stalls again, and so on.

You could put some kind of air buffer in the pipe, or an actual pressure sensor and make a servo system.

Good Luck! Rich

Reply to
Rich Grise

Well, I changed the 1uF cap on the PWM circuit to a .1uF cap, effectively decupling the frequency (and learning a new word in the process).

The result? Success! I've been playing with this thing for hours, its starting to get old, but was definitely worth all of the work.

The trick is to control the flow from the tube so that it drips at a certain very specific rate.. any faster and it collapses into a stream. Any less, and it doesnt drip at a constant rate. Luckily, thanks to some property of water (surface tension I guess) this "drip zone" is larger than the fluctations of the pump that would cause it to go into one of the unstable configurations (too little drip, or a stream). Now that the motor is running steady, the system is very stable.

Next step is to find out how to increase the liquid's surface tension so I could increase the drip rate without getting a stream, and form larger drops.

Thanks again for your help, everyone!

Reply to
randomname

I have an idea for you. forget using the pump pressure to control the dripage. Use your pump to fill a reservoir with an overflow tube, then to regulate the drip make a pulse controlled valve with a solenoid and a soft silicone tube where the solenoid is compressing the tube closed, then use the timer to activate the solenoid to control the fluid thru the tube. After the initial adjustment of a spring to control the pressure on the tube, the rest is by the timing of the drip pulses.

Reply to
James Thompson

I was thinking about this the other day... another 555 timer would just control the solenoid.

Do you know where I could find a suitable solenoid? Also, I'm not sure a water drop will fall out fast enough from gravity alone, and it might require the resevoir to be pressurized (tons of other complications)

For now, this system works, the motor has been running happily for about 5 hours. Only problem is getting the drippage to stay in teh "sweet spot" ... not too fast for stream, not too slow for inconsistancy.

Right now my "valve" is a jumper twisted around the tube, which might not provide the most consistant pressure.

Reply to
randomname

--
Once you\'ve got a steady stream of drops going, you can illuminate
them properly by changing the frequency of the strobe to coincide
 Click to see the full signature
Reply to
John Fields

John's approach seems most logical with the triggering of the strobe. Make a constant gravity feed drip like they have on medical IV tubes, then have a 2 wire drip sensor to detect when a drip leaves the tube and sync the strobe from that. When a drip is ready to leave the tube it conducts thru the liquid and the point when it leaves, it will break conduction between the 2 wire sensor. JTT

Reply to
James Thompson

--
That\'s not my approach at all.  I don\'t use a drop sensor.

What I\'m suggesting is an astable with a manual frequency control (a
 Click to see the full signature
Reply to
John Fields

John, that's what I've got already. A simply astable will do, with the "up time" flashing the strobe, and the (adjustable) "downtime" controlling the frequency of the strobes. No need for a monostable.

The problem is getting a steady drip rate, its damn near impossible as far as I can tell. It either drips sporadically or turns into a god damn stream, which really is starting to piss me off.

I think I'm going to need a drop sensor, but I feel like this is going to take a lot more work, and even if I did make a drop sensor, theres no way to get the drops to appear as if theyre moving. It would be easy to get them to freeze (flash strobe whenever a drop leaves), but constantly increasing the time period in which the strobe delays after a drop is detected takes a computer of some sort.

I've acheived some amazing results when I manage to get a stable drip, but right now it's just not happening.

Reply to
randomname

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.