IronJungle
Senior Member
Top of PCB Project Board Pic
Piezo echo
- Thread starter IronJungle
- Start date
The photos indicate you're not putting the test lead from emitter to GND, but from the motor to the collector. BIG difference if correct.
Thumping all that base current into a 20 AMP darlington with Hfe >1000 and a stalled motor in the collector will give a large current spike, especially as I don't see any decoupling caps... Adding resistance to the motor significantly reduces the spike.
Add decoupling, replace the darlington with a Tx - 2n2222 won't get hot if saturated, and try again.....
Last edited:
Armp has some wise words there.
I may have missed something , but what voltage is the motor rated?
A Darlington may have loads of gain but a healthy Vce too.
Any drooping here will make motor starting weedy.
Having a base resistor may save the PICAXE. I have seen RC in base lines too as a method of softening the switch-on.
(softening the switch cf gate-current limiting in MOSFETs).
The 'test lead' effect may just be a ground thing, I can't really follow the photos.
But the basics of wiring micros with noisy things crops here a lot.
1. As said by Armp, have decoupling/bypassing caps on the PICAXE.
In my very humble opinion this should be considered manadatory.
C2: Stick 100nF Ceramic as close to PICAXE power pins as possible.
May need a //d electro too.
Without decoupling/bypassing your 'luck factor' increases.
2. Also some kind of 'reservoir' cap where the motor is.
This reduces dips/transients getting to PIC supply.
C1: 47uF upwards without going daft.
3. As far as possible , separate the grounds from Motor-supply, from PIC-supply. e.g. my additions to your schematic.
In general, remove any 'power sharing' of supplies.
Unless you are doing fast PWM consider an N MOSFET for the motor switching. I'm not a Darlington fan.
I may have missed something , but what voltage is the motor rated?
A Darlington may have loads of gain but a healthy Vce too.
Any drooping here will make motor starting weedy.
Having a base resistor may save the PICAXE. I have seen RC in base lines too as a method of softening the switch-on.
(softening the switch cf gate-current limiting in MOSFETs).
The 'test lead' effect may just be a ground thing, I can't really follow the photos.
But the basics of wiring micros with noisy things crops here a lot.
1. As said by Armp, have decoupling/bypassing caps on the PICAXE.
In my very humble opinion this should be considered manadatory.
C2: Stick 100nF Ceramic as close to PICAXE power pins as possible.
May need a //d electro too.
Without decoupling/bypassing your 'luck factor' increases.
2. Also some kind of 'reservoir' cap where the motor is.
This reduces dips/transients getting to PIC supply.
C1: 47uF upwards without going daft.
3. As far as possible , separate the grounds from Motor-supply, from PIC-supply. e.g. my additions to your schematic.
In general, remove any 'power sharing' of supplies.
Unless you are doing fast PWM consider an N MOSFET for the motor switching. I'm not a Darlington fan.
Attachments
techElder
Well-known member
IronJungle, yes, clean up the solder job. Watch a tutorial on soldering. You are using way too much solder, and I suspect that you have "cold" joints in there just waiting to cause problems (probably from adjacent blobs.)
What you really need are some "eyes" to look at the signals that you are producing. When erratic things are happening, I usually suspect that something is oscillating rather than just switching. An oscilloscope is essential. (I looked at your blog, and you have electronic projects going back for years. I guess you were just lucky enough to have never needed an oscope for one of those.)
As someone else has said, get some decoupling on the power supply. You might even decouple that "big" transistor, too. You don't need gain at high frequencies for this project.
What you really need are some "eyes" to look at the signals that you are producing. When erratic things are happening, I usually suspect that something is oscillating rather than just switching. An oscilloscope is essential. (I looked at your blog, and you have electronic projects going back for years. I guess you were just lucky enough to have never needed an oscope for one of those.)
As someone else has said, get some decoupling on the power supply. You might even decouple that "big" transistor, too. You don't need gain at high frequencies for this project.
IronJungle
Senior Member
As always, wanted to say thanks for those that are interested enough to follow this thread and offer help.
Armp wrote>>>The photos indicate you're not putting the test lead from emitter to GND<<<
The pic of the completed project is misleading as the black wire is not a power supply GND. I am, in effect, just adding 18" of wire to make the rig work.
Dippy wrote>>>>but what voltage is the motor rated?<<<
Stamped with 3-6V. It has a cap and a diode connected to the output terminals. (It came from one of those Sharper Image hand messagers)
Dippy wrote>>>C2: Stick 100nF Ceramic as close to PICAXE power pins as possible.<<<
Easy to do and makes a lot of sense. I will work that in. See my 'scope comment below.
Dippy wrote:>>>Also some kind of 'reservoir' cap where the motor is<<<
Dippy, I assume you mean cap in parallel with the motor?
Dippy wrote:>>> consider an N MOSFET for the motor switching. I'm not a Darlington fan.<<<
Me either (anymore). This is one of the best learnings I have gotten from the project. I need a few MOSFETs in my kit for switching. This is a change I will make regardless of if I get the current setup to work or not.
Texasclodhopper wrote:>>>An oscilloscope is essential. (I looked at your blog, and you have electronic projects going back for years. I guess you were just lucky enough to have never needed an oscope for one of those.)<<<
Hey, thanks for checking out my project page! I do have a 'scope and used it to put up some signals from the piezo in post #9 of this thread. I have put it on the transistor output as well. When the transistor is off, I do see a bit of 60Hz noise (see bypass cap reference above). Other than that, the signal is nice and squared when the motor turns on and off (with emphasis on the word *when*).
So... back to the lab to try a few things. I will probably swing by a Radio Shack and pay the outrageous $2 for a MOSFET. Oh, how I long for the days when the local Radio Shack catered to the hobbyist nerd in all of us....
Thanks again and I will post up the results.
Armp wrote>>>The photos indicate you're not putting the test lead from emitter to GND<<<
The pic of the completed project is misleading as the black wire is not a power supply GND. I am, in effect, just adding 18" of wire to make the rig work.
Dippy wrote>>>>but what voltage is the motor rated?<<<
Stamped with 3-6V. It has a cap and a diode connected to the output terminals. (It came from one of those Sharper Image hand messagers)
Dippy wrote>>>C2: Stick 100nF Ceramic as close to PICAXE power pins as possible.<<<
Easy to do and makes a lot of sense. I will work that in. See my 'scope comment below.
Dippy wrote:>>>Also some kind of 'reservoir' cap where the motor is<<<
Dippy, I assume you mean cap in parallel with the motor?
Dippy wrote:>>> consider an N MOSFET for the motor switching. I'm not a Darlington fan.<<<
Me either (anymore). This is one of the best learnings I have gotten from the project. I need a few MOSFETs in my kit for switching. This is a change I will make regardless of if I get the current setup to work or not.
Texasclodhopper wrote:>>>An oscilloscope is essential. (I looked at your blog, and you have electronic projects going back for years. I guess you were just lucky enough to have never needed an oscope for one of those.)<<<
Hey, thanks for checking out my project page! I do have a 'scope and used it to put up some signals from the piezo in post #9 of this thread. I have put it on the transistor output as well. When the transistor is off, I do see a bit of 60Hz noise (see bypass cap reference above). Other than that, the signal is nice and squared when the motor turns on and off (with emphasis on the word *when*).
So... back to the lab to try a few things. I will probably swing by a Radio Shack and pay the outrageous $2 for a MOSFET. Oh, how I long for the days when the local Radio Shack catered to the hobbyist nerd in all of us....
Thanks again and I will post up the results.
"Dippy, I assume you mean cap in parallel with the motor?"
Nope. 'Top' of motor down to ground. So it acts like a reservoir.
It reduces transients and noise going up power line , which would/could also get to your PICAXE.
Hence the comparison with a reservoir... however tenuous it may sound.
Check what I said:
"2. Also some kind of 'reservoir' cap where the motor is.
This reduces dips/transients getting to PIC supply.
C1: 47uF upwards without going daft."
If you looked at my additions to your sketch that I posted you would see where C1 is....
I also sketched it to suggest your physical wiring paths. Don't share Power and Ground where possible.
Nope. 'Top' of motor down to ground. So it acts like a reservoir.
It reduces transients and noise going up power line , which would/could also get to your PICAXE.
Hence the comparison with a reservoir... however tenuous it may sound.
Check what I said:
"2. Also some kind of 'reservoir' cap where the motor is.
This reduces dips/transients getting to PIC supply.
C1: 47uF upwards without going daft."
If you looked at my additions to your sketch that I posted you would see where C1 is....
I also sketched it to suggest your physical wiring paths. Don't share Power and Ground where possible.
IronJungle
Senior Member
Got it now, Dippy. Thanks and sorry for the oversight.
I added this cap to help smooth things out.
Just kidding; that's a pretty big cap... But, I did add the caps from +5VDC and GND as suggested. One close to the power input. One close to the motor. I had two 470uF handy and used those. You can see them in the pic below.
And Dippy.... All works fine now!. It is a perfect example of "of course that is the problem" once you solve the problem. Lessons learned: 1) Add a cap to help soak up the jolts and ripples, especially if you have a motor or power device and 2) if I would have used an N Channel MOSFET for a simple switch things *may* have been easier.
The rig is in the finished box and working. I will post up on the completed projects page after I get my documentation sorted out.
Thanks again to everyone that hung in there with me and provided comments/advise.
I added this cap to help smooth things out.
Just kidding; that's a pretty big cap... But, I did add the caps from +5VDC and GND as suggested. One close to the power input. One close to the motor. I had two 470uF handy and used those. You can see them in the pic below.
And Dippy.... All works fine now!. It is a perfect example of "of course that is the problem" once you solve the problem. Lessons learned: 1) Add a cap to help soak up the jolts and ripples, especially if you have a motor or power device and 2) if I would have used an N Channel MOSFET for a simple switch things *may* have been easier.
The rig is in the finished box and working. I will post up on the completed projects page after I get my documentation sorted out.
Thanks again to everyone that hung in there with me and provided comments/advise.
Last edited:
Thanks all for an interesting project and comments... !I would isolate the PIC power from the main power with a series diode to the PIC power lines, so that when the motor punches a hole in the main supply, the diode goes into reverse (effectively isolating the PIC power lines) and you can maintain the PIC supply with an e.g. 47uF capacitor, or ideally E = 1/2 CV^2 is the energy stored in the capacitor, so work out your storage needs to drive the transistor for n-time and m-current, then size the capacitor accordingly to match the demand with some spare capacity. The motor probably does not need a nice clean supply, so no need for dustbins.
IronJungle
Senior Member
I will post up in 'completed projects', but the final version is here:
http://ijprojects.blogspot.com/2011/12/echo-box-using-picaxe-08m-complete.html
Thanks again and get ready for stepper motor questions....
http://ijprojects.blogspot.com/2011/12/echo-box-using-picaxe-08m-complete.html
Thanks again and get ready for stepper motor questions....