My 555 PWM dosen't work

Thread Starter

Scott_M

Joined May 8, 2008
37
John said:
Are there jumpers we haven't discussed? ?
No others what about IC 1 pin 2 to 6 and the current connection to pin 7 ? I could cut the trace easily enough but will need a jumper from 2-6


and a diode soldered across a resistor. You can just solder it to the leads of the resistor.
I am assuming this is the R4 1K resistor ? It is flush mounted so it will be on the back.


As for the 6V or 12V question, remember with PWM at low frequency and with a small motor, the motor will be seeing 12V,
Yes I do realize the PWM delivers 12V in Pulses , I should have said 50% instead of 6V
 

jpanhalt

Joined Jan 18, 2008
11,087
No others what about IC 1 pin 2 to 6 and the current connection to pin 7 ? I could cut the trace easily enough but will need a jumper from 2-6
Absolutely, you need to do that. I thought the earlier post was just a typo regarding the pin7 number.

Scott said:
I am assuming this is the R4 1K resistor ? It is flush mounted so it will be on the back.
Yes, R4 is correct. On the back is fine or, since the 1N4148has fine leads, you might be able to slip them under the leads on the component side and solder. Resistors are pretty tolerant to heat. Double check the orientation, though. It is easy to mix up, particularly if you do it on the back.

John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
OK
I removed the trim pot and replaced it with a jumper
I added a diode piggyback on R4 with the band end towards pin 7 on IC1
I broke the trace between pin 2 -7 on IC 1
I added a jumper from pin 2 to 6 on IC 1

I made an LED test probe with a 100ohm resistor on the anode side and tacked it to pin 3 IC 2 and left the ground end loose for now.

With power applied I have 7.5V at pin 3 IC 1 and pin 2 of IC 2
and nothing at Pin 3 IC 2

Also, this is with the pot at 100%. If I turn it down even a little it drops to .38V

Are you sure about that 100ohm resistor on the LED ? It will not light even when connected to the power in. I have double checked it with diode test function on the DVM. It is still good.
 

Thread Starter

Scott_M

Joined May 8, 2008
37
Whoops
I take it back...the LED was bad. Replaced . Now it will glow if the pot is turned up all the way. If you turn the pot down even a little it goes out.
 

jpanhalt

Joined Jan 18, 2008
11,087
OK, it sounds like IC1 may be working. I didn't check any voltages as I was using a scope. I might have expected the voltage at IC1 pin3 to read a little higher, as the duty cycle was very high, but let's ignore that for now.

To confirm:
1) You still have the pull-up resistor between IC1 and IC2, right?

Now for the LED test probe. I assumed your LED was a 10 mA device (extremely common). With a 12V supply, and subtracting the diode drop (1.6V), then you need to drop roughly 10 volts at 10 mA. I made a dumb math error, multiplied 0.01A by 100 ohm and got 10V (wrong!). The resistor needs to be 1K. Thanks for pointing that error out so politely. The free end can go to ground or to the positive supply (555's sink or source ). I would connect the anode end to the positive supply and cathode to pin 3. Then, you don't have to worry about whether you need a pull-up resistor.

The idea behind the probe is simply to put a little load on the output so you can see if it is changing. The voltmeter is virtually no load.

What may be happening is the speed pot is going from a high duty cycle to a low duty cycle very quickly. I used a 100K pot in my set-up this morning, and it was still pretty sensitive to the pot position. Do you have something smaller, like maybe a 50 or 10 K pot?

John
 

jpanhalt

Joined Jan 18, 2008
11,087
We cross posted again. Leave the LED the way you have it and try to find a smaller pot. You might increase the resistance for the led, if you want, but it seemed to be working. John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
While trying to remove the 250K pot I damaged the board ( broke the traces ) ( 2 more jumpers)
I installed the 50K pot. I have a little more "Variability" but not much. Should I go to the 10K pot?
 

jpanhalt

Joined Jan 18, 2008
11,087
Trying the 10K is worth a go. No damage will happen. IC1 creates the timing pulse; IC2 controls the pulse width. Making it too narrow won't ruin anything.

If that doesn't work, how close are you to I-70? It probably wouldn't take much time to hop down there with a scope and find out what is really happening. In return, I would like to see a demo of the Tormach.

If that sounds practical, we should probably continue that part of the discussion by PM.

As an alternative, I am sure the design in the schematic will work, as that is what I built. There may be some unrecognized bug in the PCB from something like a wire that Eagle didn't recognize should be connected. So, another possibility would be just to redo the board from a schematic that passed the ERC.

John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
I am working on the original schematic on a bread board now.

I-70 ? like Columbus -Dayton I-70 ? or did you mean I-71 ?

I am a bit north of route 18 and just west of I-71 ( about 3 minutes from exit 222 on I-71)

I am always happy to show off the Tormach.
 

Thread Starter

Scott_M

Joined May 8, 2008
37
Progress

I re-did the original schematic on the bread board and it works ! :)

With the trim pot and the 220K speed pot. I put a 80mm computer fan on the motor side and it will turn down to barley moving and restart at that setting. But I don't think it is getting to full RPM at 100% , close....

Also the speed pot acts kind of funny. 0 to about 70 % is nice and linear then it starts slowing down over 70% :confused: But we are much closer.

edit---I played with the trim pot and adjusted it so I got max rpm at 100% and it is a linear move from 0-100% :D The DVM reads 2.5V when the fan stops at about 5% and 10.7V at 100% ( Close enough ) I am confused about why the DVM dosen't read 12V all the time, Is it just that the volt meter cant read it fast enough ?

With the damage to the other board , I really want to scrap it and start over.

What changes to the original schematic would you suggest ? Maybe this new board will be the Tormach demonstration you would like to see.

Once again... I can't begin to thank you enough for all your help. It is greatly appreciated !
 

jpanhalt

Joined Jan 18, 2008
11,087
Scott said:
edit---I played with the trim pot and adjusted it so I got max rpm at 100% and it is a linear move from 0-100% The DVM reads 2.5V when the fan stops at about 5% and 10.7V at 100% ( Close enough ) I am confused about why the DVM dosen't read 12V all the time, Is it just that the volt meter cant read it fast enough ?
Great news! The answer to the DVM is yes. It essentially samples over a period and reports the average.

The schematic is working with the mods already discussed, so I would leave it alone. As a design element, you need to decide whether you want both PWM and frequency control. I don't see a need for frequency control (the pot on IC1), so I would leave it off. But, that is just a preference.

Motor windings will make noise from the PWM, so you might want to change the frequency if the 1KHz bothers you.

My comment about using the full power of schematic capture to go from schematic to board still holds. Today, I would never think of routing a board without doing ERC. You may never get rid of all of the errors and warnings. In fact, some may be intentional, but you will detect the ones you didn't know about.

I will be happy to walk you through it, if you want. It is pretty intuitive.

John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
I have redrawn the schematic and board in Eagle and it has passed ERC with no errors but DRC had 1 error that I am not concerned about. It says the hole for the TIP31 is to close to something. In reality it will be vertical not horizontal, so it is not an issue. I just have to add some pads for power in and out.

John
When I had the brushless computer fan hooked up I had to set the trim pot pretty close to the 500K end of its adjustment. I then tried a larger DC motor with brushes and it would not run until I turned the trim pot way down to the low end of its adjustment. So I think It will help "Tune" it for the specific application. Once I get it set for the actual motor it has to control I will take a measurment and replace it with a fixed value resistor.

Also With the larger motor on it The Tranistor got pretty darn hot ! I did not have a heatsink on it on the breadboard. And after a minute or two I could see that the holes in the breadboard under the transistor were starting to get bigger ( melting )

Digi Dave

Simple ??? why go simple when you can overcomplicate things ! ;)

Actually I read somewhere that the single 555 PWM's were fine for computer fans and small no load devices but the 2 timer circut worked better for low speed control of heavier load devices. I am not sure if that is true....But I read it on the internet so it must be true. :D


 

Attachments

jpanhalt

Joined Jan 18, 2008
11,087
That new schematic looks better. I think I would keep the diode across R4, so you don't have to worry about the pulse width from IC1 affecting the re-triggering of IC2. You may also want to include pads for the pull-up resistor between IC1 and IC2. If you decide later it is not needed, you can just leave the pads open. It feels good, doesn't it, when the ERC reports back no errors?

You may want to take a look at 4QD's site:

http://www.4qdtec.com/
http://www.4qd.co.uk/

There is lots of good information there. Most telling is the comment (1998, probably still there somewhere, but I didn't look) that every type of motor is different and puts different demands on the controller. That is, even with DC brushed motors, series wound, shunt wound, and PM have different needs. Thus, I am not at all surprised with your most recent results.

The computer fan motor, I believe, is DC, brushless, and the PWM control is generated by a PCB circuit inside the fan case itself. Driving that circuit with a PWM seems a lot different than driving a brushed, PM, DC motor directly.

Now that you have or will soon have a working PWM generator, it is time to consider the output. Right now, you have it configured as a high-side controller, no "main" capacitor (see: 4QD for that term) and a relatively puny diode across just the motor. That will work for small hobby motors, but as you increase the power needs, you will find it doesn't. I had a vivid demonstration of that about 10 years ago when I scaled up from a brushed, PM, 100W hobby motor to a series wound, 2 KW motor at 12V (more than 200 A starting). :eek:

At this point, it is probably a good idea to narrow down the type of motor and power class it will be.

I would seriously consider making the controlling transistor or mosfet low-side, getting a larger diode across the motor and adding a capacitor. 4QD, and I suspect a great number of other sources, will describe better what I am trying to describe.

Best regards, John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
I looked at the 4qd site and most of it is wayyyy over my head. I will hook my circuit up to the actual motor and get some current readings and see if performance is acceptable.
Is the 1N4148 diode sufficent for the R4 resistor ?
And what diode would you recomend for the motor ?
 

jpanhalt

Joined Jan 18, 2008
11,087
I looked at the 4qd site and most of it is wayyyy over my head. I will hook my circuit up to the actual motor and get some current readings and see if performance is acceptable.
You don't have to understand everything there. Remember, he makes DC motor controllers for a living. It's a good reference for some of the terms, such as main capacitor.

Is the 1N4148 diode sufficient for the R4 resistor ?
Yes.
And what diode would you recommend for the motor ?
Depends on the size of the motor. You might consider something in the 1N400x series for motors up to a few hundred watts. Remember, the motor gives off a lot of inductive kick when it is turned off by the PWM. I don't think the diode has to match exactly the on current of the motor, because the currents are pulsed, but you do need to size up accordingly in going from something less than an amp to 10 or more amps.

Finally, don't forget what 4QD calls the main capacitor. It doesn't have to be huge, but it sure made a difference in the controller I mentioned previously. I used 2400uF for that 2KW drive.

John
 

Thread Starter

Scott_M

Joined May 8, 2008
37
The reference to "Main Capacitor" must be in the members only sectio. I could not find it.
I will put a 1N4004 in for the motor diode.

Gonna go test some loads.
 
Top