Speed

           I put at least one hundred hours into seeing how fast a stepper motor can be made to spin while using cheap parts. Using a 12 volt computer supply that has a 5 ohm 10 watt load resistor on the 5 volt line will give you 11.5 volts to work with. Using my Piker board I was getting 240 rpm or 12" per minute with a 1/4-20 threaded rod. I was happy with that results and I sold the board as such.  I recommended  4 to 6 volt motors at 1 to 1.2 amps. The 4 and 5 volt motors spin a little faster if you fine tune the value of the large dropping resistors. At that time I suggested 8 ohm 20 watt Radio Shack resistors found in the audio section.

           At some point I stopped selling the Piker boards and I started listing other people's boards. Several people who bought a Piker and  then went on to buy a second controller from someone else emailed me with complaints.  I started getting low speed complaints and Emails from people who bought a more expensive system only to get a little  more speed than my Piker! Steve Manzer did not get great results with my Piker and his 1.5 amp motors. All this was sloshing around in my head!

            I contacted the other guys! Only Dave at HobbyCNC answered me honestly!  The others gave me the info and specs off their webpage. What would you do at this point?  I went back and asked the others how they determine their speed and how I tested my system.  Only Dave at HobbyCNC answered that Email!

            I decided to drop the whole thing! But the very next day someone finished a 7th Sojourn. He was running it with someone elses controller at a speedy 8" per minute. Here we go again!  I gave the person every hint I had in order to get him some more speed. Nothing worked!

             I then went on to try to simulate the problem here. At the same time I bought myself a 5 volt / 24 volt supply.  I added some resistors and my motors were spinning up a storm. The 5 volt 1 amp motors were doing 440 rpms or 22" (1/4-20 rod). I was happy with that. The 6 volt 1.2 amp motors were only spinning at 360 rpms before they would start buzzing. This was the same speed someone else was getting with their controller and motors using the same voltage.  I reasoned that larger motors don't spin fast because the rotor is larger. I was wrong!

              While simulating someone elses problem and trying to find a solution, I hit the mother load.  I found several controller designs that use four series output diodes. One diode in each output phase.  I had suggested this to people several times in the past. It never seemed to help their problem so I dropped the idea at that time. Maybe they forgot to increase the speed in the software after adding the diodes?  Maybe their controller has other design issues.

               So I added four power diodes to one of my motor output on my Piker. I then slowly increased the speed in the software. The 6 volt 1.2 amp motor goes from 360 rpm up to 480 rpms with plenty of torq! What the hell was going on! At first I thought the spikes from the motor were getting into the logic circuit. So I did some more reading! Please remember, I never wanted to sell controllers! Just by luck I picked a really old great design!

                After doing lots of reading, I found the answer! The more torq a motor has the larger the spikes the motor makes as the fields coils turn on and off. These spikes travel through the protection diodes in the circuit to ground.  The diodes are there to protect the output devices but they cause the motor to brake as the current dissapates to ground. If you use a higher voltage with your controller with a larger motor it will spin faster. The spike gets bigger so larger motor does not do as well as the smaller low current low torq motor. The larger the current the larger the spike the more braking.

                By adding the series diodes you eliminate the path to ground for the spikes while protecting the output devices at the same time. The spikes dissapate in the motor as a magnetic field rather than a flow of current to ground.  I also found you can put a single diode in series with each dropping resistor instead of  four output diodes. I tried this with a FET circuit and the outputs blew.  Somehow the spikes went around and through the motor to kill the FETs.

                A UniPolar controller running at 24 volts can blow the doors off a BiPolar controller running at 16 volts and at half the price!!! The reason is the output diodes. BiPolar circuits have 8 "Speed Bump" diodes. Just because a circuit is BiPolar it does not make it better or faster.

Speed 2



               While working on the motor speed problem I was also discussing leadscrew pitch with other people on the Hardware Store CNC conference.  They call it "pitch" I call it turns per inch. 1/4-20 threaded-rod has 20 turns per inch.  So you have to turn the threaded-rod  20 times to get the drive nut to move one crummy inch!  But that one crummy inch gives you a big torq advantage. The 1/4-20 can be attached to the 1/4" motor shaft with some rubber auto vacuum line. The friction of the nut to the motor riding on the threaded rod under load is next to nothing. To make a nice small CNC system you only need to add a little SPEED!  This is why I want all controllers to work as fast as my Piker!

                But the "what if"  happens! The gang starts turning down larger 10 turn per inch ACME rod so it fits the 1/4" shaft motors. Ten turns per inch is faster than 20 turns per inch. YES! this is true.  But, and that is a big BUT!  You have to turn down the rod down on a lathe or have someone do it. You also need to find or make a tap in order to make your own drive nut. Then you find out you have to drop the speed of your motors back if you have lower torq motors! Why?

                 A screwdriver handle is bigger than the shaft. A doorknob is larger than the axel it turns. The larger diameter to smaller diameter creates mechanical advantage. Turning a 1/2" leadscrew with a 1/4" motor has a mechanical disadvantage. You lose torq as you gain speed. Several more problems popup.  Everything is fine until you add a load.  The long ramp wound around the 1/4-20 rod is now only half as long and twice as high. The friction between the nut and the leadscrew increases because the amount of surfaces touching increases.

                 If you want more speed, get out your wallet! Every little gain of speed or torq will mean big bucks. A small machine can be built real cheap. Once you start thinking a little bigger all hell will break loose.
 

                                                                                                  Regards

                                                                                                                         John Conrad Kleinbauer
 
 

Home Info Plans     Order     Links Pics