Reply with QuoteOn the circuit board there is a 5k trimmer resistor, what is its purpose and what should it be set to? I looked through this whole thread trying to find information on it and can't seem to find any.
Bob W.
Hi,WHat happens when you hook up a pulley to it that has 40 teeth on it and the threaded rod pulley has 10? How would you configure that?
Here i give you a sample of how you can calculate the needed steps for Mach3.
Ns = 40 teeth
Nm = 10 teeth
spindle pitch = 5mm for example...
Us = 1 / 5mm (The way your axis is driving within one spindle turn)
Um = Us * (Ns/Nm)
for example
Um = 0.2 * (40/10) = 0,8 -> (Your motor has to do a 0,8 turn to move the axis for 1 mm)
Your encoder hast 500cpr. so it has a resolution of 2000.
2000 * 0,8 = 1600 steps/mm (thats the value for Mach3)
So thats the example for millimeters. same thing with inches
Yesterday i got my machine to runFirst milling tests were successful :rainfro:
You can watch it:
hehe
good night @all
Robert
My CNC-Diary (now in english, too):
www.wuselrob.de
On the circuit board there is a 5k trimmer resistor, what is its purpose and what should it be set to? I looked through this whole thread trying to find information on it and can't seem to find any.
Bob W.
Motor Maximum current adjustment. The diagram at http://www.embeddedtronics.com/ shows that the comparison voltage for this limit is approx 0.2326 times the voltage in the potentiometer cursor. The limit current set depends then on the amount of 0.1 ohm resistors in parallel in your board. (Limit current = 0.2326 x Voltage on the cursor / (equivalent sensor resistance value))Originally Posted by BobLWeiss
Thank Kruetz... although I understand most electronic theories and equations you lost me with that one. Let's say I have a servo motor that is rated at 15amps max... what would the setting be if you don't mind helping me out?
Bob
How many 0.1 ohm resistors are on your board? That is the unknown variable I need to know in order to tell you what voltage you need to adjust in the potentiometer cursor.
I used 5 per board.
Then you adjust a voltage equal to 0.086 times your maximum current (in Amperes) at the 5K potentiometer's cursor = 0.086 * 15 = 1.29 Volts.
On the previous post's equation:
Equivalent Sensor resistance = 0.02 Ohm ( 5 x 0.1 ohm resistors in parallel)
The 0.2326 factor is the division factor for a 1Kohm / 3.3 Kohm voltage divider connected to the potentiometer cursor = 1000 / (1000 + 3300)
DIY encoders
Is there any iformation avaialble on this subject ?
I 've got a box full of encoder toothed wheels and optical sensors .
Assuming I have a suitable DC motor , Do I just put such a weel on the motor shaft , and allign 2 optical sensors to have a 180° phase shift ?
It's easy to allign them with a scope connected to both channels .
Scope diagram of the two signals would be very helpfull .
Or , another way would be to use a gear reduction and have the encoder wheel on the motor shaft . Assuming a backlash free transmission .
With a redution of 5:1 and a 50 teeth weel , that would give a 250pulses / rev encoder .
Does that sound allright , or am I missing something .
Thx
Pat
Encoder pulses are supposed to be 90 degrees apart. Also don't forget that encoders read off the edges of the "teeth" or slot so that you actually see 4 pulses per "tooth". Your 5:1 reduction on a 50 tooth encoder actually comes out to 1000 pulses per revolution.
"Good judgement comes from experience, and a lot of that comes from bad judgment." - Will Rogers
I see , so they have to be 90° shifted not 180° .
Now I understand where the 4times the nr of pulses comes from .
Thank you , this is valuable information .
What would you advice me .
Put the encoder directly on the DC motor shaft , and have the large number of pulses , or put the encoder on the drive shaft to have maximum accuracy .
The reduction is made with a tooth belt and should be free of backlash .
The drive spindle on the x/y table on my mill has a displacement of 5mm/rev .
Still assuming thesame 5:1 reduction :
If I want 0,01 mm resolution , then its either a 125 slot encoder directly on the spindle . Or a 25 slot on the motor shaft .
Does that sound realistic ?
Pat
Most people will recommend that the encoder goes on the motor shaft. The main reason I see is that if there is ANY backlash it will manifest itself as servo hunting or vibration as the servo tries to go back and forth (usually at pretty high speed for a very short distance) to correct for the backlash. This can be very hard on the table mechanics, bearings, servo, driver... etc.
After reading this entire thread I've gathered that a 1000 pulse per rev encoder is in the range they recommend with this driver. But that lower count encoders will work. Will they work as well? I don't know without trying it. If you think you can get the backlash controlled then go with the 5:1 reduction as there will be more resolution and most likely finer control of the servo. If you end up with vibration go back to the lower count.
Or get your table setup and working with the lower count encoder on the shaft of the motor and cut yourself some higher count encoder wheels with cnc. You only need 250 holes for 1000 count.
"Good judgement comes from experience, and a lot of that comes from bad judgment." - Will Rogers
So I've started the uhu servo project .
Etched the boards today . Photgraphic method , etched wit FE(III)Cl .
As you can see , even a home brewn board can be made of very good quality . Price for these is close to nothing .
To bad I'm gonna have to build servo's before I can build a cnc drill .
So these are gonna be drilled manually . That's the hardest part .
Secondly , I'm gonna try to populate the boards with as much parts from the junk pcb box as I can .
Not that I mind the few $ the parts are gonna cost , but hey , this is supposed to be a fun project .
I need to order the fet driver chips and the transient absorber diodes per multiple of 25 .
Meaning , I'm gonna have a few spares .
So if anyone is interested , drop me a PM .
IR2184 : 4,5€ / 6$
P6KE12CA : 1€ / 1,6$
Shipping : only the stamp on the enveloppe
I've already ordered the parts , should arrive by the end of the week .
Pat
Added a picture of the first born
Btw , everything you see on this pcb comes from the spare parts boxes .
Missing only the fet drivers , the diodes and uli's chip .
Max. Voltage across R.16 per earler posts.
In reply to Kruetz's post on the max. voltage across R.16. On page #42 post #502 it would apear that the max voltage using 5 .1ohm power resistors at 15 amps the max. voltage would be 0.3 volt.
I.E. Rp= 5x .01 = 0.02 (BobLWeiss states he's using 5 power resistors)
I = 15 amps (per BobLWeiss)
So U = I x Rp = 15 x 0.02 = 0.7 Volt. ???????
Understand this is coming from a person who has smoked the output on six drive cards several, several times and am still going through hell trying to get it right. would you enlighten me/us
Thanks
Paul
Paul;In reply to Kreutz post on the max. voltage across R.16. On page #42 post #502 it would appear that the max voltage using 5 .1ohm power resistors at 15 amps the max. voltage would be 0.3 volt.
I.E. Rp= 5 X .01 = 0.02 (BobLWeiss states he's using 5 power resistors)
I = 15 amps (per BobLWeiss)
So U = I x Rp = 15 x 0.02 = 0.7 Volt. ???????
Understand this is coming from a person who has smoked the output on six drive cards several, several times and am still going through hell trying to get it right. would you enlighten me/us
Thanks
Paul
No, U = I x Rp =15 x 0.02 which is not = 0.7 Volt. !!!!!!,
15 x 0.02 = 0.3 volt
Look at the schematics, between the pot cursor and the comparator input there is a voltage divider, I am talking about the voltage at the potentiometer cursor (1.29 Volt), before the voltage divider, hence the difference in voltage, you are talking about the voltage needed at the input of the comparator (0.3 volt). after multiplying 1.29 times the divider factor (0.2326) the result is the same 0.3 volt at the reference input of the comparator..
Kreutz - Max. voltage ????
Kreutz,
OK, I think I'm beginning to understand... If you don't run out of patience, I may get a handle on this drive board yet !!!
Don't worry, I have made many mistakes while learning, the good thing is that I am not a doctor, most of my mistakes end up with fireworks, some noise and smoke, most of their mistakes end up underground without any noise....Kreutz,
OK, I think I'm beginning to understand... If you don't run out of patience, I may get a handle on this drive board yet !!!
BP Discovery -DIY servo drive card problems
Hello Kreutz & All,
After making the changes suggested by Kreutz, I have some improvement but yet again another problem. I've taken a picture of one of my boards that shows resistors R-24,R-25, capacitor D-7 & diode D7 after thy smoked when the servo power was turned on (134vdc with a surge of ? amps.). BTW the 12vdc is all ready applied to the drive card. the change that kreutz make for me on the snubber circuit seems to have stopped the FET's from frying.I'm showing a picture of a drive card with the damage to the above components and two fried FET's, but it was before the change in the snubber circuit. I just wanted a graphic example to show, so people could give me some suggestions.
A second area that I'd like some suggestions on is: Currently when you power up the mill, the power supply for the servo's is powered up and when you want to put power to the servo's a 24vdc circuit pulls a relay in completing the 134vdc circuit to the servo's. this is when I'm smoking my DIY dervo drive cards output section & now these additional components. Currently I'm only working with the X axis trying to keep the lost part count down !! I also have the belt off the X axis so if the servo did run away nothing mechanical will crash .... So, Two questions what would be a quick and safe way to reduce the 134vdc to say maybe 35vdc while trying to work through these problems ? or is there a better way to handle it. maybe somehow bringing the servo power supply up slower ? or later in the startup cycle ?
I really don't know and will welcome idea's suggestions, what ever.....
Thanks in advance
Paul A.K.A tenmetalman
I think you have a high dV/dt triggering the same branch mosfets due to miller effect.Hello Kreutz & All,
After making the changes suggested by Kreutz, I have some improvement but yet again another problem. I've taken a picture of one of my boards that shows resistors R-24,R-25, capacitor D-7 & diode D7 after thy smoked when the servo power was turned on (134vdc with a surge of ? amps.). BTW the 12vdc is all ready applied to the drive card. the change that kreutz make for me on the snubber circuit seems to have stopped the FET's from frying.I'm showing a picture of a drive card with the damage to the above components and two fried FET's, but it was before the change in the snubber circuit. I just wanted a graphic example to show, so people could give me some suggestions.
A second area that I'd like some suggestions on is: Currently when you power up the mill, the power supply for the servo's is powered up and when you want to put power to the servo's a 24vdc circuit pulls a relay in completing the 134vdc circuit to the servo's. this is when I'm smoking my DIY dervo drive cards output section & now these additional components. Currently I'm only working with the X axis trying to keep the lost part count down !! I also have the belt off the X axis so if the servo did run away nothing mechanical will crash .... So, Two questions what would be a quick and safe way to reduce the 134vdc to say maybe 35vdc while trying to work through these problems ? or is there a better way to handle it. maybe somehow bringing the servo power supply up slower ? or later in the startup cycle ?
I really don't know and will welcome idea's suggestions, what ever.....
Thanks in advance
Paul A.K.A tenmetalman
First, change the relay to energize the 134 volt DC power supply's transformer instead of applying DC so suddenly to the boards, that way the 134 Vdc voltage is going to build up gradually due to the power supply's capacitor charge. Second, replace All the 1N4007 diodes by HER105-T diodes (digikey part # HER105DICT-ND) the later are fast diodes and will help avoiding cross-conduction. 1N4007 diodes should not be used in the gate discharge or boostrap circuit, then replace U8 (should be damaged after the failure of R24 and R25. Recheck all the diodes and Mosfets before energizing the board again. Remember to place mica thermal pads between the Mosfets and the heat-sink, apply thermal grease to both sides of the mica thermal pads and make sure the screws are insulated from the Mosfet metal tabs and that those tabs are insulated from the heat-sink. I could not see any capacitor named D7......... on the schematics.
If you want to reduce the 134 Volt to a safer value, while testing, use a Variac between the power line and the 134 volt DC power supply transformer's primary, and reduce the voltage to the primary to a safer 40-50 volts instead of the 120 volts line voltage...
There should be no relay contacts between the 134Vdc power supply capacitors and the input to the UHU board.
BP Discovery / DIY Drivervboard sauga
Hello All & Thank You, Mr. Kreutz
Even without the Knowing what was happening, I was at the point of thinking something was going on that I couldn't figure out & wasn't presenting enough information for anyone else to help either. I had lost two motherboards to difficult, if not impossible to define reasons. One, I thought was due to a lighting storm. If this puts me on the road to having one more rung in the ladder behind me, & one rung closer to being able to program my mill, I'm going to be ONE HAPPY CAMPER ! I have some work coming in, this is my busy time of the year, but it's going to be real hard to let the mill set with something constructive to do. A off topic item, My wife & I have been doing some real work on trying to see if we have enough of a marriage left to salvage. It's hard work, especially for her. but, there may be hope. For you whom hadn't heard me talk of this I apologize, to you who expressed concern "Thanks"
This group has helped keep my mill project going at the times when I was wondering "What to do NOW"
Thanks again
Paul A.K.A. tenmetalman
Posting Permissions
