[BobLWeiss]

I just got a dc servo motor from ebay that was from a larger cnc machine. There is not model number on the motor but it has 3 cables comming out of it. I took apart the cover and 2 cables are for the brushes. There are 4 brushes on this motor and one cable (which has 2 seperate wires inside it) are on the front brushes which are larger and the other cable has 2 wires to the rear brushes. The other cable has 9 or so small 24 awg wires for the encoder which is a Renco model #77822-160. I was wondering if someone here can answer a couple questions I got?

1) How can I determine the encoder wires?

2) How do I hookup the brush wires to the UHU servo controller? Do I hook the front and back brushes in parallel or do I only hook up the front brushes by themselves? There is only 2 screw terminals on the board for the motor so it has me confused

Thanks for the help!
Bob

[smarbaga]

i do have 1 complete controller, well without the fets installed,
if kreuts would pm me his address i could send him one to test.
it will be 3 weeks before the new batch of boards arive.
with 4 brush motors you would connect the 2 fields in parallel at the motor terminals.
this would save wire and less chance of noise to be induced.
this is what i have seen on other machines,
i have not seen any connected in series, yet.
i guess this is the reason for connecting them at the motor.

[MrWild]

But is it a better option?

http://www.skyko.com/products/

A step and direction interface for anolog drivers.

[LZ1TWB]

Hello jlcortes,
Reading your reply have reminded me of my problem here I have with the controller. It seems I cannot drive it more than something like 75000 steps without having errors counted. Below this "limit" I cannot make even one. Increasing the speed leads to more advance increasing of the error. I think I have done everything right, good PS, schielded diferential encoder lines and so on, no change. I don't know what to do. I watched the 5V with scope and can see some spikes which are quite visible when I apply torque at the motor and the current is higher. I tryed with different capacitors on the voltage lines, still no change. I don't know if they are obvious on every drive and are normal or can cause problems?

Thanks for any suggestions,
Todor [email protected]

[paulC]

But is it a better option?

http://www.skyko.com/products/

A step and direction interface for anolog drivers.

Refers to a motor amplifier. There are no high current devices on board so this board needs an additional board to power the motors. It is only part of a solution.
Paul

[jlcortes]

My SMD version of Uhu board fully tested and working!!

This controller is incredible, i was using a controller called SV-500, but i had force problems with this controller, my uhu version have a lot of more torque with the same motor!!!



Jlcortex

[BobLWeiss]

I am going to be testing out the controller later today and wasn't sure if I need to supply the STEP/DIRECTION signals to the board to make sure its working. Obviously I will be doing that in the future but just to see if it works ok with the encoder and motor drive circuits I wasn't sure if I needed to do this. I can communicate with the board via the PC just fine but now its time to hook a motor up to it.

Thanks,
Bob

[smarbaga]

you can run the terminal program, set the parameters and make your motors go. maybe start out with a lo voltage first ( like 12v) just incase something on the output section in incorrect.
set the encoder error rate to its max ( 5000 ) as errors will accumulate, not going through the step/direction terminals.

[L98FIERO]

I was looking around on the forum and came across this http://www.cnczone.com/modules.php?n...threadid=35212 Has anyone attempted to auto-tune the UHU board like Dubble is doing with his? Can it be done?

LaVerne

[smarbaga]

LaVerne
to have a self tunning feature
one would have to get into the drive controller
software and integrate this feature monitoring and
adjusting the current against the various parameter changes.
i think ???

[MrWild]

I have some hytachi servo motors with magnetic encoders. They are 500 PPR and have six wires. Can they be hooked up to the UHU boards and if so, how are they wired in? These are for experimenting with the process of wiring up a BOB, computer communication, and EMC2. After I understand the basics, the drives and BOB will be used to refit my Crusader M and photo encoders.

[ulihuber]

Hi MrWild,

these Hitachi Motors are small but good. I use them for testing.
If you open the case of the magnetic encoder, you'll see easily what the supply voltage pins are and where the signal lines are located.
The motors seem to have different cables in different applications, be carefull and don't make connections identical without double checking.

@LaVerne: I made some Tests with autotuning and found that it's not worth the effort. It brings by no means the same results as an intelligent and sensible person could do. The UHU will run in most applications without any tuning, just using the preconfigured setup. Getting it better than that needs to include the controller as well (ramps...), different torque situations and impulses. Mutch too much für a tuning function in the servo loop controller.

About the tracking error problem a few days ago. There is no known software problem which could cause such an effect. The controller does not have any loops which can longer than sub-second. The longest one would be a blocked motor with very small error and up-counting integral error.
So I assume that the setup is not free of electrical problems. 75.000 steps in not really a big deal. You should have millions without losing a single one.
Check the encoder error counter. If it's counting up, you really need to take care of your encoder lines.
Check the output from the parallel port. If it's a laptop, you'll probably not have more than 3.3V. In this case, reduce the resistor at the opto coupler input.

Best regards
Uli

[FredCNC]

Does anyone have another type of file format other than the PLT and PDF?
This is because the company that does the PCBs for me (and his service has unbelievable quality and cheap cheap prices), doesn't want to do from a PDF, because it would compromise quality, and the PLT conversion with the program that i found that does that (Gerbview) superposes the fine tracks in the middle, making the drawing useless. I asked if he could do directly from the PLT, but he doesn't know how to...

Any sugestion on previous experience or programs to do the PLT conversion, it would be greatly apreciated.

[tenmetalman]

Hello All,
I finely got to put ten hours in on my mill project. I'm currently limiting my efforts to the X axis. my hope is to iron out most problems on this axis and the Y & Z will go much faster I'm using the orginal servos & smarbaga UHU drive cards. When I start the mill the X axis drive comes on along with the PC4CNC Multifunction card getting the 12 & 5 vdc on the machine side. when the PC is turned on the second 5vdc for the computor side is supplied to the breakout board and the pump comes on activating the output on the breakout board. At this point I have a toggle switch that turns on a relay to pull in the contactor that activates the circuit for the power supply for all the axis drives. when I switch on the toggle switch the X axis servo turns on max rpm and most of the time the 10 amp fuse on the servo power supply blows..... Even when it doesen't, the servo will slow to a stop from max rpm after 5/10 seconds and will not respond to any requests for movement, from the mach3 program or the UHU terminal program. I have the belt off the servo so the table isn't going nuts. I've had to change the 1.8kohm 1watt resistor on the UHU drive to a 1.8kohm 5 watt resistor to keep it from going up in smoke. anyone got any ideas ????

[contactirfu]

Hello All,
I finely got to put ten hours in on my mill project. I'm currently limiting my efforts to the X axis. my hope is to iron out most problems on this axis and the Y & Z will go much faster I'm using the orginal servos & smarbaga UHU drive cards. When I start the mill the X axis drive comes on along with the PC4CNC Multifunction card getting the 12 & 5 vdc on the machine side. when the PC is turned on the second 5vdc for the computor side is supplied to the breakout board and the pump comes on activating the output on the breakout board. At this point I have a toggle switch that turns on a relay to pull in the contactor that activates the circuit for the power supply for all the axis drives. when I switch on the toggle switch the X axis servo turns on max rpm and most of the time the 10 amp fuse on the servo power supply blows..... Even when it doesen't, the servo will slow to a stop from max rpm after 5/10 seconds and will not respond to any requests for movement, from the mach3 program or the UHU terminal program. I have the belt off the servo so the table isn't going nuts. I've had to change the 1.8kohm 1watt resistor on the UHU drive to a 1.8kohm 5 watt resistor to keep it from going up in smoke. anyone got any ideas ????

Reverse the polarity of the motor power supply near the motor, I had a similar problem and kreutz and smarbaga and others had give me this solution a few threads back...........BTW what is the rating on your PS and whats the voltage and current ratings on your motors.

I sent a mail to you earlier, kindly respond to that email.

Regards
IRfanulla

[wuselrob]

Hi,

I had the same problem with one of my axis. Like contactirfu said, simply change the polarity of the motor on your UHU.
Try to double the resistors value to maybe 4,7kOhm / 5W and the capacitor beside it to the half value. Then the resistor will not be glowing anymore

Regards,
Robert

[kreutz]

Hi,

I had the same problem with one of my axis. Like contactirfu said, simply change the polarity of the motor on your UHU.
Try to double the resistors value to maybe 4,7kOhm / 5W and the capacitor beside it to the half value. Then the resistor will not be glowing anymore

Regards,
Robert

While you might control the heating problem with this solution, you are completely defeating that RC circuit purpose, and may as well take it out. The purpose of that circuit is protect the Mosfets from transient voltages created during the switching of the high current on the inductance. The might be a problem with non proper components used on that circuit, a higher than normal motor inductance, brush arching, or a plain design OOps, The only way to safely do a change is checking what transients are there and redesigning the snubber to properly absorb and dissipate the energy without overheating.

[contactirfu]

Hi,

I had the same problem with one of my axis. Like contactirfu said, simply change the polarity of the motor on your UHU.
Try to double the resistors value to maybe 4,7kOhm / 5W and the capacitor beside it to the half value. Then the resistor will not be glowing anymore

Regards,
Robert

HEllo Robert,

That's a nice update on your website........the new controllerbox is excellent.........how did u get your front panels made.........is it decals or engraving on metal boards.......

Kreutz did u get the boards from smarbaga to redesign the rc circuit?

Thankx and regards
IRfan

[kreutz]

Kreutz did u get the boards from smarbaga to redesign the rc circuit?

Thankx and regards
IRfan

Not yet, he is waiting for parts in order to finish the boards. On the other hand, it is good he didn't send them yet because I have been too busy at my job lately. Hopefully I will be back to the normal workload within a couple of weeks.

[ejkoeze]

While you might control the heating problem with this solution, you are completely defeating that RC circuit purpose, and may as well take it out. The purpose of that circuit is protect the Mosfets from transient voltages created during the switching of the high current on the inductance. The might be a problem with non proper components used on that circuit, a higher than normal motor inductance, brush arching, or a plain design OOps, The only way to safely do a change is checking what transients are there and redesigning the snubber to properly absorb and dissipate the energy without overheating.

This is getting interesting. A post from Stefan8051 from Januari on cncecke.de shows that he does not exactly know how to calculate the snubber net (http://5128.rapidforum.com/topic=125...119&startid=3). He more or less states to go by feeling and says that when R gets too hot the only thing that works is increase the value of or or lower the value of the C
The examples he gives are 10n and 6,8K for a 90V system.

Looking forward to whatever this may bring.

Erik Jan