[dakimber]

How many seconds since the start of movement?

Kreutz.

It seems to happen right away - probably less than one second.

Doug

[kreutz]

It seems to happen right away - probably less than one second.

Doug

Have you made any tests with a lower voltage motor power supply? What are the 1.5 uF capacitors type? (ceramic, metalized polypropylene film, electrolytic, tantalum, etc)

Have you checked (with a scope) the output of the opto couplers after loss of control?

It looks like your bootstrap charge circuit is unable to hold the charge for a long time, your decision of not turning on the lower Mosfets at the same time, during PWM off, is affecting the upper Mosfets gate drive circuit by not giving the bootstraps circuit a low impedance circuit to recharge periodically.

Using the right capacitor type is also very important on the bootstrap circuit.
I also recommend using a couple of 1/2 Watt carbon composition 100K resistors from the motor leads to ground respectively, by doing this you will avoid having the upper mosfets' source voltage to float due to the high impedance of the open lower MosfetsMosfets which will also avoid locking up the upper Mosfet drivers at startup, or after a period of inactivity.

What is your PWM frequency?

[dakimber]

Have you made any tests with a lower voltage motor power supply? What are the 1.5 uF capacitors type? (ceramic, metalized polypropylene film, electrolytic, tantalum, etc)?

I tried a 24V supply on the workbench and never lost communication with the Pluto, despite using a laptop in that config. It did exhibit the 50% problem, though.

The caps are metalized polyester film, Digi-Key part EF2155-ND: http://www.panasonic.com/industrial/...bd0000ce23.pdf
I calculated the minimum value according to the International Rectified app notes and got something like 0.05uF (I don't remember exactly), so I thought I was being conservative with the 1.5uF. Maybe I wasn't conservative enough or just mis-calculated.

Have you checked (with a scope) the output of the opto couplers after loss of control? )

It isn't totally consistent, but mostly one opto is stuck high and the other low. This should be driving the motor, but it doesn't (the motor does typically "jump", then stop).

It looks like your bootstrap charge circuit is unable to hold the charge for a long time, your decision of not turning on the lower Mosfets at the same time, during PWM off, is affecting the upper Mosfets gate drive circuit by not giving the bootstraps circuit a low impedance circuit to recharge periodically.

My decision not to drive the lower pair ON during PWM off is looking worse and worse....

Using the right capacitor type is also very important on the bootstrap circuit.
I also recommend using a couple of 1/2 Watt carbon composition 100K resistors from the motor leads to ground respectively, by doing this you will avoid having the upper mosfets' source voltage to float due to the high impedance of the open lower MosfetsMosfets which will also avoid locking up the upper Mosfet drivers at startup, or after a period of inactivity.

What type of capacitors do you recommend? And how do you calculate the proper value?

I added the 100K resistors but didn't see a significant change in circuit behavior.

What is your PWM frequency?

19.5 Khz - the default value for the Pluto-servo HAL component. I haven't changed any of the FPGA programming or the Pluto-servo component.

I can see how the H-bridge could cause the servo not to work correctly, but I'm still missing the connection between that and why I lose communication with the Pluto-P. Possibly some type of noise feedback corrupting the FPGA?
After this happens the encoder values don't change when I turn a servo.

Thanks for all the help!

Doug

[kreutz]

I tried a 24V supply on the workbench and never lost communication with the Pluto, despite using a laptop in that config. It did exhibit the 50% problem, though.......................
Doug

Please, PM me with your e-mail address. The capacitors are the right type. There are a few hints in your anwers that might help solve the problem.

What are the minimum and maximum duty cycles? Can you show us the oscillogram of the signals driving the opto-coupler (with the 24v power supply)?

Thanks,

Kreutz.

[epineh]

From memory the PWM signal is inverted for the op amp to what a H-Bridge would use, I had heard that Lawrence had made a MOSFET version of the code, but I haven't seen it, and I may have not heard correctly...

Do some searching for Max-Mod's version of the DSPic, I believe his version is suitable for H-Bridge use, in which case Samco's Bridge should be a great match. It also has some tuning software that would be very handy, it may be in French though...

Cheers.

Russell.

[epineh]

No Prob Kostas, sorry I don't have a link for you, I had one to the forum that contained the code and more details but I cannot find it right now, if I do I will post it straight away.

Cheers.

Russell.

[jalessi]

http://www.max-mod-shop.com/index.php

[mojalovaa]

Forgive the stupid question maybe, but I'm a beginner in this business, whether your servo controller has the ability to connect it to the encoder, or is this just another driver, so you have some control electronics.
I have a BLDC motor from a battery drill 24V 2.6A and I would like him to make a DC servo motor so I wonder whether there is some kind of web pages on that topic for a complete beginner or if you have some sort of scheme to send me a private post, or that they have laid him here that all beginners can learn something.
Regards

[dakimber]

mojalovaa-

The servo control basically has 3 parts (not counting the motor and encoder). First, there is a personal computer running Ubuntu Linux that runs EMC2, the brains of the system. The PC is connected to a Pluto-P board through the parallel port. The Pluto-P is an FPGA board that breaks out a number of the control signals and handles some of the real-time intensive things such as counting encoder pulses. The Pluto-P also provides the control signals that drive the H-bridge. This thread was originally started because I was having trouble getting my H-bridge to work. Thanks to all the help I got from other forum members the H-bridge (and everything else, for that matter) has been running my CNC mill successfully for several years now.

I wasn't sure from your post exactly what you intended to do with your servo. If you want to control it from a PC take a look at EMC2 (which apparently has been re-named to LinuxCNC since I last looked!):

LinuxCNC.org

I hope this helps!

Doug

[mojalovaa]

I want to make a servo controller for 48V and 10A max. electricity, but whatever scheme I find a microcontroller and since I do not have the source code, and I do not know a whole lot of programming, I wanted someone sends me a simpler scheme, or if you can page from which one can download the source code, if it is because of him Atmel I have a programmer but if not I'll make for Pic
Is someone trying to make the car Wiper wiper, since their motor servo motor.
I'm sorry my English is very bad, so I use google translator, but I see that he is not even that much better than me.