H-Bridge for comment - Page 4


Page 4 of 6 FirstFirst 123456 LastLast
Results 61 to 80 of 115

Thread: H-Bridge for comment

  1. #61
    Member
    Join Date
    Feb 2007
    Location
    canada
    Posts
    962
    Downloads
    0
    Uploads
    0

    Default

    I use 2104's on all my drives, but lately i the price has almost doubled.



  2. #62
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by acondit View Post
    I edited post #52 to replace the zip file with an upgraded schematic. I fixed the connections on the capacitors and revamped the connections a little (removed the -18V connection and hooked two pins to ground).

    Alan
    Hello;

    How do you plan to drive the output stage?

    Sign-magnitude?, Locked Anti-phase?

    If you employ the former you might have to add a charge pump to help recharge the bootstrap capacitors when the motor is running in one direction for a long time.

    If you employ the latter, then you don't need those high capacity caps on the bootstrap circuit (depending on the PWM frequency).

    Regards,

    Kreutz.



  3. #63
    Gold Member acondit's Avatar
    Join Date
    Apr 2005
    Location
    USA
    Posts
    1778
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by kreutz View Post
    Hello;

    How do you plan to drive the output stage?

    Sign-magnitude?, Locked Anti-phase?

    If you employ the former you might have to add a charge pump to help recharge the bootstrap capacitors when the motor is running in one direction for a long time.

    If you employ the latter, then you don't need those high capacity caps on the bootstrap circuit (depending on the PWM frequency).

    Regards,

    Kreutz.
    I don't know the answer to that yet. I was planning on using the Pluto-P that Samco was using and that Russell blew up. I don't know what its output is. Maybe Samco or Russell can jump in and enlighten me.

    I really just wanted to try and provide current limiting that was missing from this design (sort of to see if I could help). I have some other projects to finish before I actually do this pcb.

    Alan



  4. #64
    Registered
    Join Date
    Mar 2005
    Location
    germany
    Posts
    27
    Downloads
    0
    Uploads
    0

    Default Current sensing methods

    Hi guys,

    I have some questions about the current sensing section, when we used here so small sensing resistors 0.02 Ohm were it not much better to have an op amp which multiplies the measured current let say about factor 5 or 10 or more? So the voltage level could be near 3Volts.

    Which op amp should be used for such multiplier? Is the tlv2772 good enough to do that?

    I read in the tlv2772 datasheet that this device is obviously an op amp and not a comparator why did we use here an op amp instead of a true comparator like e.g. LMV7219? (7 nsec, 2.7V to 5V Comparator with Rail-to-Rail Output)

    The schematic is not able to sense fast/slow decay. Fast decay currents will be hot-wired in the shottky diode D9.

    If we connect on both legs of the bottom FETs a sensing resistor it must be possible to measure fast/slow decay currents, but for me it's not clear how to do that exactly. Has anyone a good idea to that?

    br
    helmut



  5. #65
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by Helmute2000 View Post
    Hi guys,

    I have some questions about the current sensing section, when we used here so small sensing resistors 0.02 Ohm were it not much better to have an op amp which multiplies the measured current let say about factor 5 or 10 or more? So the voltage level could be near 3Volts.

    Which op amp should be used for such multiplier? Is the tlv2772 good enough to do that?

    I read in the tlv2772 datasheet that this device is obviously an op amp and not a comparator why did we use here an op amp instead of a true comparator like e.g. LMV7219? (7 nsec, 2.7V to 5V Comparator with Rail-to-Rail Output)

    The schematic is not able to sense fast/slow decay. Fast decay currents will be hot-wired in the shottky diode D9.

    If we connect on both legs of the bottom FETs a sensing resistor it must be possible to measure fast/slow decay currents, but for me it's not clear how to do that exactly. Has anyone a good idea to that?

    br
    helmut
    You don't need to amplify the current because a comparator can work at mV levels.

    You don't need a fast comparator for this application because it is more sensible to switching noise (since there is no blanking time), and then you will need to introduce a better filter at the input delaying the switching anyways.
    Current on the motor does not increase suddenly due to the coil inductance, so there is not a big penalty (to the average motor current) to delay the switching off of the bridge (less than 1 uSec, due to the use of the op-amp).

    There is no control or monitoring of the decay current due to the design of the circuit to work only in 2 quadrants, so decay current will never be higher than conduction current (when high Mosfet and low mosfet on opposite sides are on) controlled by the chopper. D9 protects the input of the op-amp from negative voltages higher than its own Vf (0.3 volt)

    Is the tlv2772 good enough to do that?
    I used that chip on the Mardus-Kreutz chopper at 39.2 Khz and it works perfectly fine.

    Note: All your questions make perfect sense on a PID or PI controlled current loop...


    Regards,

    Kreutz.



  6. #66
    Registered
    Join Date
    Mar 2005
    Location
    germany
    Posts
    27
    Downloads
    0
    Uploads
    0

    Default

    Hi Kreutz,

    thanks for your replay, yes indeed I'm thinking of a PID or PI controlled current loop, but for me it's not clear how to do that in a best efficient way.
    I'd like to reduce the needed components for doing that.

    Also it's difficult to collect all current trails. At first I thought it could be a good idea to have two sensing resistors in each bottom FET leg. But the fast decay current has a special behaviour.

    1. Initial current
    This current can be measured directly on the sensing resistor on the bottom FED side which was switched on. One problem is, if I like to use only one comparator I need a analog switch like an DG411 - DG413. But I have no experience with such a component. Is it good enough for such way?
    Another idea from my side was to use two comparator devices in parallel.

    2. Slow decay
    Nothing to do, I need the same switch position as in point 1 described. No problem at all.

    3. Fast decay
    Here it is necessary to measure the current on the other side of the bottom FET leg sensing resistor also the polarity of the comparator has to be changed.

    For me it's unclear if it's possible to use such analog switches for doing that.
    Or do you know a nice schemantic to do that in a very efficient way?

    br
    Helmut



  7. #67
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by Helmute2000 View Post
    Hi Kreutz,

    thanks for your replay, yes indeed I'm thinking of a PID or PI controlled current loop, but for me it's not clear how to do that in a best efficient way.
    I'd like to reduce the needed components for doing that.

    Also it's difficult to collect all current trails. At first I thought it could be a good idea to have two sensing resistors in each bottom FET leg. But the fast decay current has a special behaviour.

    1. Initial current
    This current can be measured directly on the sensing resistor on the bottom FED side which was switched on. One problem is, if I like to use only one comparator I need a analog switch like an DG411 - DG413. But I have no experience with such a component. Is it good enough for such way?
    Another idea from my side was to use two comparator devices in parallel.

    2. Slow decay
    Nothing to do, I need the same switch position as in point 1 described. No problem at all.

    3. Fast decay
    Here it is necessary to measure the current on the other side of the bottom FET leg sensing resistor also the polarity of the comparator has to be changed.

    For me it's unclear if it's possible to use such analog switches for doing that.
    Or do you know a nice schemantic to do that in a very efficient way?

    br
    Helmut
    Hello Helmut;

    Using analog switches is going to introduce a lot of switching noise when changing channels (just try them switching two ordinary audio channels...).

    I suggest you to read the following application note: Unitrode Application note U-112 (Now Texas Instruments) http://server.oersted.dtu.dk/ftp/dat.../apps/u112.pdf .

    Remember that the by using a PI/PID loop you will make the setup motor dependent, and most of the people don't have a scope to tune the fast (about 20Khz required loop frequency) PI/PID current control loop.

    Regards,

    Kreutz.



  8. #68
    Registered
    Join Date
    Sep 2007
    Location
    Mexico
    Posts
    7
    Downloads
    0
    Uploads
    0

    Default

    How do you plan to drive the output stage?

    Sign-magnitude?, Locked Anti-phase?

    If you employ the former you might have to add a charge pump to help recharge the bootstrap capacitors when the motor is running in one direction for a long time.

    If you employ the latter, then you don't need those high capacity caps on the bootstrap circuit (depending on the PWM frequency).
    In the past I had used LMD18200 and the charge pump frequency is generated internally. However, when using IR2104 I start to believe that I MUST generate the charge pump frequency myself. Am I correct? I was choosed IR2104 because the high and low sides are mutually exclusive, so I do not have to worry about turning on both sides at the same time. But now what happen if I just held permanently high for instance the IN of one IR2104 and permanently low the IN of the other IR2104? (that it is the equivalent of having 100% PWM on on leg of the bridge and 0% PWM on the other leg) Does this mean that I must not deviate from 10% ~ 90% of the PWM?

    Thank you for any advice.



  9. #69
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by jc40 View Post
    In the past I had used LMD18200 and the charge pump frequency is generated internally. However, when using IR2104 I start to believe that I MUST generate the charge pump frequency myself. Am I correct? I was choosed IR2104 because the high and low sides are mutually exclusive, so I do not have to worry about turning on both sides at the same time. But now what happen if I just held permanently high for instance the IN of one IR2104 and permanently low the IN of the other IR2104? (that it is the equivalent of having 100% PWM on on leg of the bridge and 0% PWM on the other leg) Does this mean that I must not deviate from 10% ~ 90% of the PWM?

    Thank you for any advice.
    In that case you need to implement your own charge pump (you are working on 2 Quadrants). That is precisely the case I was talking about. Most of the literature talk about 10% to 90% PWM limits for 4 quadrant (Locked anti-phase) drive mode, but it all depends on PWM switching frequency and charge time window versus RC time constant of the bootstrap capacitors..



  10. #70
    Registered
    Join Date
    Sep 2007
    Location
    Mexico
    Posts
    7
    Downloads
    0
    Uploads
    0

    Default

    Thank you for your reply kreutz,

    my IR2104's should arrive today, so I will be testing it in a 24 volts circuit (the final circuit is 90 VDC). The frequency should be 20 KHz only. I do not need more than that. And for simplicity I am planning only to drive the motor in a full forward or full reverse mode. Well, I believe that "full" it will mean 90% PWM for one leg and 10% for the other leg, just to maintain the charge pump. 10% it will be 5 microseconds only, but I understand that those chips work with switching times so much small than that. And if I need to stop the motor in a particular position, I guess the Shutdown line will be the way to do it BUT with the PWM and /PWM active and cycling all the time.



  11. #71
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by jc40 View Post
    Thank you for your reply kreutz,

    my IR2104's should arrive today, so I will be testing it in a 24 volts circuit (the final circuit is 90 VDC). The frequency should be 20 KHz only. I do not need more than that. And for simplicity I am planning only to drive the motor in a full forward or full reverse mode. Well, I believe that "full" it will mean 90% PWM for one leg and 10% for the other leg, just to maintain the charge pump. 10% it will be 5 microseconds only, but I understand that those chips work with switching times so much small than that. And if I need to stop the motor in a particular position, I guess the Shutdown line will be the way to do it BUT with the PWM and /PWM active and cycling all the time.
    How do you plan to control the motor current?



  12. #72
    Registered
    Join Date
    Sep 2007
    Location
    Mexico
    Posts
    7
    Downloads
    0
    Uploads
    0

    Default

    Do you mean sensing the current through the H-bridge? Or maybe doing trapezoidal acceleration? Well, in fact the motor is not doing a very hard work. The starting only peaks to almost 3 amps only. I am not planning to do other control than to start the motor, run it for a seconds, reverse it, etc. Or maybe do you mean another concept that I am not understanding?



  13. #73
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by jc40 View Post
    Do you mean sensing the current through the H-bridge? Or maybe doing trapezoidal acceleration? Well, in fact the motor is not doing a very hard work. The starting only peaks to almost 3 amps only. I am not planning to do other control than to start the motor, run it for a seconds, reverse it, etc. Or maybe do you mean another concept that I am not understanding?
    I suppose you are trying to control speed or positioning. You should have a current limit just in case the current go over the continuous current limit for the motor, that is a protection for the drive, the motor, and the load in case of stall.



  14. #74
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default

    Yes. I am hoping between fuses and following error set in emc2 may cover those bases.. Until I get around to adding current limit. (yes I know that is scary ;0)

    sam

    Quote Originally Posted by kreutz View Post
    I suppose you are trying to control speed or positioning. You should have a current limit just in case the current go over the continuous current limit for the motor, that is a protection for the drive, the motor, and the load in case of stall.




  15. #75
    Member
    Join Date
    Aug 2006
    Location
    USA
    Posts
    2758
    Downloads
    0
    Uploads
    0

    Default

    Guys,

    I will be out of town for at least a week, with limited or no internet connection.

    Regards,

    Kreutz.



  16. #76
    Registered
    Join Date
    Sep 2007
    Location
    Mexico
    Posts
    7
    Downloads
    0
    Uploads
    0

    Default

    Just for curiosity...
    does anyone has tested the IR2104 with DC power rails fluctuating?. That is, for instance to control a DC permanent magnet motor of 90 VDC, 2 A, using the 110 VAC line, bridge rectifier, and then from there supplying the power to the H-bridge (the power supply for the IR2104 Vcc through zeners, and some filtering). I know that the peaks of the rectified voltage will be very higher than 90 volts, so what other method could be used to power such a motor, without using transformers?
    Thank you for any help in advance.



  17. #77
    Member
    Join Date
    May 2006
    Location
    Australia
    Posts
    2420
    Downloads
    0
    Uploads
    0

    Default

    It has been a while since I had much to post on this thread but I am posting to say thanks to those that contributed, my vers.1 H Bridge has been powered up and works a treat !!

    I ended up building my own version and kept it very basic, I used 2184 gate driver chips as they are the only ones my normal supplier keeps, I had actually etched a nice double sided board and populated most of the components that I had at hand then went to order the 2104's and ... well lets say that board makes a nice paperweight. At least it was a great test bed for me on double sided board making.

    After that I decided to make the board single sided, and seperate the processor (I built the first drive based on the elm-chan servo drive) and get it back to a H-Bridge only. Ordered the gate driver chips at this stage and built the new design board. Chips turned up and they were surface mount instead of the DIL packages I had designed the board for , another paperweight.

    New board and using the surface mount chips (only type my supplier stocks) and all is well, I already had a PWM generator built so plugged it in and voila !!!

    I have not looked at current limit circuitry yet, I intend to make something, just not sure how to go about it, at the moment I am building the processor board, once that is done I will add some refinements.

    Again thanks all, this stuff isn't exactly my field of expertise but it has been fun getting the old grey matter working again, I guess if I was a smart (and not so stubborn headed) person I would have just bought 3 Gecko's and have made lots of stuff on my router by now...ah well

    Russell



  18. #78
    Registered
    Join Date
    Jan 2006
    Location
    USA
    Posts
    121
    Downloads
    0
    Uploads
    0

    Default Current Limiting Explanation

    I am trying to understand how the current limiting works in this schematic. From page 5 of this thread.

    http://www.cnczone.com/forums/attach...0&d=1192069293

    Particularly D7, D8, R14, R15 and IC7A. Something fishy?

    Thanks,
    Jon



  19. #79
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default

    Yah - I tried to follow it but got a bit lost. I did my own similar to it...

    Untested but close. Also - you need to 'OR' the two pwm signals into the flipflop otherwise it won't reset.

    sam

    Attached Thumbnails Attached Thumbnails H-Bridge for comment-newschem5-jpg  


  20. #80
    Registered
    Join Date
    Jan 2006
    Location
    USA
    Posts
    121
    Downloads
    0
    Uploads
    0

    Default Connected wrong?

    Referring back to the original UHU schematic, I believe that the end of R14 that connects to R15 should go to ground instead. R14 would be R24 in the original schematic. Although I'm still not sure how it works.

    Jon



Page 4 of 6 FirstFirst 123456 LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

H-Bridge for comment

H-Bridge for comment