Hi All,
I have been suffering for months now in designing a H bridge that controls a 36V DC motor with 3 Amps stall current and 1 Amp running current. The Transistors used to blow (TIP122 & TIP127) until I added a flywheel diode ( although there is once included in the transistor, however, my main problem now is that the controller (PIC16F84A) re-triggers and changes the motor direction when the power is cut out from an external limit switch. I have two limit switches witch paralleled diodes to stop the motor motion at the end of the stroke. I have added RCD snubber on the H bridge circuit at the +36 V and ground nodes but still I think the motor arcs and retriggers the controller. This phenomena only occurs at limit switch opening cutting the current from the motor. I am using a satalite dish linear actuator motor and using its cam and built in limit switched without amendment.

Anyone's help is mostly appreciated

 

Can you do a simple schematic of the power circuit, and describe the power supply and the filtering for the power supply used by the H bridge.

It sounds like you only have an RC filter across the incoming voltage supply, or are you saying they are from the motor nodes to the rail/ground.

How is your PIC powered and the drivers implemented.

Does the motor use brushes?

 

Hi Tim, Thanks for your prompt reply. I have an AC source (36VAC which gives a peak of 52VDC after the bridge) smoothing capacitors are added before and after LM318T variable voltage regulators. I use several of these to drop down from 52Volts to 36volts (where backup batteries are connected with OR diodes, then step down to 12Volts for the remote control circuit, then from the 12V to 5 Volts using 7805. My guess is that there is a huge surge arc that happens when the limit switches cut the current from the dc motor.

The drivers are TIPs NPN side gets the signal directly from pic with pull down R at base, while the PNP TIP is controlled with 2n2222 as i cant driver the upper side with micrcontroller input. The bridge has been tested with 10 OHM 100 Watts resistors and works great. but when inductive load as motor is connected, the story begins. I have tried an RC snubber (Voltage clipper mode) 1n4007 Diode in parallel with 1K and both in series with 1000uF Cap. the surges seems to have been reduced that now the self triggering happens like once every 10 strokes. I measured the inductance of the motor previously and it was 4.039mH and ?I simulated it on SPICE with 5ohm resistance and 4.04mH coil. (there is how I got the snubber value where the surge was reduced from 900 volts to around 36 volts. I also measured the limit switch timing using osilloscope and it was 1mS.

Anyways, will post the schematics and wait for your help. Thanks.

 

optical isolation or a separate power supply, not connected to the motor circuit.
Either one might be all that you need to do.

Will wait for the schematic to be posted.

 

Do you have a 0.1uF decoupling cap across the power pins on the PIC?

I'd also add a 22uF electrolytic on the power line, as close as possible to the PIC supply pins. The 0.1uF should be as close as physically possible, sometimes as a SMD cap underneath a DIP package.

 

Hello,

As thatoneguy sais are there decoupling capacitors in place?
This thread will give you a lot of info on it:
Decoupling or Bypass Capacitors, Why?

On snubber circuits there are a lot of links on the EDUCYPEDIA:
http://www.educypedia.be/electronics/powerelectronics.htm

Bertus

 

Hi All, thank you very much for your replies. I dont have a 0.1uF cap next to the PIC, I only have a 100uF electrolytic after the 7805. Please find attached the main parts of the schematic, please let me know if there is anything else missing. Will add the 0.1uF cap and give it a try.

Please find attached as well Limit switch connection to the motor.

 

Without looking further into your design than just power path, I suggest capacitor bypassing the output of the first LM338 (ie. H bridge) with probably at least 1000uF, and also the input of second/third LM338 (say same cap level). You should use a star 0V arrangement for the main filter caps, and the TIP122 emitters.

Add an RC across the motor itself, and also each limit switch contact.

Tim

 

Thanks Tim for the reply, All is clear except the for the star 0V arrangement, would you please explain what is it? Thanks.

As for the RC circuits, I tried 1K ohm + 1000uF Cap and also 450 Ohm + 0.1uf cap and both clipped the voltage to +36 max, which one is the most correct?

 

I would suggest that the RC snubbers are in place mainly for higher frequency ringing suppression, even for the motor (brush noise) - rather than lower frequency (1k/1000uF) transient control. If you have a waveform capture cro you can look at the current waveform through the snubber via the resistor voltage waveform - or use a LEM module.

If you search on 'star grounding' techniques, and try and appreciate where the transient current loops exist (eg. 36V filter cap through motor) - then you don't want that motor transient current going through a section of 0V cabling that supplies power to the uP - rather you want it to only connect between sections at a single point/star. A lot of good stuff with respect to audio circuitry (eg. http://valvewizard2.webs.com/Grounding.pdf).

Tim

 

Hi All, thank you very much for your replies. I dont have a 0.1uF cap next to the PIC, I only have a 100uF electrolytic after the 7805. Please find attached the main parts of the schematic, please let me know if there is anything else missing. Will add the 0.1uF cap and give it a try.

Please find attached as well Limit switch connection to the motor.

Add the 100nF (0.1uF) cap (I use Mica or film caps) between the power on the uC, and let us know the results.

Adding caps (0.1uF) across the supply pins of comparators will give additional benefit, if their outputs are glitching with the main power switching.

If that doesn't help, please provide photos of the circuit in question. The ground should be a "plane" on the PC Board that the uC and comparators are mounted on, and this should be liberally "sprinkled" with both small/fast decoupling caps (mica/film), and larger electrolytics with low ESR. If you have thin round wires between circuit boards, change them out for flat braided cable, it looks like solder wick, only 10 times bigger. Try to maintain very low impedance power and ground links to the supply. Flat and wide are somewhat immune to EMI, and can also provide a lower parasitic resistance.

I wouldn't take the steps of rewiring the entire circuit until decoupling caps are placed across the power and ground connection of every IC, though.

If you have a scope, it would help us find the issue quicker, so let us know if you have access to one (100Mhz or so).

 

Thanks Tim and Thanks for thatoneguy for the info, regarding the CRO, is there any way to see the surges on an analogue one. I have a pretty old Oscilloscope (HAMEG HM412-5 up to 40Mhz) with a max. of 10x probe available.

I will try all recommendations and let you know the results today.

 

Aussa I think everybody answer already the question with a great explanation but I have one question. The battery back up circuit, have you tried if that is functioning? Well I believe the purpose of having the back up battery is to provide a continuous supply to the circuit once the main supply is off, is that right?

 

Yes Newbies Hobbyist, it is working fine, the diodes act as an OR gate, paralleling diodes provides a higher current (each SR can withstand 3Amps). preferable to use fast switching diodes like the SR360.

The working supply will be the one with the higher output voltage.

In my schematic, I rely more on the batteries as they are charged with a solar panel. and as soon as their volts go below 36.75Volts, the transformer takes over (In my case batteries are main supply and Transformer is back up supply)

 

I see, I thought your battery is not your main supply. Because I have a circuit that is using a back up battery when the main voltage is lower than battery the battery will supply the whole circuit and I wired the two diode different, one end is in the output of my main supply and the other diode is connected to battery.

 

Just to clarify - I interpreted the battery backup as only for the uP supply, not for the motor - is that right?

 

tim yeah I think you are right it is just only for the uP that is whay I asked aussa to clear my thoughts also.

 

Guys, the batteries are an alternate 36 volts source for the h bridhge and microcontroller. Please check my chematic on the first page.

 

aussa - I think your schematic is in error - as the diode circuits as shown would prevent the backup battery from energising the H bridge.

 

aussa, I completely agree to tim. First of all those two diodes before the 2 OR'd circuit they are in reverse bias second if you are saying that the battery will work in the absence of your main supply by monitoring the voltage level of the two supplies then you wired the two diodes in back up battery circuit wrong. Please review your circuit.