Tronic's DIY dual Mosfet H-bridge (DSMhb)

Re: Tronic's DIY dual Mosfet H-bridge_

I think next week i get my ordered parts and then i start to make the PCB. Thanos, or someone else, can you help me with my question above?.

Re: Tronic's DIY dual Mosfet H-bridge_

@ Thanos, or anyone else who may have a good background in electronic design.

A few days ago I hooked everything up; my simulator (which I have now definitively given the name ICE-sim, Interactive Cockpit Environment Simulator), AVR-DSMhb electronics module, X-sim control software, R-factor, and PC into a fully functioning 2DOF system. It was initially powered up with only 12v driving pretty strong motors (much stronger than wiper motors) capable of 110A each. It ran beautifully over 2 days with a total of 6 hours continuously + 4 hours continuously. R-factor was set to A.I. mode running the BMW Sauber F1. This initial testing was done without a load typical of a human body aboard.

I then increased the electrical drive to my final target of 24v, reduced to about 60% the motion extensions of the ICE-sim from the maximum possible I was able to attain at 12v, this as a precaution. After starting, all hell then broke loose...Not sure of the precise sequence of events because it happened so quickly. Either, due to the increase to 24v the ICE-sim reached its maximum lateral-G position and jambed itself to one side as it tried to reach the required position, or just as it reached the maximum position it drew too much amps through the DSMhb and popped the lateral-G H-bridge low-side mosfets (IRLZ44) and 74HC08 chip, and then jambed itself before it could stop. I think it jambed first and then tried to draw more amps than the mosfets could handle, and so they popped! The long-G (pitch) motor actually popped the 20A inline fuse first, probably due to the destruction of the 74HC08 which then allowed a short through this legs mosfets. All this happened in about 3-5 seconds!

I repaired the DSMhb and tested with the original mosfets in the long-G (pitch) leg, but one of the mosfets immediately cracked, so it had sustained some damage also. Then I replaced all active components on the DSMhb as insurance. This was successful and the test wiper motor worked again. But I tried something a little different on one of the H-bridge legs. On the high-side mosfets I have been using the STMSTP80P device with complete success. For the low-side I decided to try its complimentary device, the STMSTP80N. That leg has been incapable of activating the attached wiper motor. The other leg with the low-side IRLZ44 mosfet works just fine.

The side with the all STM devices activates it's corresponding status led's when there is no motor attached to the outputs. With the motor attached, I can see the led quickly flash but not light continuously, and the motor initially makes a click sound, but will not run. It seems to not have enough voltage provided to the low-side mosfet to turn it on and then turn the motor.

I have just enough knowledge of electronics to be a danger to myself.

But if possible, I would really like to use both the P and N STM mosfets together, as they seem to be a bit more robust. I need to be able to turn on the low-side, N device. Yet it appears that maybe I cannot, because it is driven directly by the output of the 74HC08, which I measured as 4.96v out. There is going to be a drop in voltage through the mosfet, and it requires 2min-4max volts at the gate to turn it on. This is the same as the IRF44, but it may be so close, that it cannot trigger it to turn on. I have been using the IRLZ44 which only needs 1min-2max at the gate. And it has worked fine at 12v, but without a human load in the ICE-sim.

Ideally, I can physically configure the DSMhb to take 2 paralled mosfets in piggyback mount at each of the connectors but then, can the drive circuits to the mosfet gates be adjusted to handle the increased turn-on voltages needed? Especially at the low-side mosfets.

I am researching online, and have seen Thanos' suggestion to Igor sharky to change the 3k6 and 1k5 resistors, but that only affects the high-side mosfets, which still seem to trigger. If anyone has expert suggestions, it would be very, very welcome to me.

I have also documented many voltage readings (but not current) at points all over the DSMhb in various states with both types of N mosfets. If this is helpful I could attach those results here in another post, as well as links to mosfet data sheets in pdf.

Thanks!
R-eng

Re: Tronic's DIY dual Mosfet H-bridge_

Hi R-eng,

I think that you need some kind of voltage amplification of the N mosfets gates above the Vcc level that the 74HC08 chip can provide. The link below explains some things about it, and has a voltage doubler circuit that you can try...

http://www.cadvision.com/blanchas/hexfet/nch-brdg.htm


I wish I could help more, but my design knowledges are extended only in the digital area of electronics, and still trying to figure the analog area thru experimenting...

Regards, Thanos

Re: Tronic's DIY dual Mosfet H-bridge_

I just want to remind you to use end-limit switches if possible and safety cut-off switch for the motors or the powersupply.

The idea is to let the motor protect itself when reaching the end-limits even if this was not detected from the controller or the over voltage fuses!

Look on the next diagram:



The switches can be placed on the travel of the motor so they cut the power to the motor when they pressed. The diodes serve as reverse direction helper, so the h-bridge can still command the motor to free himself from the end-stop by giving him power to the other direction.

Here are some large switches that can be used for this purpose along with the heavy duty 6amp diodes and a line cut-off safety switch


The line cut-off safety switch cuts the power line of the power-supply and locks in place!!! Can be unlocked if you turn around its head. One positive of this huge switch is that you can press it even with your feet!!

Regards, Thanos

Re: Tronic's DIY dual Mosfet H-bridge_

Thanks Thanos, I have previously read most of Blanchard's H-bridge related website, and was looking at the better mosfet h-bridge, as shown here. What do you think about it?

http://www.cadvision.com/blanchas/hexfet/np-s.htm

I also found another N-chn mosfet which is identical except that it has a much lower Vgs(th) gate to source threshold of 1.6v (typ), with significantly lower Rds(on). This is nearly identical to the IRLZ44 in Vgs(th), and would give me the needed headroom.

Another option perhaps is to drive the 74HC08 from another regulator on the DSMhb with a 6.0v output. It can handle a max of 7.0v, so this may give me another +1v margin to drive the mosfet. Would I then need to increase the 3k6/1k5 resistors to maintain the original voltage at the base of BC546 you spec?

R-eng

Re: Tronic's DIY dual Mosfet H-bridge_

@ Thanos,

Out of curiosity, in the firmware control of the pwm/motors, do you use a brief moment of engine braking between movements to reduce back EMF, as suggested by Blanchard?

http://www.cadvision.com/blanchas/hexfet/index.html
stopping a motor

Perhaps the DSMhb circuit does not lend itself to this method.

R-eng

Re: Tronic's DIY dual Mosfet H-bridge_

As a followup...I tried the last option first, and used the output of a 78L06 +6v regulator to successfully activate the N-chn mosfet. It ran the test wiper motor, so indeed, it needed just a touch more voltage to the gate. No changes to the resistors.

R-eng

Re: Tronic's DIY dual Mosfet H-bridge

What could happen with the mosfets if i didn`t use a battery as buffer? and what happen with the backfiring (regeneration) energy?. Could it be that the mosfets become very hot or overheated?
I want to us a power pack ( 400VA AC-Transformer with a electric rectifier = 15V DC) without a battery.

Re: Tronic's DIY dual Mosfet H-bridge_

I have a couple of questions:

- Wich is the latest partslist?
- I use molex terminal connoctors like in Thanos's partslist, so wich pcb do i have to use? 1.2b?
- What kind of diodes to use for the limit switches? (I use wipermotors)
- Do i have to use a groundcable because my frame is made of metal? If so, where do i have to place it?
- Will a 12v 30a powersupply do the job for the motors?

Re: Tronic's DIY dual Mosfet H-bridge_

Hi Thanos,
I'm trying to finalise the fusing of the DSMhb1.2. I have at present one main line fast-blo fuse rated at 30A wired for a single 12V battery and 15A individual fast-blo fuses for each wiper motor. What do you recommend considering that you use a joyrider frame and a flight program with a lot of turbulence. I think rapid oscillation (turbulence) can cause huge spikes. I had earlier lost a set of MOSFETs (2 sets actually, one for each bridge!), even though heatsinks and cooling fan were OK, so I want to be doubly sure this time. I don't have an inline ammeter, but I doubt spikes would show.

Would it be a good idea to use a circuit breaker(s) of suitable rating.

I have noticed in the WW motors I use that there is no interference/spark suppression circuitry (caps/ resistors/inductors etc) I have seen in some of your pictures here, and in other motors. Can this cause spikes which can damage the bridge? Do you have an external circuit that can do this job? I love the turbulence, and I hate changing MOSFETs!!!

Thanks & Regards,

Erich.

Re: Tronic's DIY dual Mosfet H-bridge_

Hi Erich,

I think you should use smaller 10 amp rated fuses per motor to start with, as they must burn first preventing damage not only to the mosfets themselfs but to your wiring and the board too!

On my joyrider it was easy to reach 10 amp, only on the Pitch motor that has the more mass to move (LCD monitor, controls etc.). So I used only two 10 amp fuses on my sim, without the 30amp main fuse you do.



Anyway I suggest you to get an inline ammeter and voltmeter too. Its easy to add them on the battery power line and can save you from having problems with load after all. By watching the ammeter you can balance perfectly the Pitch axis motor. If each axis is balanced well the ammeter will increase only when there is motion involved! Also a emergency press shut-off switch (big red button) is handy to have to enter or exit the sim without electrically stressing the motors.




I don't think that you will have any trouble with WW motors, since all motors act as a huge inductor/resistor. The only thing you can do is to add some capacitors between the two poles of the motors, only if you are facing interferences in the ADC potentiometers feedback. In that case you could always use alternate potentiometer solutions like the SS495A ratiometric Hall Effect sensors!!


Best Regards, Thanos

Re: Tronic's DIY dual Mosfet H-bridge_

Hi Thanos,
Thanks for your quick reply. I wonder if you have found a formula for not sleeping! Please share.

10A!!!! So I'm way over the top. I was thinking that since the IRF44 had a 49A rating, and is the lower of these partners, I could take a lot more liberty. Apparently not---and I can appreciate that. And the 30A fuse appears irrelevant--funny, but that's the one that keeps blowing--any ideas?

Meantime, here's something I just found that covers a lot, but I refer specifically to
Chapter 23-11 which covers spike and noise suppression. Since I
have no way of measuring the effectiveness, which of the circuits
shown do you think might be of relevance to our requirement? Ok, you've already more or less answered that.
50nF-1uf across the brushes, and 50nF-1uF from each brush to earth is what is suggested.

http://books.google.co.in/books?id=DM6j ... pg=PT587&l
pg=PT587&dq=noise+spike+suppression+small+dc+motors&source=
web&ots=kKIrCMO8A6&sig=lZdjWRU3U5VYKYxh97EbXaUom9Q&hl=en&sa=X&oi=book_result&resnum=10&ct=result#PPT587,M
1

Actually, this is quite a general electronics treasure house. :yippiee:

Thanks,

Erich.

PS. I think that Voltmeter-Ammeter board costs less than a set of MOSFETs, so it goes on tomorrow---at least the Ammeter.

Re: Tronic's DIY dual Mosfet H-bridge_

Hi Thanos,
Got myself the Ammeter 0-30A, at least. Also, for abundant precaution, I'm adding 3x1uF caps per motor, one directly across the brushes, and one from each brush to motor body, and motor body to sim ground. This Nissan model WW motors have no internal suppression components, or they may have been externally fitted, I don't know.

Besides this, I feel I was a careless in the setting up process. Here are some things I should have done:--

a) I should definitely have had at least the Ammeter, if not the Voltmeter, in the main line

b) I should have connected only ONE AXIS at a time, got the feedback pot and motor directions correct, disconnected this, and then set up the second axis. I had the misfortune of having both pot AND motor orientations incorrect when first connecting. Anything can happen in this confusing situation (at least, for me!).

c) I should definitely have started with lower rated fuses, Maybe even 8A per motor. I used 15A fuses per motor to start with, under the mistaken impression that it was OK because the bridge could give upto 40A easily. I didn't take peaks/spikes into account---possibly they are present!

Sad to report that my gallery of defunct MOSFETs has increased to 4 sets ( and I have the 'NO TC' PCB!!). (Anyone have a good foolproof circuit for testing MOSFETS?) Sadder for sure, but definitely wiser!! :?

I shall report the results when things work out.

I hope this post is of use to others as well, approaching this stage of building their sims.

Or put it all down to 2008, and hope 2009 is a better year!!

HAPPY NEW YEAR!!!! :

Erich.

Re: Tronic's DIY dual Mosfet H-bridge_

Two things I would like to offer.

One, I experienced another meltdown of my DSMhb when attempting to test a heavy weight in the ICE-sim. I want to caution others not to do what I did while trying a shortcut in procedure. I needed to change the polarity of the X-sim effect for one of the motors. I decided to leave battery power connected to the DSMhb, and stop the X-sim profiler as well as pause rFactor. I should have known better, and I do, but I was lazy! The moment I changed the [invert] option of the effect channel, the corresponding Schottky diode melted. This was not good ;-) I replaced it but other components were damaged, but not visibly. Probably more mosfets were on permanent vacation again!

Do yourself a favor...remove the battery power from the DSMhb when making polarity changes in X-sim Profiler. It's a fraction of a second, not hours, and it will definitely cost less.

Two, I am driving large motors and when attempting to drive the heavy weight in the ICE-sim, I was popping 20-amp fuses between the battery and DSMhb almost immediately. I want to parallel mosfets with a modified DSMhb to drive my high amperage motors so that a continuous 60-70 amp load is feasible. I have been researching online for info to do this properly. I am kindof tired of buying and replacing mosfets seemingly regularly. In my search I came across this website of products and good info.

http://www.robotpower.com

Check out the info on the OSMC project. No offense Thanos, but this looks pretty sweet! The difference hear is that only N-chn mosfets are used, so the very special IRFB3207, 180A, 75V device is useable. Anyhow, maybe all this will be of help to others.

p.s. There is also a link to a current limiter device which has been discontinued, but the documentation is there.

R-eng

Re: Tronic's DIY dual Mosfet H-bridge_

Happy new year everybody!!!

My powersupply for the wipermotors has arrived. It's a 2-16v 40amp max. supply. I think that will do the job. The question that i have
right now is... What kind of diodes do i have to use for the limit switches on the motors?

Re: Tronic's DIY dual Mosfet H-bridge_

Hi R-eng,
Thanks for your advice. I read your Posts #88 and 90 with interest. So did you replace the current MOSFETs with STMSTP08P/N's with good results?

Regards, Erich.

Re: Tronic's DIY dual Mosfet H-bridge_

@ erichans

I did use the STMSTP80PF55/STMSTP80NF55-06 (or 08) and it has worked very well. But as I explained in post #88-90, the N's required me to drive the 74HC08 with an additional regulator which I mounted next to it, a 78L06 +6v regulator. But be cautioned...the 74HC08 cannot take more than this +6v. When done this way, it was able to turn-on the N-chn mosfet with no problems.

I also had to disable the +5v line at the molex connector to the DSMhb, and then add the wire link to +12v output on the AVR board. I use a 12v dc adapter, but the 9v Thanos has suggested is also fine. The additional filter capacitors are 100uf/16v with a 0.47uf/50v mylar bypass.

R-eng

Re: Tronic's DIY dual Mosfet H-bridge_

Thanks, R-eng. It seems obvious that you did the same tests as in posts #88 and 90 would also be very helpful if you could detail how, if you did them differently, and for how long you used the new MOSFET setup, max temps, if you measured., etc. Overall, would you say this setup gives you much more headroom with the same circuitry of DSMhb? Barring, of course the required 6V change.

Thanos, I was hoping, as suggested by Ego, to go for the 2 motor-per-axis setup once and for all, as I think my heavier metal frame would need more zip to reach your agility. Would the current DSMhb setup still be OK for this, assuming I 'phutted' mine out of sheer stupidity! Your Athens Show video was excellent---single motors-12V!!!

I'm still stuck in midair some might say is a good place to be in a flight simulator! :?

Thanks, Erich.

Re: Tronic's DIY dual Mosfet H-bridge_

Hi Thanos,
I found a MOSFET seems to be a closer complement to the IRF4905 than the IRFZ44N. It is the IRF3205 and I took down some key figures for all 3 for a quick scan:--

Device------IRFZ44N N-Ch-------IRF3205 N-Ch-------IRF4905 P-Ch
VMax-------------55V----------------55V---------------------55V
Current----49A @ 25 Celsius---110A @ 25 Celsius----74A @ 25 Celsius
-------------35A @ 100Celsius-----80A @ 100 Celsius----52A @ 100 Celsius
Dissipation------94W----------------200W----------------200W

I took the above figures from their respective Datasheets. You think this might work to give us a bit more headroom--possibly more safely take on that extra motor per axis?

Regards, Erich.

Re: Tronic's DIY dual Mosfet H-bridge_

Hi,
Thought I'd share this quickly: http://www.4qdtec.com/pwm-01.html I got led to it directly or indirectly from here. I particularly draw your attention to the matter under Main Capacitor.

Could account for the fact that my motors ran well off load and in close proximity to the battery---but with the two negatives added, load AND longer battery leads when installed on the Sim, conditions were ideal for overheating >> burnout. In the Sim, the battery leads are probably a huge inductance. The positive goes through the support wheel switch**--4'--comes back to the sim motion on/off switch in the cockpit--4'---from here, to the Controller--6'--all told 14'~~!! The negative is shorter, Batt. Neg to Controller---3'. IT GETS WORSE!! Each wire pair to each of the MOTORS are 5' and 6' long, respectively, for the roll and pitch motors!! And what about the addition of the Ammeter as mentioned in the article? But I won't run without that! And I have reduced the individual motor fuses to 10A as a starting point, as advised by Thanos. I will also temporarily retain the main fuse, but at 15A, possibly 20. If that starts blowing, then possibly wiring inductance is the culprit.

And did I twist the leads as suggested in the article?--NO, not yet!! I didn't even know about this--but I'm onto it right now! What's a mystery is why the 30A fuse in the main battery line blew several times, yet the individual 15A fuses to each motor didn't blow even once!

Erich.

**Support Wheel--A sort of retractable mini-undercarriage resting on the ground that prevents heavy load on the WW motor plastic pinions when the pilot/driver enters/leaves the cockpit. For safety, a switch is linked to this to prevent motion being started if its down.