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

Hi Thanos And R-eng,
HOORAY :clap: !!!! I wish you great luck in the sourcing of parts! As you did for the DSMhb I trust they will be easily available everywhere and of reasonable cost. I'm guessing the midpower model you have in mind will be the most popular. The 12V option and 24V options should both be available. Is it possible to operate from mains supplies also? It would be really nice if one could have something where the builder could decide his power reqmt. and it could be built up on common PCBs. For example, you could select to use from 1 to 4 MOSFETs per quadrant, depending on your power reqmt. The system then becomes really versatile! Since heat sinking is so important, I hope adequate space will be between pairs which have a common heatsink, so that larger area copper heatsinks can be used, perhaps with a more powerful, centrally placed, fan. Senetor mentioned that its good to balance temperature in common heatsink MOSFET pairs. As you know, in my DSMhb, a thick walled copper tube links the common sinks, which I now intend to replace with thick aluminium or copper blocks, taking full advantage of the ~12mm gap between common heatsink pairs. Its an excellent move to leave the front end of the DSMhb untouched.

(Have you given some consideration to using thyristors for higher power?)

The DSMhb IMHO is great, and I am determined with a few small physical changes to enable it to handle 2 WW motors per axis. Thanks to you I haven't blown any further MOSFETs (yet!!) A few of the things I've done/ intend to do are:- Use wider heatsinks. Already done. Increased width from 10mm to 15mm bringing them dangerously close together. Used Araldite 'spacers' to prevent accidental contact. (See pic). Much time was spent ensuring the sink strips in direct contact with the MOSFET's metal pad were absolutely flat (fine flat file and finish off with 0 grade emery paper) for good heat transfer. This seemingly small change gives an ~30% surface area increase! These will be replaced by 14/15mm copper strips as soon as I can get hold of them. A lot of the air blown by the fan escapes at right angles to the strips without doing much work. Two cardboard guides were added to force the air to do more work by flowing along the strips before escaping. The air velocity from the open ends is now much greater. Two fan support bars were cut away to remove some obstruction to the airflow. Hopefully better cooling allround. While changing MOSFETs in the NO TC model, one casualty was the PCB stripes at the gates.(Maybe bad quality PCB material!) My new PCB will have these wider, as also the width of the solder-thickened tracks for increased current capacity and lower inductance. The long wires from the two motors (totally 14' long!) to the DSMhb were tightly twisted to reduce inductance effects (thanks to info on Pg 1 of the 4QD website). Tha attached pics will say all this much better.

I hope some of the above points will have some relevance while designing the new motor controller.

Regards, Erich.

Hi all,

I've been looking at Thano's great project for a while here and seeing how much effort that goes into the designs. Nice work indeed!

I don't pretend to have much knowledge on this subject but found this today. Probably not
much of interest, but thought I'd run it by you guys.

http://www.arrowne.com/innov/in214/f_1389.shtml

Anyway, please feel free with the flames...

Cheers, Phil.

Hi all, here's a circuit for a Simple Mosfet Tester I thought might be of interest.

Simple Mosfet Tester.pdf


Senetor.

Hi,

I decided to upgrade my existing DSMhb boards, using the IRF3205 mosfets. I already ordered some very cheap from ebay:

http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 0070728987

25 parts are enough for all my DSMhb's and will have spare parts too!



-The design of the new stronger h-bridge is going well. Here is a small teaser view of it (not complete yet!!!!)





Best Regards, Thanos

Hi Thanos,
Great news about the IRF3205 for the DSMhb--and more great news, I have them in hand. Can't believe it. But the faithful old IRFZ44N's are holding up this time.

The HP Controller teaser looks good too!! :clap:

Early congrats! Great stuff! :

Regards, Erich.

Hi,

Has anyone found on-line the HIP4081AIP, cheap and available in 1-2 pieces order?

I found it on Digi-key but they don't seem to sell to individuals...



Regards, Thanos

Tronicgr,

It seems that AVNET also sells this part. I don't know if they deliver internationally though. If this is holding up your project, I have a corporation that I can use to purchase the item and forward it to you.

:

I`ve ordererd from them some time ago as a student. I simply marked it on the form.
The shipping is pretty high for small quantities though...

Thanks. if I can't get my local electronics supplier to bring them, I might need your help... :


Regards, Thanos

Hi all, Farnell Electronics have them, they can be bought online in any quantity you like, plus postage. In Australia they are $6.35 ( Aussie dollars ) which is about $4.30 US. Plus $10.00 postage per order. They have online stores all around the world, so give them a try.

Senetor

Thanks,

I tried through them, but the closest point I was redirected was their UK store, that had minimum order charge £20... Plus their shipping service is only UPS that charge £15 for packages up to 1kg...

Darn...


Now I can see how if feels, for many people out there, been desperate about sourcing some parts of the AMC & DSmhb, even if they were pretty common... :


Regards, Thanos

Hi, found this on Intersil's web site,

HIP4081, Erratic Delay Times, here' a fix for the problem, and some other related reading for the device.

<a class=postlink href=http://www.intersil.com/rnwscripts/cust onclick=window.open(this.href);
return false;>http://www.intersil.com/rnwscripts/cust</a> ...


Senator.

Hi Thanos been looking around and found that device you were looking for .. i may be able to get one and send it to you found at RS Components <a class=postlink href=http://uk.rs-online.com/web/search/sear onclick=window.open(this.href);
return false;>http://uk.rs-online.com/web/search/sear</a> ... &x=24&y=12 Will try and place an order on monday through my company and then ship it to you ... let me find out how soon i can get hold of it and then i'll let you know when i'll post it .. Hvae to check stock and availability but we do a lot of business through them so fingers crossed ..

Hi,

As you know there is always the possibility of something to go wrong while powering up a motion simulator especially in start, where the motor limits aren't yet known or the direction is opposite or maybe a design flaw blocks the motor and make it stall. A stalling motor have big possibilities of burning down its coils or the h-bridge that drives it.

The best way to monitor and prevent such situations is to add an ammeter (perhaps a digital meter) in series to the motors and watch for the rising of the amperes having your hand on the kill switch for the power source. But what happens if we cannot have our attention on the amp readings, if we sit on the simulator on example and using it. We must somehow automatically monitor the current and have the motion controller cut the power when a current limit is reached that is defined by the h-bridge used and the ampere rate of the motor.

In this case we can use a current sensor. There are several solutions that I'd like to present here, hoping that more cheap ones may be found from others too. Some of them are commercial and others completely DIY.


1.Shunt current sensor: Its a piece of wire that we let the all the current pass trough and measure the fraction of the voltage that produces on its two ends, some mVolts, that is increasing as the current increasing in linear manner. The simplest way is to use a piece of 10AWG wire as described on this link: <a class=postlink href=http://www.rcgroups.com/forums/showthread.php?t=393591 target=_blank>http://www.rcgroups.com/forums/showthread.php?t=393591</a>
I already made such shunt and tested it successfully:

This method requires some effort since I had to calibrate the readings on the gray DVM, by putting in series my accurate Fluke meter connected as ammeter, and slowly cut the shunt piece smaller until it gets close. Its a cheap way, not more that a 1-2 euros for the cable. Also some additional circuit is needed to interface is to my AMC1.5 ADC ports.

Another solution that is more accurate is to use a commercial shunt current as this one:
<a class=postlink href=http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=350153196811 target=_blank>http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 0153196811</a>


Here how the current shunt working principle:




2. Hall effect current sensor clamp: Its more expensive but it uses hall effect sensor and a clamp that the power cables from the motor goes though and measure the current that pass though them without interfering in cabling in any way.
<a class=postlink href=http://scienceshareware.com/how-to-measure-AC-DC-current-with-a-hall-effect-clamp-.htm target=_blank>http://scienceshareware.com/how-to-meas ... clamp-.htm</a>





3. But the best solution I found on current sensors is this:
<a class=postlink href=http://www.nuhorizons.com/featuredproducts/volume7/Allegro/current_sensor.asp target=_blank>http://www.nuhorizons.com/featuredprodu ... sensor.asp</a>

Its a pure beauty! We could even place such a part on the h-bridge itself!!!! Its the safest way to use sing the measurement hall effect connects on the motion controller using only some analog output, 0 to +5v.


Any suggestions or comments are welcome.


Regards, Thanos

Well price of ACS750SCA-050 is really good <a class=postlink href=http://cz.farnell.com/allegro-microsyst onclick=window.open(this.href);
return false;>http://cz.farnell.com/allegro-microsyst</a> ... dp/1457175 its easy and clean solution.

Hi,
i have some Problems with the H-Bridge, i think it will be the H-Bridge.
The output voltage for the motors should be the same as the source voltage (Battery), right?.
So the battery has 12,4 V, Motors are not mounted to the simulator.

1. If i run only Motor1, the output voltage is 11,54V in one direction and in the other direction 10,35V.
2. If i run only Motor2, the output voltage is 11,65V in one direction and in the other direction 9,82V
3. If i run Motor1+Motor2 the output voltage is 10,8V (M1) and 11,2V (M2) in one direction and in the other direction 9.3V (M1) and 8,35V (M2). For example, if i first start Motor1 and then Motor2, Motor1 go slow down with the speed, because of decreasing of voltage

So i have different speed, because of changing voltage.
The battery has a stable Voltage from 12.2-12.4V.
If the Motors are not connected to the bridge, i have a maximum output voltage as high as the battery voltage.

What do you think is the Problem?. Could it be not correct or bad soldered connections?, or why i am loosing some voltage?.

:clap: Ha Ha Thanos I spy with my little eye Hmmmm Could that be the eagerly awaited prototype for the new H Bridge PCB seen in the bottom right hand corner of PICT2304usm .jpg Just bought The new Mosfetts off Ebay so looking forward to the rest LOL ... Gonna Need that Board for the Big Flight Sim :lol:

You spotted an h-bridge, but this isn't mine. Its Roland's h-bridge that he kindly sent me to do tests with. It takes analog input 0 to +5volts, with 2.5v to be the center position. He even added an RC filter to be able to connect it directly on PWM outputs of the AMC1.5. Here is a link to this motor driver he designed: <a class=postlink href=http://simprojects.nl/electrical_drive_3.htm target=_blank>http://simprojects.nl/electrical_drive_3.htm</a>

On the photo you see it driving Monast's Festo servo motor with 24V power supply. This servo motor is amazing...! It draws 1/3 of the current that wiper motors, and its way more quiet!!!




Regards, Thanos

I have done some Tests and i think the problems could come from the Mosfets (IRFZ44N) :? . I have ordered some new Mosfets and i will try it again if they arrive.

Hi,
Is it possible to mount on DSMhb another power mosfets, that can support more current?
For example I want to use IRF1405: <a class=postlink href=http://www.irf.com/product-info/datasheets/data/irf1405.pdf target=_blank>http://www.irf.com/product-info/datashe ... rf1405.pdf</a>
It support 169 A.


thanks
Stefano