AMC Compatible H-Bridge

I would like to present a new and improved H-Bridge plan that is compatible with the AMC. I will cut to the chase, so here are some details:

Schematic.pdf
Component values to be finalized through testing.

Improvements to the DSMhb:

-Uses 4 N-Channel MOSFETs instead of 2P+2N channels.
-IR2110 driver IC ensures fast switching and maximum saturation of the transistors for efficiency, also provides a layer of isolation for the AMC.
-High Power MOSFETs can be used, maximum current will depend on the actual devices used (100A+).
-Fuse protecting from short circuits. Current rating depends on the transistors.


For consideration:
-Supply voltages between 12V and 24V due to the driver limitations. Most FETs have a maximum Vgs of 20V, which inevitably asks for a protection circuit when used with a 24V supply.
-There is no hardware shoot-through protection. An extra XOR gate and 2 more AND gates are needed. The 74HC08 already has the 2 extra AND gates needed.
-High-side MOSFETs cannot be turned on continously (driver limitation), maximum PWM duty cycle about 99%
-Additional flywheel diodes may not be needed if the MOSFETs diode is sufficient for the currents.


I'm sure the lists will grow as this is still in the planning and design stage, and I would like the community's input and opinion.

Please leave your ideas, suggestions, features, and also your preference for the physical design of the circuit such as: connector types and layout, size, cooling, or anything that comes in mind.

Regards,

Adam

Frakk, why don't you use the HIP4081A Full H-Bridge driver chip, only one required per H-Bridge and about 40% of the cost of two IR2110 IC's plus all the other benefits of charge pump and boot strapping for upper Mosfets. Available in PDIP or SOIC

HIP4081A.pdf

HIP4081A- Application Note.pdf

Senetor.

I have the HIP4082 and I tried to get the bridge working with 2 different chips. The first probably burned out, the second one I was more cautious about, but it was overheating pretty bad and the output pins didn't work as they should. Maybe I got a faulty chip for the first one, but the second one acts the same way. I tried for about 2 weeks and couldn't figure out whats wrong with the circuit and decided to ditch it. One day I will go back to it and see if I can get it to work.

I have read about these drivers from people in the industry using them. The IR2110 is a more reliable mainly due to the isolated digital circuitry. And the other good point beside it is that I could get it to work Also I think they are more widely available and easier to get, which IMO is a big deciding factor for a DIY project. If you have to order 2 IC's from out of your country, their price will go way up.

Where do you get your prices from? The HIP408x series is hard to find in some countries (I could only order the 4082 from one supplier in Hungary), especially in the DIP package. At the time I bought the HIP4082 for the same price as the two IR2110's, about $6USD. I noticed that Digikey and Newark sells them for ~$6USD a piece, which is ridiculous.

I am open for ideas and different components, if there is something better, cheaper, easier to get, I will be more than happy to use it.

edit:
Found these on ebay: 5x IR2110 for $15 + free shipping, that's 2 H-Bridges plus you have an extra in case something goes crispy critter.
http://cgi.ebay.com/5-PCS-NEW-ICS-FETS- ... _819wt_939

Hi Adam!

This circuit looks more complicated and expensive than Thanos HBridge. But perhaps it would be more reliable and complete device. In any case, I'll try to do it after you finish and share all the necessary information.

WBR, Nickolay.

The only significant cost difference is the IR2110's, add ~$6/H-Bridge, so $12 for the dual setup of DSMhb. You can add $1 extra for all the small pieces like capacitors, schottky diodes, fuses. Block terminals can be replaced by other connectors, I prefer banana plugs and heavy gauge cable audio cable for high currents. Also the transistor terminals can be eliminated by soldering.

If there is a need for a kit of parts, I can look into the cheapest bulk prices and offer it for you guys.

Hi Adam!

It may be interesting to us. But first we would like to see the completed device. We are waiting for news about your new Hbridge!

P.S. What gauge are you going to use cables?

WBR, Nickolay and ru.zlokot

The cables depend on the currents, how long they are and of course the price. The connections will be a lot more critical.

Here is a good table of the specs:
http://en.wikipedia.org/wiki/American_wire_gauge

I would say 12-16AWG is pretty safe for our applications.

Just a little update:

I put together the AMC and the H-Bridge for proof of concept. The digital circuitry and the IR2110 control logic is running off of 5V, the H-Bridge and gate driver part is taking 12V from a small 2A PC power supply. Everything went just as I expected, so it is great success! :thbup:

Top right is the H-Bridge, red and green LEDs show the direction and speed of the motor by changing brightness.Bottom middle/right is the AMC, bottom left is the AND gates, part of the H-Bridge circuit.
The 4 LEDs show the states of the MOSFETs (ON/OFF). Red and Green is the two bottom ones, the two yellow beside are for the top MOSFETS across them.



I changed the circuit just a little bit around the AND gate. The new configuration turns the bottom transistors ON and pulses only the top one across it, instead of pulsing both at the same time as in the original DSMhb. This reduces switching losses, improves efficiency and makes the PWM more accurate in the case there are switching delays of the top and bottom sides.

Updated schematic with some parts and values:
Schematic1.1.pdf

The next step is to solder and test the circuit under a load.

I will also start designing the PCB layout, so please give me your ideas and recommendations regards to that.

Thanks for reading,

Adam

ı will soon try your circuit too.
if u get success in load (over 5A with low pwm) then you prove your work thank you for improving hbridge.
Waiting for good news.

5A will not be a problem because the transistors can handle 49A continuous and it is easy to just put a higher current FET in it. I will try to push it to the limit and see how it reacts. Making a PCB is also a big challenge. It needs power lines to handle the currents, good connectors, efficient cooling that isolates drains, practical size and form factor. But first and foremost it needs to be safe!

What kind of fuses would you prefer? I have been looking into the ATC type fuses. Cheapest and smallest solution, up to 40A I believe.

Currently waiting on some perf boards and connectors.

Frakk, I have been using these pcb mount blade fuse holders and find them easy to solder and they use Automotive blade fuses.





Senetor.

That is what I've been thinking too. I am looking for distributors that sell them for cheap.

http://au.farnell.com/multicomp/mccq-12 ... tt=1586596

http://au.farnell.com/keystone/3522/fus ... tt=1608218

Auto Fuse Clips.pdf

Senetor.

The 3550 horizontal clip is perfect on that catalog page!

I was also thinking about just regular blade connectors that are big enough for the fuses. Something like this:
http://www.sjscycles.co.uk/images/produ ... mnon_l.jpg

First prototype is ready to be tested. :clap:



I just ran some initial testing, after re-soldering some misplaced connections everything was fine!


Have to install some cooling fins, bullet connectors and the fuse still.

I'll take some scope shots when I get a chance, and also get a REAL DC motor! :

Hi Adam!

Glad to see that your project is progressing! I look forward to the results of tests!

WBR, Nickolay

I will have time on Thursday to measure things, so until then I will write a small firmware for a controller to produce a PWM signal with varying frequency and duty cycle.

Thanks Adam. We expect good news from you.

i ve started to draw the board soon i will add the pics

Pin 11 is Shutdown. It is directly connected to pin 2 (Ground/Common) to enable the IC. I just drew it the way I did, electrically doesn't matter where they are connected.