Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

H Bridge trouble

Status
Not open for further replies.

amsm

New Member
I am building a CNC router. I have the whole thing essentially done, except the control of power to the motors. To this end, I am trying to build H-bridges, which I have never done before. I acquired some MJE3055T power transistors, and I've looked at the basic literature. I constructed the standard H-bridge, but it isn't working. Many of the schematics I've seen for power H-bridges have both PNP and NPN transistors in them. Is this a requirement? The data sheet says that the gain should be between 5 and 100. A few that I have are around 250. Does the gain have to match all around? Is the low gain going to be a problem? I tried to use a single transistor to turn on a motor. When I arc the base to positive, the motor will run when it is below the transistor, but not above it. I am so close to finishing this. I am short on time and money, and this is to supplement the family income. Please help.
 
When I arc the base to positive, the motor will run when it is below the transistor, but not above it.

You know the emitter is the one connected to the motor, and the collector to the positive supply rail when it's on top like that, right? I've seen people get this one backwards.

But it's a little tricky even wired correctly. The base voltage generally needs to be higher than the main motor supply voltage for that NPN transistor when it's at the top of the H-bridge like that. It seems like it SHOULD turn on (albeit with an extra .7V drop) but it starts sucking a lot of current through the base and through whatever you are driving it with and causing a big voltage drop on the high side.

Part of the problem is The 3055 has LOUSY gain, you can only count on an hfe (beta, current gain) of maybe 10-15. Measure it with a meter - you will probably see some positive voltage on the motor, but too low to get it moving (stall and start are like 10x or more the running current on motors).

One way to deal with it (back in the days before MOSFETS and IGBT's) was to actually build a separate supply just to drive the bases of the final stage transistors. Even with a PNP-NPN H-bridge you would sometimes do that, just to lower the power burned off in the base current limiting resistors. In this case you probably need a good 5 amp supply sitting 3-5V above the main positive supply for the motor.
 
I tried uploading a pic of the circuit I am working with - take a look in my profile if you are curious. Looks like I was too hasty in my choice of transistor. I have to pull 10.5 amps peak very seldom, but nominally about 5 to 7 amps. My main concern is for the Z-axis of my machine. It's gonna draw a lot during a jog, pulling the Z carriage back up. I'm not really concerned about efficiency for this iteration, I'll worry about that next time around. Could I just drop a power resister in front of the positive of the H-bridge and be done with it, or is it more involved than that? What value would you suggest? Could I use the carbon rod out of an alkaline battery for the resistance?
 
How much current can I expect going into the base of the MJE3055T to drive it? The data sheet says 6 amps, which seems absurd! I MUST be looking at the wrong spec!
 
Last edited:
6 amps is the max. You probably only need 1 amp. What is your voltage?

Your schematic is horrible (please use real symbols next time) but it looks like you have the motor connected to the BASE on those two high-side drivers! They are probably destroyed now. Also never "arc the base to positive" without using a current limiting resistor on those low-side drivers or you will wreck them.

The Z-axis rapid traverse up is always the hard one (I have a MAXNC-10CL). Nobody ever puts a constant-force spring in these machines to counterbalance the weight of the spindle assembly, they just expect the motor to haul it up.
 
Yeah, I drew it in openoffice, sorry.

I'm using 12 volts. My motors pull 10.5 amps wide open, but that's at about 180 inches per minute. I won't run them this fast too much because the leads wip really bad, and there is alot of vibration.
The pinout for these is B-C-E, but still, I think you might be right. They got pretty hot soldering and desoldering. On my meter, two have a Q of 50 and the other two are around 245, or so. Some of the others I haven't used are this high too, though. It seems kind of weird.

I'm needing three H-bridges, obviously. I have 12 3055s, and enough diodes to protect them. I have to get a 5 volt Arduino signal to open the 3055s. I was thinking I'd drop the voltage through a resistor to the H-bridge positive and run 12 volts to the bases through smaller transistors, but the only other transistors I have are the puny bulk Radio Shack kind you see in transistor radios. Can run these in parallel at all?
 
The pinout for these is B-C-E, but still, I think you might be right. ...

Look at your diagram, if that base is on the left, 2 are connected to the motor.

I have to get a 5 volt Arduino signal to open the 3055s.

CLOSE them. This transistor is normally off. If you have one that's on all the time, it's probably blown.

Voltage isn't the problem so much as the current. You need a midsize driver that can handle a couple of amps. I don't know why, but these are usually harder to lay your hands on than the big stuff or the small stuff. Use Digikey or Mouser.

Don't trust those unloaded beta readings too much - trust the spec sheet. It may tempt you with a high gain reading, but under load you get disappointing results that match the numbers on the spec sheet.

the only other transistors I have are the puny bulk Radio Shack kind you see in transistor radios. Can run these in parallel at all?

Yes - you have to give each one it's own resistor, though. Each bipolar transistor has a slightly different voltage drop, the lowest one winds up shouldering almost the whole load. The resistor balances this out.
 
Dude! I just got an idea!

See, my whole problem started because All Electronics mislabeled their product. I bought these motors, thinking they would only need three amps. So, I got these two L298Ns to drive them, as well. I finally get to the point where I'm hooking the motors to the chips, and they just hum and twitch a little. I hook the motors to the power directly, with a meter in-line, and the motors are pulling 10.5 amps!

So, I have these duel H-bridge chips. Can I use them to drive the MJE3055Ts? They can handle 2 amps, 3 peak.

How far above the motor supply would I need to drive the bases?
 
Yeah, that looks like a good idea. I'm a little concerned about the PIV rating on the 3055's base with that approach. The base on your 3055 can only handle like -7V before it avalanches... but you could fix that with a diode, and it still looks like a smart way to do it because it gives you the logic level conversion and those important gates to lock out the "shoot through" state (very important! never trust a couple of port pins to do it).

>>How far above the motor supply would I need to drive the bases?

Only a couple of volts (1.8V @ Ic = 4A), but there's a catch - the L298N drops a couple of volts, too, and the diode will add close to another volt, so Vs on the L298N needs to be a good +5V or more above the 12V rail to saturate those drivers. Plus you want a little over so you can put a dropping resistor in it. Say 20V.

The NPN on the top is going to get around -12V on it's base using the bridge, so you need a diode.

See attachment for what I'm suggesting. Notice the size of these dropping resistors on the base - two of them are 5W and the other two need to be 20W!

I love All Electronics, bought many things on there, but like any surplus place you have to just be grateful for any specs they give you but not take them too seriously. The motor specs are the unloaded running current (if they even give you that) so figure 10x for start and stall and something in between for loaded running current.
 

Attachments

  • hdrv.GIF
    hdrv.GIF
    8.5 KB · Views: 189
Thanks for drawing that up for me, duffy! I appreciate that! BTW, did you do this in Eagle CAD?

I went to school for electronics, but it was way back in '89. I've only used it when I needed it, so I'm pretty clumsy. Looks like it's coming back around again.
 
Could I use the carbon electrode out of a flashlight cell for the power resistors? What about "D" size batteries?
 
Last edited:
I used Orcad.

I'm not so sure about that mythical carbon electrode in the flashlight cell. As a kid, I had more that one jackass tell me about that thing and get me to waste time in the dangerous activity of prying open a battery full of caustic poisonous sh!t only to find out there's no carbon rod inside.
 
I've acquired a few myself. I was going to use them in my electric furnace until I realised that Cal-Rod glows red without any harm to it. I tried both battery carbon and pencil lead. I learned that carbon will vapourise pretty quick if there is too little carbon and too much power.
Some batteries have carbon and some have a metal (stainless?) rod instead. I'm just don't know if this circuit will generate enough power to burn the carbon.
 
Dude! Your circuit works great! Only thing now is that my 298s went bad somehow. I got ahold of some TIP31Cs. I'm going to put them between the Arduino and the H-bridges.

The literature says that the TIP31 has a gain of 25 at 1A. The pins on the Arduino are only capable if 40mA, whereas the H-bridges take an amp or so. I obviously need a resister between the Arduino pin and the base of the TIP31. How do I go about calculating the value of that resister?

Is the current running into the base limited by the current being drawn through the emitter?
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top