H-bridge inverter, N only

wayneh

Joined Sep 9, 2010
16,102
Here's an H-bridge circuit I could use some commentary on, as I've never built one. I'm trying to use components I have on hand, and stick to N-channel MOSFETs. Just wondering how likely this is to fail due to shoot-thru or other problems.

Not shown in the schematic:
Inductor load is a ~10:1 transformer, which I want to run at ~1A. Is there a transformer in LTSpice?
My MOSFETs are IRF540N, by I used IRF530 because that one is in LTSpice
Frequency range = 400-2000Hz, set by R2. I'll use a pot there, with fixed resistors to limit the range.

The diodes D1 and D2 form a Dickson charge pump to get the voltage up for the high side switches.

The goal is a DC-AC inverter to drive some EL (electroluminescent) tape. My plan is to use a 555 oscillator driving an H-bridge to pulse current thru the secondary coil of a wall wart used backwards, to get ~120VAC and ~100mA at the output.

Will the charge pump provide enough juice to drive the high side?
Is there any problem using both the 555 Output and Discharge outputs? I don't see this done very often but it seems a handy way to spare an external transistor.
Any suggestions for component values? (Don't worry about the timing components - I can handle those.)
Anything I'm missing? (More gate resistors?)

Attachments

• 67.1 KB Views: 273
Last edited:

wayneh

Joined Sep 9, 2010
16,102
PS: This is my first LTSpice project of any size. Can I actually get a simulation out of this? Obviously I'll have to specify the components.

I've attached the LTSpice file if anyone cares to tweak it.
View attachment My inverter2.asc

I haven't drawn any flyback diodes yet.

Last edited:

ronv

Joined Nov 12, 2008
3,768
Wayneh, Can you attach it as a dot asc file instead of a png file?

wayneh

Joined Sep 9, 2010
16,102
The attachment in #2 is not a .asc?

ronv

Joined Nov 12, 2008
3,768
It shows up as a .php which I think is a png file.
The asc you can find at LTC/LTspiceIV/then the file with the transistor.
There are some transformer examples at LTC/LTspiceIV/Examples/educational.

ronv

Joined Nov 12, 2008
3,768
Hmm. Looks like the site doesn't like asc files.

wayneh

Joined Sep 9, 2010
16,102
Odd. It leaves here as an asc file and appears that way when I view the post above.

Anyway, here's a zipped version of the same file. Maybe it will better survive the transit.

View attachment My inverter2.zip

ronv

Joined Nov 12, 2008
3,768
Worked like a champ! Let me play a while.

wayneh

Joined Sep 9, 2010
16,102
Hmm. Looks like the site doesn't like asc files.
Perhaps the new server is blocking them? Wild speculation, but there is a new server.

ronv

Joined Nov 12, 2008
3,768
Pretty good!

I picked some diodes, (just right click on the component),added a value for L, added a ground, added u after the cap values and reduced the battery resistance. I think the cap for the charge pump might need to be bigger and it needs clamps, but other than that I think it will run so you can see shoot thru and stuff. I'll look a little more, but it looks like your well on your way.

ronv

Joined Nov 12, 2008
3,768
Hmm, I always thought discharge was inverted from the output but it isn't. I guess you will need an inverter for one side or the other and a little work on where the base resistors go. I made the storage cap smaller for the charge pump to make it easier for the 555. I really like this circuit!

wayneh

Joined Sep 9, 2010
16,102
Nuts, you're right, I misunderstood the description that the discharge is "in phase" with the output. I thought that meant it went to open collector when the output is high, but it's the opposite. It was so clever the way I had it!

Glad you like the circuit. I think it may help others, since I was able to find a lot of folks asking for an all N-channel H-bridge, and there were not very good solutions. This one will be capable of quite a good current if I can get it working for real as well as it looks on paper. On the other hand, there are H-bridge ICs that would be better if you don't already have a pile of parts to scavenge.

Were you going to attach your improvements?

ronv

Joined Nov 12, 2008
3,768
Sure, here is what I did. You can probably do better given some time. I think you will love spice.

Attachments

• 1.4 KB Views: 27

wayneh

Joined Sep 9, 2010
16,102
Haha, I most certainly do NOT love it! It's so incredibly unintuitive and un-Mac-like. But I'm learning to use all those "function" keys I never used before!

I don't love it, but I do appreciate its power. I just haven't learned to harness that yet.

Thanks for your input. I'll keep tweaking tomorrow.

ronv

Joined Nov 12, 2008
3,768
I wonder if it could be made to work for a single pulse?
Yea, wait till you start adding components. %$#@** Thread Starter wayneh Joined Sep 9, 2010 16,102 You removed any resistance between the 555 output and the charge pump or the MOSFET gate (was R1, 47Ω in my drawing). That was a question I had. None needed? I understand the added base resistors. I was planning to change D1 and D2 to Schottky diodes, so thanks for that and all the flyback diodes. Q1, Q2 and M5 - I'll just be using whatever I have on hand but it's good to have parts named. The 555 discharge pin 7 can just float open, I think? What does the ".tran 100m" at the lower left mean? Total noob here. Thread Starter wayneh Joined Sep 9, 2010 16,102 I wonder if it could be made to work for a single pulse? Yea, wait till you start adding components. %$#@**
Huh? I guess I don't understand either sentence. But I did just have a nice grasshopper cocktail with mint from my garden.

ronv

Joined Nov 12, 2008
3,768
The tran 100m is how long the simulation runs. When you press simulate it will ask you haw long you want it to run. I used 100m (ms) so I could see the rise time on the charge pump.

The other 2 sentences--- I was thinking maybe it would work for things like dc motor control if the 555 ran at say twice the frequency of the pwm so it had time to charge before the top fet was ask to turn on. This may be worth some time and effort.

The second was about the syntax for spice - ugly in my mind.

wayneh

Joined Sep 9, 2010
16,102
I just read something that blew my mind. If you replaced the simple 555 clock in my schematic with a PWM signal under control of a micro, you could simulate a sine wave to the transformer. Something to think about after I get v1.0 working.