Hello there. I am currently trying to design a basic boost converter which will utilize a single AA alkaline battery and step up its voltage to 5v (at around 150 mA ideally). I believe I have a basic understanding of how such a circuit works and of the design equations needed. My current problem lies in selecting a suitable power mosfet to serve as my switch, one that could operate via the 1.5v source. If such a device doesn't exists, what kind of mosfet driver circuit would work best? If any additional information would be helpful, please let me know. Thanks in advanced.

 

The Sanyo P/N: MCH3484 (an On-Semiconductor Co) MOSFET shoud do. It's gate threshold voltage is only 0.8V, so a 1.5V battery should work well even down to about 1 V. It's a 4.5A device, so 150 mA is no problem. It is a surface-mount device (SOT-70 / SOT-323), but it's not so small that you can't breadboard with it. For the output diode, you'll need a very low voltage drop shottky diode like the 1N5817 or 1N5818. I don't know what you're going to do for the logic/oscillator/control IC. There are low-voltage types that should work from a 1.5V Battery, but you didn't ask that so I assume you have that under control.

Good luck,
Kamran Kazem, V.P., CTO
Magnetic Design Labs, Inc.

 

Kamran, thank you for your quick reply. I will try to simulate my converter with the suggested parts and respond with the results as soon as I can. As for the control, I was considering some sort of astable oscillator. Perhaps there is a simpler/more elegant solution to this in the form of an IC?

 

i used a small IC for a boost convertor, i only needed 12V at 50mA and while it took two pages of calculations its damned good..

 

LTC3526L boost converter (with integrated switch, I suppose) from Linear can operate from 0.7V, but the max o/p current depends on the input voltage. Might just be able to deliver the 150mA at 1.5V!

 

I apologize for my previous post, as I should have been more clear. My goal is to design my own boost amplifier as opposed to buying one off the shelf. When I mentioned a more elegant solution in the form of an IC, I was referring to the control for the mosfet switch. Ideally I would like to construct this myself as well, but I was looking for existing IC's that I might be able to base my design on. Thanks again for the suggestions, I should have simulations of the converter up by this afternoon.

 

So currently I'm using modelsim to test my boost converter. My output doesn't increase in voltage, but instead rings around my input voltage. The current through the inductor also appears to simply ring around 1.5uA, which seems strangely low. I think it might have to do with my power mosfet not turning on properly, as the voltage above the mosfet never seems to drop significantly. Increasing the voltage of the pulse does not seem to help things either. I've included some waveforms and my circuit, and I'll update if I figure out anymore.

 

Could you do something based on the Joule Thief?
http://en.wikipedia.org/wiki/Joule_thief

 

I've included some waveforms and my circuit, and I'll update if I figure out anymore.

There's no variation in even the last digit of the y value in your plots, and yet the trace is a wave. Something seems wrong.

 

Wayneh, you might be right. I think the spice model for the transistor I retrieved from the company's website might not be working correctly. I've been testing it separately from the rest of my circuit and I can't get the channel to conduct. I'm going to look for alternate models and try again.

 

I apologize for my previous post, as I should have been more clear. My goal is to design my own boost amplifier as opposed to buying one off the shelf.

You can build one from a LMC555 free running timer and a coil. Guaranteed to run on 1.5V.

 

Bountyhunter, thanks for the suggestion. Is anyone familiar with writing spice models? I was able to get a really detailed model of the MCH3484, but multisim didn't like it because it was a four terminal model. I did my best to adapt it to a three terminal model by writing in a sub circuit line, and while multisim accepts it now, the transistor still never turns on regardless of the gate voltage applied. I've attached the spice code I've been using. Any help would be really appreciated...

Also, I do still get an warning when running simulation. Apparently multisim doesn't recognize parameter NQSMOD

 

Bountyhunter, thanks for the suggestion. Is anyone familiar with writing spice models? I was able to get a really detailed model of the MCH3484, but multisim didn't like it because it was a four terminal model. I did my best to adapt it to a three terminal model by writing in a sub circuit line, and while multisim accepts it now, the transistor still never turns on regardless of the gate voltage applied. I've attached the spice code I've been using. Any help would be really appreciated...

Also, I do still get an warning when running simulation. Apparently multisim doesn't recognize parameter NQSMOD

Now you know why I always went to the lab and built it up. By the time they got the sim to work, I was already done with real bench data.

 

So it has been awhile since I provided updates on this small project, but I am happy to say that I have made some progress. It turns out that I was not setting up the spice models properly with respect to their symbols in Multisim. Fixing this made both me and the simulator happy. Below I've included my current circuit and output waveforms. My next step is to decide on a control IC (I have decided against the astable multivibrator as it is not easily adjustable) and to look into a comparator which will drop pulses when the output voltage gets too high (mainly when a new load is applied and the output is swinging). The output seems to stabilize fairly fast, though I'm not sure why I need such a large load in order to achieve my desired current (100-200mA).

Edit: Just a quick thought... would some kind of voltage snubber be useful for preventing initially higher voltages?