# Trouble with transistors

hello dose anyone know of any type of transistor that would work with a 2.7v max source? I am trying to make a 2v-18v DC to AC inverter.

I was using 2 Mosfet transistors with a 60hz oscillator, but the problem was it was barely turning the transistors on and over all not efficient at all.

The other idea I had was to use a logic circuit in its place or a digital logic circuit with maybe a raspberry pi.

any input is very much appreciated.

Related Electrical Engineering News on Phys.org
meBigGuy
Gold Member
Bipolar transistors (as posted above) will do better at low voltages.

MOSFETs typically have pretty high turn on voltages and high gate capacitance.
But, you can look for low threshold power MOSFETs. like
http://www.vishay.com/docs/71692/71692.pdf

Digikey search will find high current MOSFET devices with logic gate inputs: (note you can select ranges of values for search).
http://www.digikey.com/product-search/en/discrete-semiconductor-products/fets-single/1376381?k=logic mosfet n channel

No idea what your circuit looks like, what currents you want to switch, how much voltage the switching devices will see, etc etc

Last edited:
DocKiru and Jeff Rosenbury
meBigGuy
Gold Member
He is trying to create 60Hz AV at 18V for some reason. Not sure if it is 18V PP or RMS, what power he requires, what topology (transformer based?) what he is going to do with it, whether square wave is ok, etc etc etc

DocKiru
Thank you everyone for your input. I am currently making a diagram on the computer and Ill post that as soon as I am done. I am also going to type an in depth response to what I am trying to make and do. Il get that posted latter today.

okay so what I am trying to attempt to do is power a Dayton 110v 8A max split phase A/C motor. If my math is right then this motor consumes .13333333 Amps per min; 7.9999998 Amps per hour, 880vA.

So my goal is to see how low of a source power supply I can get it to run off of (with no load on motor) and to try and figure out where improvements could be made to improve over all efficiency. the voltage I want to use for this is 2.4v-2.7v NiMh M cell's in parallel to get 25,000AH and (29) 2.7v 3,000F Super caps in parallel to make a hybrid battery.

so from the battery I was thinking about using a LDO 2.7v Regulator then go into the inverter and I have found a few different ways of making the inverter, but they are all really close to the same. I was going to use a oscillator to make a 60Hz modified sinus wave. Then I was going to use a transformer to step up the 2.7 voltage to around 8v and go threw another transformer to go from 8v to 110v or just do one step up transformer to go from the 2.7v to 110v then to (20) parallel 110v 1200 mFd ac run cap bank then everything goes into a motor controller. from the controller the excess current not consumed or what is not going towards the motor is redirected into another cap bank that goes to a small generator that produces 5.5v then threw a four way bridge rectifier that produces 5.5v and then that threw a burp charger to recharge some of the battery during operation to extend the run time.

also I am still trying to figure out a way to use the remaining current that is going to ground to charge caps to be used again. (this is one of the hardest thing i came across to try and do)

When I get done making the full diagram I will post it its not going to be perfect because i am making it in adobe illustrator; at the moment I am trying to find all the right parts to use for the inverter and then ill move onto the rest.

I am sorry if it all doesn't make a lot of since or if it seems like its scattered all over the place. but that is a quick run down on what I am doing

NascentOxygen
Staff Emeritus
There is no "excess current" in the proposed arrangement. There is no point in using the motor to drive a generator to partly power the motor; far from extending the motor's runtime from a particular battery, this will reduce the run-time by the mere fact of introducing additional losses.

A 110V 8A motor when powered from 2.4V via inverter circuitry will draw at least 370A from that 2.4V source at full load. At no load it will draw much less, but driving a generator as you envisage is not a "no load" condition. When you load a motor it draws extra current, in accord with "power out = power in".

Sorry, but there is no such thing as a free lunch.

meBigGuy
Gold Member
I totally agree with NascentOxygen that using motor power to drive a generator lowers the system efficiency by the efficiency of the generator.
The motor will only draw what it needs to power what it is connected to, and overcome its internal losses. The generator is a bad idea (sorry).

BTW, there is no such thing as .133 "amps-per-minute". Maybe it uses 0.133 Ah in a minute.

A class A 60 Hz oscillator running at 370 amps is a problem. Forget about doing that!!

You need to research power inverter circuits that use switching transistors. Essentially you drive the transformer with a squarewave.
There are lots of examples on line. This one (just an example) uses schmitt trigger inverters to drive 2N3055 transistors. http://www.circuitstoday.com/100-watt-inverter-circuit

One of the biggest efficiency issues will be running off 2.7V. The saturation voltage of the switching device and the time it spends switching are the primary power losses in your system. If you were switching 5.4 volts, for example, the current would be half, and the saturation voltage lower. The power loss at a constant voltage varies as I^2.

But, putting batteries in series to get 5.4V presents its own problems (especially with charging).

Maybe you could switch +2.7 and -2.7 (two banks, grounded in the middle) and a dual charging system. (maybe that's how people do it now? no idea)

DocKiru
rbelli1
Gold Member
maybe that's how people do it now?
All of the lead acid based UPS's I have worked on just stack the batteries and pray. I would hope the lithium ion ones have some balancing involved. With NiMh/NiCd it would probably depend on the charger.

BoB

all the lithium battery's are very picky in the way they want to be charged and if you don't charge them right then you destroy them not to mention that they could explode and cause a lithium fire which is scary but when you charge them they have to have a balanced charger so yes they do and most Nimh/NiCd also do but it dosent work so well the NiCd/NiMh because the balance charger doesn't monitor the temperature which is more important; so that why I am building a fast burp charger that can charge up to 16 cells at a time and it is monitoring both temp and voltage.

By no means is this project about getting a free lunch this is a resurch project that I am doing and the first part of it is a no load situation, where it is hooked to nothing, no generator no nothing. the first part is all about just getting the motor to run off of that voltage and improve upon it then go into the second phase of it which is where I will add a load to it.

BTW, there is no such thing as .133 "amps-per-minute". Maybe it uses 0.133 Ah in a minute.
Thank you; yes 0.133 Ah per min.

One of the biggest efficiency issues will be running off 2.7V. The saturation voltage of the switching device and the time it spends switching are the primary power losses in your system. If you were switching 5.4 volts, for example, the current would be half, and the saturation voltage lower. The power loss at a constant voltage varies as I^2.

But, putting batteries in series to get 5.4V presents its own problems (especially with charging).

Maybe you could switch +2.7 and -2.7 (two banks, grounded in the middle) and a dual charging system. (maybe that's how people do it now? no idea)
I don't see why this couldn't work, but I'll have to look into that more.

rbelli1
Gold Member
Maybe you could split this into two parts. First get a power supply that can be set to 2.7V and build your circuit. Once you get that worked out you can measure the current and design a battery system that is a suitable size.

You are talking about HUGE currents here and if you estimate wrong you will either be very disappointed at the results or you will have spent way to much money on the batteries. Overbuying the power supply just means that your next project can be bigger than this one.

BoB

Maybe you could split this into two parts. First get a power supply that can be set to 2.7V and build your circuit. Once you get that worked out you can measure the current and design a battery system that is a suitable size.

You are talking about HUGE currents here and if you estimate wrong you will either be very disappointed at the results or you will have spent way to much money on the batteries. Overbuying the power supply just means that your next project can be bigger than this one.
thank you I think that is a great idea and would be the smarter thing to do. I think I will do that first. Thank you rbelli1 :)
also the current I will be dealing with is also lethal.

I have HV gloves and insulated tools for this I am big on safety

rbelli1
Gold Member
I have HV gloves and insulated tools for this I am big on safety
Those are for working on or near live circuits. It is best to put suitable bleed resistors on your capacitors and never touch anything when it is live. Make sure there are power lamps on all voltage rails and don't touch it if any lights are on.

The battery pack is another story. It is in part always live but the voltages you have there are not dangerous. Just don't get any conductors across the terminals.

BoB

meBigGuy
Gold Member
Get your wedding band across a battery bank and ...... ouch

I can't believe I wrote that power is I^2 at constant voltage. I meant constant resistance, like a switching fet.

The lower the voltage you start with, generally the worse your efficiency, since currents are higher for a given power. Power loss in transistors and transformers is very current dependent.

hello dose anyone know of any type of transistor that would work with a 2.7v max source? I am trying to make a 2v-18v DC to AC inverter.

I was using 2 Mosfet transistors with a 60hz oscillator, but the problem was it was barely turning the transistors on and over all not efficient at all.

The other idea I had was to use a logic circuit in its place or a digital logic circuit with maybe a raspberry pi.

any input is very much appreciated.
My recommendation would be a hexfet ( Hexagonal Metal-Oxide Field-Effect Transistor Read more at http://acronymsandslang.com/HEXFET-meaning.html). Such as the irfz46n hexfet. This hexfet runs very close to a mosfet, the main difference is that these can handle high currents and high switching speeds and they require a low turn on voltage to the gate. I use these for inverters and amplifiers.

berkeman
davenn
Gold Member
2019 Award
Then I was going to use a transformer to step up the 2.7 voltage to around 8v and go threw another transformer to go from 8v to 110v or just do one step up transformer to go from the 2.7v to 110v then to (20) parallel 110v 1200 mFd ac run cap bank then everything goes into a motor controller.
That's not going to work either
You need to learn how capacitors work

NascentOxygen
Staff Emeritus