MC34063: Is this an appropriate way of generating 5V 2A from a few Li-ion cells?

[technix]

MC34063 can withhold only up to 1.5A but to pump out 2A I need something beefier. In this circuit I kept the MC34063 part intact but the output transistor is, instead of actually handling the current, connected into a makeshift low impedance MOSFET driver 74LVC1G04 (marked as 'HC1G04) which in turn drives a 40A MOSFET IRF7470. The The inductor have 100uH of inductance at 5A, and the Schottky diode handles 3A on average.

My question is, will this circuit work? Or will I burn something?

By the way, can I cook up a SMPS doing the same job by using a 2.5V reference like MC1403, a comparator like LM339, a power MOSFET like the IRF7470 in question, and a crystal oscillator that have an output enable pin (or a crystal resonator, a 74LVC1GU04 and a 74LVC1G08, a resistor and two caps)

[bktemp]

Never use a 34063 in stepup mode without current sensing. Otherwise something will blow up when used at higher power levels.
Better get something more modern. There are lot of much better, more efficient and smaller stepup converter ics.

[technix]

Never use a 34063 in stepup mode without current sensing. Otherwise something will blow up when used at higher power levels.
Better get something more modern. There are lot of much better, more efficient and smaller stepup converter ics.

If this is your suggestion then please show me a way using a chip that costs less than three US cents each in quantity of 50.

By the way, the 0.04 ohm sense resistor can be achieved by using an ultra thin trace on the PCB.

[darkness_electro]

Use an external mosfet connected trough the output pin. For best and quickest switching and to increase efficiency use a push pull bjt gate driver. Or at least a pnp pull down transistor. Will do wonders.

[tszaboo]

Never use a 34063 in stepup mode without current sensing. Otherwise something will blow up when used at higher power levels.
Better get something more modern. There are lot of much better, more efficient and smaller stepup converter ics.

If this is your suggestion then please show me a way using a chip that costs less than three US cents each in quantity of 50.

By the way, the 0.04 ohm sense resistor can be achieved by using an ultra thin trace on the PCB.

I would never dear connecting a MC34063 to any battery. That thing seriously lacks any kind of protection. Including UVLO, so killing your battery or going up in flames when the voltage drops is too dangerous. What happens when the voltage drops below 3V? The inductor will conduct, the diode will conduct, your output will be 2,5V or so. Powering whatever you have there.
We dont need yet another cheap crap quality Chinese product.
Learn for the west: Safety is priceless.

[technix]

Never use a 34063 in stepup mode without current sensing. Otherwise something will blow up when used at higher power levels.
Better get something more modern. There are lot of much better, more efficient and smaller stepup converter ics.

If this is your suggestion then please show me a way using a chip that costs less than three US cents each in quantity of 50.

By the way, the 0.04 ohm sense resistor can be achieved by using an ultra thin trace on the PCB.

I would never dear connecting a MC34063 to any battery. That thing seriously lacks any kind of protection. Including UVLO, so killing your battery or going up in flames when the voltage drops is too dangerous. What happens when the voltage drops below 3V? The inductor will conduct, the diode will conduct, your output will be 2,5V or so. Powering whatever you have there.
We dont need yet another cheap crap quality Chinese product.
Learn for the west: Safety is priceless.

I have a separate battery protection already in place and the DW01A-based two-chip solution is transparent to the charger and the powered device, so it is being irrelevant here. I may originate from China but I am also fed up with local el-cheapo. This is part of my Raspberry Pi UPS project (because I use Raspberry Pi as router so I need to make sure it stay powered, and most power bank just won't cut it due to internal switching when charging.)

[technix]

Use an external mosfet connected trough the output pin. For best and quickest switching and to increase efficiency use a push pull bjt gate driver. Or at least a pnp pull down transistor. Will do wonders.

Schematic example please?

Also in this circuit I am driving MOSFET with CMOS output so I think it will work better than your solution.

[darkness_electro]

Just copy the part atached to pin 2.

[technix]

Just copy the part atached to pin 2.

I cannot afford that much voltage drop across BJTs. This is why I went CMOS as fast as possible once outside the chip, using the 74LVC1G04, a CMOS chip, to drive the chopper MOSFET.

[bktemp]

You could use a BJT instead of the mosfet. The efficiency won't be great, but it will be cheap.

[technix]

You could use a BJT instead of the mosfet. The efficiency won't be great, but it will be cheap.

Well I am going to use AO4606 instead of IRF7404 here because it saves part count, and AO4606 is cheaper than a lot of IRF and Fairchild power MOSFETs in SO-8 package. Remember the perfect diode circuit Raspberry Pi requires when back-powering it that calls for a PMOS, and this boost circuit calling for a NMOS? AO4606 happens to contains one of each. I can use the NMOS as the chopper and the PMOS as part of the perfect diode circuit. (By the way, I also happen to have a few BCV62 matched PNP pair with C1, B1 and B2 internally connected, which is the other semiconductor component I need for the perfect diode circuit, and thanks to the greater integration, the board can be a lot simpler)