Go Down

Topic: MC34063A external MOSFET boost converter output voltage rapidly rises (Read 462 times) previous topic - next topic

TomGeorge

Hi,
If you look at the mc34063A datasheets, you will se that they have a recommended PCB layout.
Because of the currents involved and the frequencies and the fact that most are squarewave, the layout of the circuit is critical to proper operation.
If yours is basically long wires and plug type connections, then you are pushing uphill for a result.

SMPS are hard to design and make from prototype, and the fact you do not have a scope to check your circuit operation will hamper your results.

Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

kaseftamjid

Hi,
If you look at the mc34063A datasheets, you will se that they have a recommended PCB layout.
Because of the currents involved and the frequencies and the fact that most are squarewave, the layout of the circuit is critical to proper operation.
If yours is basically long wires and plug type connections, then you are pushing uphill for a result.

SMPS are hard to design and make from prototype, and the fact you do not have a scope to check your circuit operation will hamper your results.

Tom... :)
Okay, so reliability is a definite issue when dealing with long wires. I tried to reduce this by separating the actual boost converter part (inductor-MOSFET-capacitor-diode, basically the things with high current passing through), So the only "long" wire is the 4cm long wire that connects to the gate of the mosfet from the mosfet driver. So i hardly doubt that the long wire is a issue right now.

Even so, could you confidently say that long wires are causing this problem for me? My ultimate goal is to make a pretty pcb according to spec to bring everything in one place, but to do that, i have to make this boost converter part work first.

Judging from the question before, I may be a beginner, but im not a complete noob to make silly mistakes that i think is right, like not using correct traces/wires for high current applications. 

TomGeorge

Hi,
This "inverter" will not properly invert, how do you think it works?



Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

TomGeorge

Hi,
If we simplify your inverter, you do not have either of the BJTs biased properly.



Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

raschemmel


I don't se anything pulling up the TC4427 inputs.
(no pullup resistors ) so where is the voltage comin from ?
Did you measure it on the inputs ?


kaseftamjid

Hi,
This "inverter" will not properly invert, how do you think it works?



Tom... :)
Hi,
If we simplify your inverter, you do not have either of the BJTs biased properly.



Tom... :)
This is a bit weird... okay, I just ran a few simulations and am wondering how I missed this, anyways, the more you learn. The reason i 'used to think' this is an inverter circuit is because i temporarily ran my mosfet with this configuration, and the mosfet would turn on when the input was low, Being lazy, i didnt think of checking that twice, look where that got me.

I have this alternative plan of using a simple 2 npn nor gate as not gate. like this
I tested this on simulation, and works fine with both inputs tied up.

Am i missing anything? i think i might use higher value resistance, how much does resistance affect the switching speed of the transistors? should i be concerned about this?

P.S, i am using 1nF timing capacitor, which in turn gives me about 35-40 Khz of switching frequency. The transistors i have, d882, can handle up to 1Mhz. So, what effect does resistance have on switching speed?

kaseftamjid

By the way, this would explain the voltage rise, the pulse width of the ic changed but the output was inverted, for now im surprised that the faulty circuit provided any output at all!

kaseftamjid

This is a bit weird... okay, I just ran a few simulations and am wondering how I missed this, anyways, the more you learn. The reason i 'used to think' this is an inverter circuit is because i temporarily ran my mosfet with this configuration, and the mosfet would turn on when the input was low, Being lazy, i didnt think of checking that twice, look where that got me.

I have this alternative plan of using a simple 2 npn nor gate as not gate. like this
I tested this on simulation, and works fine with both inputs tied up.

Am i missing anything? i think i might use higher value resistance, how much does resistance affect the switching speed of the transistors? should i be concerned about this?

P.S, i am using 1nF timing capacitor, which in turn gives me about 35-40 Khz of switching frequency. The transistors i have, d882, can handle up to 1Mhz. So, what effect does resistance have on switching speed?
EDIT: Its been a long day and i just realized that this is the thing i tried to avoid at first, with the pullup, anyways, ill try the simple way, with 1 transistor and see if its good or not.

raschemmel

I haven't seen you acknowledge thatthere's nothing pullingup the inverter outputs

kaseftamjid

I haven't seen you acknowledge thatthere's nothing pullingup the inverter outputs
I acknowledged the whole thing was wrong, including the wrong that the so called inverter i have been trying to use has no pull up

Go Up