First attempt at designing a low input voltage boost converter

[Gasboss775]

Of course, another point is that you probably couldn't have a source sense resistor anyway as it would drop too much of your precious v(in).

I was thinking about using a very low value current sense resistor 0.033R or 0.022R and an amplifier than can have inputs down to zero, probably CMOS which could also accommodate the low supply voltage also.

Probably the MCP602 as I have a couple of these.

 

[Flyback]

yes I suppose you can amplify a low signal, of course, if your F(SW) is too high then the amplified voltage may be nothing like the signal, but then all you have to do is reduce f(sw)...and anyway, with current mode control, the current signal doesn't have to exactly replicate the actual signal, as your error amplifier will take account.
I wonder if analog devices might have something for you in that respect of amplifying low signal

 

[Flyback]

But how about cheating and using , as discussed, a low vin boost controller such as MIC2571 to give you a nice healthy bias voltage rail of 5v or so. (Or more if you want, it can go up to 36v output) ..then its a simple job to get a controller which can use that to boost the batt voltage up to 9v.

MIC2571
**broken link removed**

mcp602 may do the job yes, but from the fact it only draws 230ua iq you can tell that its not one of the faster opamps.....as you know, fast opamps always draw more bias current

 

[Gasboss775]

I wonder if analog devices might have something for you in that respect of amplifying low signal

If possible I'd prefer to utilise parts that I already have if possible and NO smd, as I have shaky hands and these tiny components are near impossible for me to use!. Of the opamps I have, I think Ca3130, Ca3140, lm358 ( probably too slow ), mcp602, lmc6842, all have inputs that can go down to ground. I suppose that opamps that require a higher supply voltage could be powered from the output. The circuit just needs to be able to start up reliably.

 

[Flyback]

see your point, and the current sense resistor will help limit staircasing current at startup, or should be able to.
Another point is that you could try "constant off time control", if you are going to use a sense resistor, since current mode control will require a slope compensatory ramp if you are going to do it "properly".
since COT uses just more or less a comparator and an error amp to get smooth control...........it doesn't need an oscillator as the ramp of the inductor gets kind of used as the 'oscillator'.
im sure you understand cot control, its just switching up to the sense current limit, (or the amplified sense current limit) then you go off for a fixed amount of time...and benefit is it needs no slope compensation.

 

[Gasboss775]

im sure you understand cot control, its just switching up to the sense current limit, (or the amplified sense current limit) then you go off for a fixed amount of time...and benefit is it needs no slope compensation.

Actually I'm not familiar with this topology but thanks for bringing it to my attention. My adventures with switched mode power supplies only began in mid October last year, so I've still got much more to learn.

 

[Flyback]

im just wondering if youd be better using a current sense transformer than an opamp....there are unipolar setups which can be used with a boost converter.....you just have to be sure and reset the transformer in every switching period...........it amounts to little more than a resistor, and diode, the CST itself, and the burden resistor....put a D/zener in aswell for extra protection in startup or transeints.
IF YOUR WORRY IS SWITCH CURRENT STAIRCASING AT STARTUP, THEN MAYBE YOU COULD JUST SKIP THE CURRENT SENSE RESISTOR, AND USE A MICRO TO SWITCH THE FET, AND DO IT IN SUCH A WAY THAT IT GIVES YOU A SOft start (sorry about the capitals, im rubbish at typing)........anyway, you could get the micro to just do short fet on pulses at first, spaced wider apart, so that your inductor current cant staircase up too high at startup...then when sufficient vout has built up, the vout itself will do a good job of stopping staircasing of the inductor current, because the vout-vin will be across the inductor at fet off times, and will force the current in the inductor to ramp down suficientlty.

 

[ronsimpson]

two cell to 5V, Change RF1 to get 9V.
Current is most of 1/2 amp.

 

[Flyback]

looks great, but OP only wants thru hole parts...lm2623 is only 4 by 3mm

 

[Gasboss775]

two cell to 5V, Change RF1 to get 9V.
Current is most of 1/2 amp.
View attachment 96751

Thanks Ron. I don't want to use a one chip solution as I want to learn more about different switch mode topologies, through practical experimentation. The circuit is not intended for any specific task but rather for the challenge of learning about a new type of circuitry.

__________________________

Apparently I have just broken the 100 post barrier. Where do I collect my prize?

 

[Flyback]

Good luck, boosting up from potentially as low as 1.8VDC to 9VDC is setting yourself quite a tough challenge for starting off...specially since you want only thru-hole components, and only wish to use a single stage, and appear to want a smooth controlled flow of power (smoothly switching pulses).
On the other hand, if you do it as on/off with big cout, then I suppose its doable, in fact as you already seem to have shown.
Why not just have an overcurrent comparator instead of full current mode control via the sense resistor....ie stop flyaway fet current due to staircasing.

 

[Gasboss775]

Why not just have an overcurrent comparator instead of full current mode control via the sense resistor....ie stop flyaway fet current due to staircasing.

I think I might try that tomorrow.

I also think that I'll switch to using 3 cells ( 2.7 V minimum ) to reduce requirements somewhat. As I said elsewhere the circuit is not intended for any specific role it's about learning through doing, so Iam at liberty to move the goalposts if necessary!

 

[Flyback]

Too right,
By the way, powering stuff like 9V LED lamps from a single NiMH cell or two, with a single stage, is a bit of a “golden chalice” in electronics.
Few have succeeded with great results.
The Chinese would be doing it if it was possible to do it without recourse to specialist integrated circuits like Ron Simpson suggested.
The Chinese are not doing it, indicating that its not really practical without Ron’s IC type.
However, you deserve a crack at the chalice for succeeding to actually get the 9v

 

[Gasboss775]

Too right,
By the way, powering stuff like 9V LED lamps from a single NiMH cell or two, with a single stage, is a bit of a “golden chalice” in electronics.


However, you deserve a crack at the chalice for succeeding to actually get the 9v

Thanks.

Another boosting topology that's worked with low voltages is using an lm386 as a square wave oscillator feeding a diode capacitor multiplier. I never managed 9 volts, I expect this would only be possible with very low current drain, perhaps hundreds of microamps output at best.

 

[Gasboss775]

Good luck, but I believe the only really effective way to do a boost converter from 2 NIMH up to 9V (?power level) is to use a “Pilot” mini boost converter to give you an “at_least_4V” bias rail which can be used to power the main boost pwm controller.

That sounds plausible but I am thinking that powering from 3 cells could make the task a lot more manageable.

 

[Gasboss775]

Do you think this could work? Its my first attempt to design in current mode.

 

[Gasboss775]

MIC2571

Another great new product spoiled by only being available in SMT. Bah!

 

[ronsimpson]

Another great new product spoiled by only being available in SMT. Bah!

 

[Gasboss775]

The problem of soldering it onto the adapter still remains sadly.

 

[ronsimpson]

I do it all the time.