what's wrong with this circuit?

Hello fellas

I breadboarded the circuit in this magazine article:

formatting link
and eventuallly figured out how it works. In the text, the author never zeros in on what actually makes the circuit switch. All that's really happening is that the draw from the inductor sucks charge out of C2, turning off Q2 which flips the inverter U1a, turning Q4 off, then C2 charges up, the mosfet turns back on etc. I found this out because I tried to drive the circuit from a low impedance power supply set to .65 volts and it didn't work, but with a little resistance in series with the .65 volts it ran. I also found out that C4 has very little effect; I simply yanked it out of the breadboard. The circuit kept running, and the frequency changed less than 10%. The circuit below is a lot simpler than what's in the article and does exactly the same thing (view in monospaced font):

,------+------------------------------------------, | | | | 1R ) | | ) 160uH | --- solar ) | - 0.65v | --- ,--------+--------------------------|------------------+------, | - | | | BAT54 | | | | | ,-------, +-->|--+--2R--+----+ | | | +---|1 14 \ 1N4148 | | | | | | gnd | | \ |--' | | | | | 1M | 3 )o---+-->|----+--|| 100uF | --- | | | / | | |--, | | | batt - | +---|2 7 / | < |IRF530| | --- | | '-------' '------| | | 100uF - | / | \ | | | | | +---100k-+--| PN2222 | CD4093 2N3906 | | | | | | | | > | | | | | zener | | | | | | | | | | | 10uF 1M | | | | | | | | | | | | | | | | | | '--------+----+--------+-------------------+------+------+------+----+------' | gnd

What's funny is that I'm in a college EE class where the professor assigned us this circuit to build in the lab and do a report on. He posted the diagram but not where it came from. The professor repeated the explanation from the electronics design article. Everybody went in the lab, built the circuit and did reports on it, including me. It was actually our final project of the semester, and each lab group had to work up a power point presentation and give it in front of the class. While we were in the lab building the circuit, I figured out that C4 was pretty much pointless, the timing was done at C2 and said so during our presentation, but it was only last nigh that I twigged to the fact that three of the gates and a bunch of passives are useless as well. So this morning I got on the computer and was able to find where our prof got that circuit. I think I also located the author. There's a George Woolcott living in Harrodsburg KY, near Lexington, which is where the author's employer Lexmark has its headquarters. I'm considering contacting him and telling him our college class built his circuit as an assignment and ask if he would like to comment on it. But the semester is ending in just a few days. I don't have hiis email and would have to call him on the land-line number I found. I suppose I should just talk to the professor instead.

Reply to
Michael Robinson
Loading thread data ...

charge

then

simply

[snip mangled ascii art] Okay, that was too wide and line-wrapped. Here it is again: ,-----+---------------------------------------, | | | | 1R ) | | ) 160uH | --- solar ) | - 0.65v | | --- ,--------+-----------------------|-----------------+--, | - | | | BAT54 | | | | | ,-------, +-->|--+--2R--+---+ | | | +---|1 14 \ 1N4148 | | | | | | gnd | | \ |--' | | | | | 1M | 3 )o--+-->|--+--|| 100uF | | --- | | | / | | |--, | | | - | +---|2 7 / | < |IRF530| | | --- | | '-------' '----| | | 100uF | - | / | \ | | | | | +--100k-+--| PN2222 | CD4093 2N3906 | | | | | | | | > | | | | | zener| | | | | | | | | | | 10uF 1M | | | | | | | | | | | | | | | | | | '-------+----+--------+----------------+------+------+------+---+--' | gnd

--------------------------------------- Posted through

formatting link

Reply to
Michael Robinson

formatting link

'--------+----+--------+-------------------+------+------+------+----+------'

This's simpler still:

formatting link

--Winston

Reply to
Winston

formatting link

Does the improved circuit oscillate stably under all conditions? It looks like the circuit might latch up at high light levels. Also, it looks like the frequency of oscillation will vary with changing light levels - it seems the original circuit was designed to keep the frequency high enough to prevent the inductor from saturating.

Reply to
Bitrex

formatting link

On second thought I don't think it would latch up, the coil resistance would have to be pretty high for that to happen.

Reply to
Bitrex

formatting link

It's a cool circuit. High enough input voltage fires it, the inductor current builds up, that drags down the input voltage, the fet shuts off, and it flyback-pumps the battery. If the L saturates, easy to avoid, no big deal... you just lose a little efficiency.

I wonder if the dual 100u caps and the 2r resistor are necessary.

The class assignment should have been to do what you did: clean up that ED hairball.

John

Reply to
John Larkin

formatting link

'--------+----+--------+-------------------+------+------+------+----+------'

Good point. It's hard to beat a blocking oscillator.

John

Reply to
John Larkin

formatting link

But then what's to be done with the other sections of the 4093? :( One could probably get a similar circuit to work with a Schmitt trigger using just 2 transistors.

Reply to
Bitrex

Here is one I did for a solar cell which I had on hand to charge a 6V (or 12V) SLA. What is not shown is the microcontroller that measures the battery voltage and discontinues charging if the float voltage exceeds a threshold. It has been working since spring of 2007 as a stand-alone supply for my weather station. The extra components surrounding the transformer are measured strays, etc, for simulation. The transformer was a dual-winding inductor whose part number I do not have before me at the moment.

5 components not including the solar cell and battery.

Version 4 SHEET 1 1092 852 WIRE 128 144 -160 144 WIRE 384 144 192 144 WIRE -688 224 -784 224 WIRE -592 224 -688 224 WIRE -368 224 -592 224 WIRE -160 224 -160 144 WIRE -160 224 -368 224 WIRE -112 224 -160 224 WIRE 112 224 -32 224 WIRE 256 224 208 224 WIRE 384 224 384 144 WIRE 384 224 336 224 WIRE 560 224 384 224 WIRE 112 240 112 224 WIRE 208 240 208 224 WIRE -784 256 -784 224 WIRE -592 256 -592 224 WIRE -160 304 -160 224 WIRE 384 304 384 224 WIRE -688 320 -688 224 WIRE -784 352 -784 336 WIRE 112 368 112 320 WIRE 208 368 208 320 WIRE -688 400 -688 384 WIRE -592 432 -592 320 WIRE -160 480 -160 368 WIRE 112 480 112 448 WIRE 112 480 -160 480 WIRE 208 480 208 448 WIRE 384 480 384 368 WIRE 384 480 208 480 WIRE -592 528 -592 512 WIRE -368 560 -368 224 WIRE -304 560 -368 560 WIRE 208 560 208 480 WIRE 208 560 -224 560 WIRE 112 592 112 480 WIRE 112 592 -32 592 WIRE 112 640 112 592 WIRE -32 656 -272 656 WIRE -272 688 -272 656 WIRE -32 720 -32 656 WIRE 560 720 560 224 WIRE 560 720 160 720 WIRE -272 768 -272 752 WIRE 112 768 112 736 WIRE -32 816 -32 800 FLAG 112 768 0 FLAG -784 352 0 FLAG -592 528 0 FLAG -688 400 0 FLAG -32 816 0 FLAG -272 768 0 SYMBOL ind2 96 224 R0 SYMATTR InstName Lp SYMATTR Value 100µ SYMATTR Type ind SYMBOL ind2 192 224 R0 SYMATTR InstName Ls SYMATTR Value 100µ SYMATTR Type ind SYMBOL ind -128 240 R270 WINDOW 0 32 56 VTop 0 WINDOW 3 5 56 VBottom 0 SYMATTR InstName Llp SYMATTR Value 1.5µ SYMBOL ind 240 240 R270 WINDOW 0 32 56 VTop 0 WINDOW 3 5 56 VBottom 0 SYMATTR InstName Lls SYMATTR Value 1.5µ SYMBOL res 96 352 R0 SYMATTR InstName Rp SYMATTR Value .343 SYMBOL res 192 352 R0 SYMATTR InstName Rs SYMATTR Value .343 SYMBOL cap 368 304 R0 SYMATTR InstName Cs SYMATTR Value 4p SYMBOL cap -176 304 R0 SYMATTR InstName Cp SYMATTR Value 4p SYMBOL res -208 544 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R1 SYMATTR Value 47 SYMBOL current -784 336 R180 WINDOW 0 24 88 Left 0 WINDOW 3 24 0 Left 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName I1 SYMATTR Value .4 SYMBOL cap -704 320 R0 SYMATTR InstName C2 SYMATTR Value .1µ SYMBOL schottky -48 592 R0 WINDOW 3 -94 16 Left 0 WINDOW 0 -50 -14 Left 0 SYMATTR Value 1N5817 SYMATTR InstName D5 SYMATTR Description Diode SYMATTR Type diode SYMBOL voltage -32 704 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 WINDOW 0 -45 5 Left 0 WINDOW 3 -51 96 Left 0 SYMATTR InstName V1 SYMATTR Value 6 SYMBOL schottky -608 256 R0 SYMATTR InstName D2 SYMATTR Value BAT54 SYMATTR Description Diode SYMATTR Type diode SYMBOL voltage -592 416 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value .72 SYMBOL nmos 160 640 M0 WINDOW 0 5 -10 Left 0 WINDOW 3 -126 31 Left 0 SYMATTR InstName M1 SYMATTR Value IRF7343N SYMBOL cap 192 128 R90 WINDOW 0 10 70 VBottom 0 WINDOW 3 -18 -7 VTop 0 SYMATTR InstName Ct SYMATTR Value 14p SYMBOL cap -288 688 R0 SYMATTR InstName C1 SYMATTR Value .1µ TEXT 104 192 Left 0 !K1 Lp Ls 1 TEXT -552 728 Left 0 !.tran 0 .002 0 1u TEXT -744 128 Left 0 ;Solar cell TEXT -152 760 Left 0 ;Battery TEXT -64 80 Left 0 ;Transformer TEXT 688 256 Left 0 !;.step temp -25 75 100 TEXT 704 344 Left 0 !;.temp 75 TEXT 176 640 Left 0 ;Actually, Fairchild FDN335N TEXT 160 672 Left 0 ;----------------------- TEXT 96 528 Left 0 ;1 TEXT -224 208 Left 0 ;3 TEXT 216 528 Left 0 ;4 TEXT 472 208 Left 0 ;6 RECTANGLE Normal -464 592 -880 144 RECTANGLE Normal 32 848 -160 688 RECTANGLE Normal 464 512 -208 96

Reply to
John KD5YI

Then the error is deliberate, and you should not rely on others to determine what it is.

Reply to
DarkMatter

ge

hen

I

imply

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|

=A0 =A0 =A0 =A0 =A0 =A0 =A0)

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0) 160uH

=A0 =A0 =A0 =A0 =A0 )

=A0 =A0 =A0 =A0 =A0 =A0 =A0| BAT54 =A0 =A0 =A0 =A0 =A0 =A0|

=A0 =A0 =A0 =A0 +-->|--+--2R--+----+

=A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 =A0|--' =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

| =A0 =A0 =A0 100uF =A0 =A0|

=A0 =A0| =A0 |--, =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 < =A0 =A0 =A0 |IRF530| =A0 =A0 =A0|

=A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0100uF

=A0 =A0 =A0\ =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0zener

=A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

+-----=AD-'

ed

tion

the

ab

e
y

of

ge

ting

and

and

ould

There is a subtlety to the ED circuit you and the others are missing- apparently differential equations are beyond most EEs in this day and time- there is much more to this circuit than merely getting it to oscillate. Your experiment with placing a resistor in series with the power supply simulating the PV making it work is a clue ....To make a short story shorter- your circuit is a low performance joke ,and the original circuit is in a different league, the league of circuits that are useful and productive.

Reply to
Fred Bloggs

Here is one I did for a solar cell which I had on hand to charge a 6V (or 12V) SLA. What is not shown is the microcontroller that measures the battery voltage and discontinues charging if the float voltage exceeds a threshold. It has been working since spring of 2007 as a stand-alone supply for my weather station. The extra components surrounding the transformer are measured strays, etc, for simulation. The transformer was a dual-winding inductor whose part number I do not have before me at the moment.

Breadboard looks like:

formatting link

Reply to
John KD5YI

(...)

Very cool John!

Nifty circuit!

--Winston

Reply to
Winston

article:

formatting link

'--------+----+--------+-------------------+------+------+------+----+-----­-'

I mentioned above that the frequency of oscillation of the "improved" circuit is going to change with the amount of light falling on the cell. Lots of light will cause the IRF530 to stay on for shorter and shorter periods of time, and since the boost converter is running open loop it means the output voltage at the battery terminal will fall with increasing light to the cell. Perhaps you have a different issue in mind.

Reply to
Bitrex

article:

formatting link

'--------+----+--------+-------------------+------+------+------+----+-----­-'

I mentioned above that the frequency of oscillation of the "improved" circuit is going to change with the amount of light falling on the cell. Lots of light will cause the IRF530 to stay on for shorter and shorter periods of time, and since the boost converter is running open loop it means the output voltage at the battery terminal will fall with increasing light to the cell. Perhaps you have a different issue in mind.

Reply to
Bitrex

Hey, thanks. I didn't spend much time on it to make it efficient because I figured solar power is free. No need to worry about it.

Cheers, John

Reply to
John KD5YI

Here is one I did for a solar cell which I had on hand to charge a 6V (or 12V) SLA. What is not shown is the microcontroller that measures the battery voltage and discontinues charging if the float voltage exceeds a threshold. It has been working since spring of 2007 as a stand-alone supply for my weather station. The extra components surrounding the transformer are measured strays, etc, for simulation. The transformer was a dual-winding inductor whose part number I do not have before me at the moment.

Breadboard looks like:

formatting link

And here it is under a little florescent lamp. Yes, I know the current is low. So is the influx:

formatting link

Reply to
John KD5YI

formatting link

That's a fair question indeed. What did the OP do with the other 4093's? I hope he remembered to tie the inputs either high or low and not leave them floating.

Reply to
David Eather

:

uit

harge

, then

se I

5

the

I simply

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0)

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0) 160uH

=A0 =A0 =A0 =A0 =A0 =A0 )

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| BAT54 =A0 =A0 =A0 =A0 =A0 =A0|

=A0 =A0 =A0 =A0 =A0 =A0 +-->|--+--2R--+----+

=A0 =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 =A0 =A0|--' =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

--+--|| =A0 =A0 =A0 100uF =A0 =A0|

=A0 =A0 =A0 =A0| =A0 |--, =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 =A0 |IRF530| =A0 =A0 =A0|

=A0 =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0100uF

=A0 =A0 =A0 =A0 =A0\ =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0zener

=A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

=A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0 =A0| =A0 =A0|

-+----+-----=AD=AD-'

igned

he

anation

lt the

nal

nt

e lab

the

only

ch of

rof

eorge

tacting

nt and

ail and

should

r

nd.- Hide quoted text -

Yes- my issue relates to 1/2 C V^2=3D 1/2 L I^2 and the fact that the PV goes into photocurrent mode with a much smaller current than that supplied by the capacitor. So PV trickle charges C2 up to VBE(ON) of Q2, the latch swtches , discharging C2 through the boost L, and the R4- C4 is timed to switch everything off (via the latch and overriding the Q2 input because it is High) just when I=3DIpk. Turning off sooner or later means less energy transfer to the battery per cycle. This is what the description of R4-C4 time constant being 1/4 x 1/sqrt(LC) is all about- it's not perfect but at least it tries. Allowing the voltage across C2 to determine switch off time by dropping below the VBE threshold of Q2 almost certainly does not fully discharge the available energy.

Reply to
Fred Bloggs

article:

formatting link

then

simply

 '--------+----+--------+-------------------+------+------+------+----+-----­­-'

and

Hide quoted text -

Any energy left in the cap can be used next time.

John

Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.