Driver Circuit Thread

[random person]

Was taking a break from working on my boost drive and thought i'd make a fun thread for circuit nerds like me and possibly a resource for other people who are interested. Post your driver circuits here:

Discrete Linear Voltage Regulator (Constant Voltage Out)
Here is how i'd make a linear regulator with discrete components. I'm sure there is more ways to do it but this seemed the easiest. Caps are not necessary but they are helpful.

Vout = ~1.2(R1/R2+1)

more to come.... please feel free to point your designs here as well as a resource for others.
Regarding my boost driver design, it is about 2 months from sales(provided my proto pcb build goes well) and will cost roughly 12 usd for assembled and 6dollars for kit+pcb. 3A out, 6+V out Max ratings. Abit more complicated then the guy above this.

 

[GBD]

Nothing too excited for me,
Here are two I recently did.... one was a boost converter for an LED, and the other a linear regulator with some simple modulation for a laser diode array. (diodes driven in series)

Note for the boost, the ICs like the op amp and AND need decoupling capacitors. (I drew the caps for the timers, but forgot for the logic XD )

Working on some boosts based on LT ICs (various), so I can really limit the part count with a boost IC to work with.

 

[jib77]

Here is an easy dual rail driver based on a LM358N Op-Amp and a NPN/PNP Transistor pair. Useful for things that requires a + and - voltage like OPHIR Thermopile heads.

 

[random person]

Note for the boost, the ICs like the op amp and AND need decoupling capacitors. (I drew the caps for the timers, but forgot for the logic XD )

Thanks for sharing, obviously puts my simple linear reg schematic to shame. However, I do find beauty in simplicity. As for your circuit, there doesn't appear to be a feedback to the PWM control. It appears your controlling the duty cycle and freq with pots. Whats regulating the output then? The zener on the output? Very interesting schematic!

Very cool Jib! I like how you go about getting the dual supply with a single power supply via the op amp. Op amp method is very handy I hear. No ugly ripple from DC to DC ICs. Haven't had a chance to use something like that yet.

...to add to the goodies, heres a simple discrete linear constant source.

 

[GBD]

The output is kinda half-assed regulated by that zener, and then by a feedback transistor that varies the voltage on the base of that darlington (the voltage devider that leads to the NPN tied to AND's output to ground)

That clock feeding the darlington Is more or less independant from the boost circuit, so that is a pretty clear downfall of this design, so the only real limiting I have to the LED itself was either control the duty, or use the LED's V/I curve itself for current limiting.

However, what you can do is add another AND (to make 3 input AND) and add a comparator with a voltage devider tapped to the output of the boost converter, so once the referance voltage from the devider exceeds "X" value, it will turn the clock off momentarly and will restart upon refereance drop (which in turn is voltage drop across the output side). (same idea as the undervoltage shut-off I included as I was running from a SLA, but a different layout for feedback) that op-amp is really just bieng used as a comparator and can be replaced as such by the LM311 series and the likes. That is more or less what my little zener/npn on the output clock does (not the zener on the output to ground), but a comparator would probably be far better in responce and regulation. (ive yet to test it out with this circuit in regards to the comparator method)

Which I guess is the beutty of the ICs Linear tech and texas instruments provide, everyting all prepared for you in one package, just add a few passives on the outside rather then deal with a mess of parts and your gold

 

[random person]

Which I guess is the beutty of the ICs Linear tech and texas instruments provide, everyting all prepared for you in one package, just add a few passives on the outside rather then deal with a mess of parts

Quite right! LT has amazing ICs, documentation, and simulation software. Outside of work though I don't think i'd ever use them. Kinda hard to make a high current boost driver under 3 dollars with a LT IC. :cryyy:

oh... and heres my 18V to 2KV voltage driver... works great for umm... LPC826s and umm ... other... functions =P

 

[jib77]

Very cool Jib! I like how you go about getting the dual supply with a single power supply via the op amp. Op amp method is very handy I hear. No ugly ripple from DC to DC ICs. Haven't had a chance to use something like that yet.

The only drawback of course is now you have a different ground ref for anything being powered by it, like my 16bit ADC. I will now have to level shift the I2C lines to talk to my Arduino (which is external and powered by USB). :/

 

[GBD]

You and your multipliers random. didn't anyone ever give you a CRT flyback transformer?

 

[rhd]

Here is an easy dual rail driver based on a LM358N Op-Amp and a NPN/PNP Transistor pair. Useful for things that requires a + and - voltage like OPHIR Thermopile heads.

http://laserpointerforums.com/members/jib77/albums/other-stuff/5604-dualrail.jpg

This is technically a question for Jib, but since I know he's not around here often, I'd love anyone's input.

I'm trying to utilize Jib's schematic, but I'm running into trouble with the LM358N, and specifically, with the way Eagle handles this part. I'm fine with the way Eagle deploys two symbols for the same package. However, for some reason, the "A" symbol is only revealing 3 pins for me, whereas it revealed 5 pins on Jib's design.

Any thoughts? Screenshot:

 

[123splat]

The two symbols are because the package had two circuits in it. In such cases, usually the 1st circuit in the package is the one with the power & gnd leads pinned.
What *.lbr are you getting the package from? My version of 'linear.lbr' displays the LM358 member just like jib's. (I have tried to attach it as linear.txt, if that works, save it to your eagle\libs\ and rename it to linear.lbr, then look up the LM358 member). Hope this helps.
Good luck

 

[rhd]

The two symbols are because the package had two circuits in it. In such cases, usually the 1st circuit in the package is the one with the power & gnd leads pinned.
What *.lbr are you getting the package from? My version of 'linear.lbr' displays the LM358 member just like jib's. (I have tried to attach it as linear.txt, if that works, save it to your eagle\libs\ and rename it to linear.lbr, then look up the LM358 member). Hope this helps.
Good luck

I think this is probably not a Library/Eagle problem, but rather a user problem (a ME problem). IE, I'm probably just not understanding some usage issue with Eagle.

See, in the preview window, the part (from the linear.lbr) looks correct:

But when I actually place both symbols on my schematic, I don't get those two extra pins on the "A" symbol. As seen in the example I posted previously (mine is on the right):

 

[BShanahan14rulz]

Is there a setting where assumed things like powering ICs is hidden on the schematic? The software I used, very elementary, didn't have anything like that, but who knows, could be an option in the workspace preferences or something like that.

By the way, all you contributors rock. I was going to post a schematic to my LED car dome light array of 5mm LEDs, but I don't think it really belongs with the likes of these. It's just a simple 12V reg, 555-based PWM generator, and a series-parallel array of LEDs, more on the complexity level of the "DDL" lm317 cc circuit people use here.

 

[lazeerer]

I think this is probably not a Library/Eagle problem, but rather a user problem (a ME problem). IE, I'm probably just not understanding some usage issue with Eagle.

See, in the preview window, the part (from the linear.lbr) looks correct:
PICUTRE
But when I actually place both symbols on my schematic, I don't get those two extra pins on the "A" symbol. As seen in the example I posted previously (mine is on the right):
PICUTRE

Ive actually got one of these from Rick for my Ophir Head. It works Great.

Ill talk to him for you and Let you know what he says to do in Email. Iam going to talk to him tonight anyway.

Mine Shows Up at 5 IIRC. Iam at work so i will have to check when i get home.

Well thats if you dont figure it out before then.

 

[rhd]

Ive actually got one of these from Rick for my Ophir Head. It works Great.

Ill talk to him for you and Let you know what he says to do in Email. Iam going to talk to him tonight anyway.

Mine Shows Up at 5 IIRC. Iam at work so i will have to check when i get home.

Well thats if you dont figure it out before then.

Hey,

That would be great - thanks!

(I haven't figured it out)

 

[rhd]

Any luck?

I've tinkered some more, and can't figure our how to get three pins.

As lame as this sounds, could someone just place that one IC on a schematic, and send it to me? Maybe my library just has a bad copy.

 

[benmwv]

Hey rhd, you have to right click the lm358, click invoke, then click the power option. For whatever reason they set it to be available by request.