LM350T adjustment calculation

[manie]

I have looked at the datasheet but afraid I do not understand it............!!!!!! Too old or something ............
I need to calculate the R2 resistance for a 3V output............ datasheet shows R1=120ohm, datasheet attached.View attachment LM350.pdf

Anyone that can help ????????????

Manie

 

[manie]

Which other 3V 1A regulator can I use with a pass transistor to increase amps to 10A ????

Any help greatly appreciated........

Manie

 

[nick12ab]

The equation in the datasheet is:
Vo=1.25(1+R2/R1)+IADJ
The datasheet says that IADJ<0.1mA. The datasheet actually isn't very helpful.
Using trial-and-error, I calculate 170 ohms as the value you need for 3V output. Test it with a multimeter.

 

[srnet]

Data sheets says that the regulator will maintain 1.25V between the output and the adjustment terminal, i.e. across the 120R.

So (ignoring the very small adjustment terminal current) for 3V output, the voltage drop across R2 needs to be 1.75V

You could work out the in the 120R current and then the current for R2 at 1.75V, but its the same for each resistor so the answer is simply;

R2 = (1.75/1.25) * 120 = 168R

 

[MFB]

Although not exactly for the same device, the following LM317 on-line calculator may be of interest. http://www.whatcircuits.com/lm317-calculator-v2/

 

[manie]

See that what I am talking about !! Nick says 170 ohm and SR says 168 ohm so more do you want ??? Thanks guys !!!!
Manie

 

[hippy]

The equation in the datasheet is:
Vo=1.25(1+R2/R1)+IADJ

Actually it's Vo = 1.25(1+R2/R1) + IADJ R2

As IADJ is less than 0.1mA, rearrange the equation to determine R2 and then solve for IADJ=0 and IADJ=0.1mA and that will give the range of resistor value it should be. Then you can decide what actual resistor value to use.

 

[manie]

Thanks Hippy, got that.
manie

 

[BeanieBots]

Hi manie,
Now that you know how to calculate the resistor, it might be worth keeping your calculator out to work out how much POWER your pass transistor will dissipate and then work out how HOT it will get.
10A will give 10W of heating for every volt dropped across your little tranny! We don't want any more smoke;-)
So, what's the input voltage? what's the output voltage and what is the degrees/watt of your heatsink?

 

[bluejets]

LM196/LM396 or similar plus appropriate heatsink.

 

[John West]

I'm kinda partial to 5K 10 turn trim pots for setting regulator output voltages, but that's just me.

 

[booski]

I'm kinda partial to 5K 10 turn trim pots for setting regulator output voltages, but that's just me.

Ditto! Really though, is there much need to use an adjustable reg when your going to use a pass transistor?

A little 100ma to92 case 3v reg would be fine, the output coupled to the base of a few 3055 transistors with a 1w 1ohm resistor.

I did a similar think in my amp build. lm7818 with 2v zener on the ground brought it upto ~20v. The output went via 1ohm 3watt resistor to a pair of bd911 transistors. Little heatsink and pwm, temperature controlled fan, lovely!

food for thought.

 

[manie]

Hi BB.....
I am dropping from 24 to 15V=9V drop on one (for 10A 3V out=30W finally), = +-3.5A= +-32W (in case of 24V input).
For 12V input I am dropping to 6V=6v drop = +-5A = 30W
For the 3V I am dropping 3V at 10A = 30W.
I will use BD712's as pass trannies doubled and trippled up (wattage = 75W each)
I think doubling the tranny will give 150W possible with +- 15W per tranny actual. Should be OK yes ???

Manie
Edit:
Oh yes, the heatsink is the metal box which is 130mm x 200mm, 1mm thick.

 

[SAborn]

Depending on your application you could take a different approach and do a better design using the LM723 chip, i done a article on building a variable power supply that also allows current limiting, its a simple circuit that works well and could be set to a fixed output.
have a look here for the circuit (LM723)
http://www.thebackshed.com/windmill/articles/VariablePowerSupply.asp

 

[Dippy]

Is 'Magic Smoke Manie' melting semiconductors again?

I think I must have wrong end of stick.
In summary; you are dropping 24V to 15V -> 3V @ 10 Amps?
BB has given a good clue.
Remember, with linear your Amps go 'right through'.
What a lot of watts.

Can you do a schematic as I can't quite grasp this.
I worry that even your metal box won't sink the watts.

Can you go switched-mode manie?
Converting power is better than converting volts.

 

[MartinM57]

I think I must have wrong end of stick.

Chopstick? You been and returned yet?

 

[bluejets]

Don't know why you don't simply use a device like lm196 designed for the job.
Seems if you start to parallel transistors as what I think you are suggesting, you could be headed for a big smoke up if not done correctly. Thermal runaway.
Bolt the regulator to a d9 dozer blade for heat transfer and you're done.

 

[srnet]

Seems to be a complicated way to build a patio heater .................

 

[manie]

Hi Bluejets
"Don't know why you don't simply use a device like lm196 designed for the job."

If I can find it - ANYWHERE - I would probably use it..... every site I have tied comes up empty including National/TI......

Do you know where I can find it for trial and then probably 100's later ??

Manie

 

[bluejets]

Well, I did find this ........ http://www.ic2ic.com/search.jsp?sSearchWord=LM196

But like you said, scarce as hens teeth, probably due to the fact that switching regulators are much more common for higher currents as someone commented before.

I did find this for you though........ might be some help.....(down towards the bottom of the page)
http://www.reuk.co.uk/High-Current-Voltage-Regulation.htm

Looks like he used fixed regulator but should work the same with adjustable.
Doesn't mention any problem due to paralleling.

 

[Dippy]

Why are we (you) getting hung up on ancient/obsolete/hard-to-find devices?

Granted that without a schematic or block diagram it's difficult to make proper suggestions but consider the following.

1. Go to Farnell/RS Components/Digikey and do a PARAMETRIC search.
At least it'll give you some options on current (pun) devices.

2. Many regulators can work with pass tranistors.
You may have seen an on-line circuit but do more searching on designs. Don't get stuck on one design.
But please be aware that it will get chuffing hot and I'm not sure your box will sink adequately.

3. Look at switcher designs unless there is some specific reason to avoid them.
The extra bit of noise may or may not affect your downstream circuit. Check.
Your component specification and circuit layout will make a big difference to performance.

4. Consider Switcher control ICs plus a suitable external transistor.
Similar caveats as above.

5. Think about a Switcher doing the bulk work and a linear hanging on the end to fine-tune.
Some noise will still sneak through. Don't imagine a linear will get rid of the noise totally.

6. A DIY Switcher using PWM , maybe with a linear at the end.
Similar caveats as above.


There are many options. Please don't get obsessed with a single design because that was the first one you found on some site.
Switchers were 'made' for this type of job unless noise is a problem.

 

[manie]

Hi Dippy
No noise is not a problem, current and current limiting though is NB. The devices I've found here in S.A. are mostly SMD and I don't know whether I'll have the board built yet or not (certainly not the first trial board). Time constraints are a problem as well, Ive got +- 2 weeks left............... I am looking at switchers yes (see SMD note) but not much luck so far...
Manie

 

[Dicky Mint]

Hi

I'm building a Bench Power Supply from a LM723 and a laptop power supply as in the attachment to post 14 by SAborn

I want to limit the current to 3Amps adjustable and I can't seem to find the correct information.

From my calculations the two 0R47 power sense resistors in parallel give a max of about 3Amps but I thought it supposed to be a 10 Amp supply?

Sorry if I'm missing anything apparent but I'm a bit lost.

 

[SAborn]

@Dicky,

Did you find the original thread on TheBackShed forum i posted the article in, as i think there was more information there (been a while)
If you get stuck i will see what i can sort out for you when a get some spare time.
It will output 10 amp but not for extended time as the resistors get very hot, although i never found i required long periods of time at 10 amp on the bench.

What sort of amp range were you hoping for, as i think mine only goes down to 1 amp, but it also goes above 10 amp too.
I also find the good old analog needle amp meter is better than a digital meter as you do notice the needle taking a big hit if there is a short or something.
With current regulation i dont find i need much less than a 1 amp setting as it just limits to that without distroying itself, i often end up with the leads shorted on the bench and no problems, in fact to test the current limiting i short the leads and dial the amps up or down.

 

[Dicky Mint]

Hi SAborn

Thanks for getting back to me.

I've got a convenient 20V 3.25A power supply and I think I'll use that to power the LM723

I guess the set up in 'thebackshed' is calibrated for 10Amps.

I don't really want to draw more than 3 Amps max as I don't want to over load the laptop supply and
I really don't deed much over an amp for my projects so a plump 3amps will give me loads of head-
room.

I'm using analogue meters, Aluminium sheet for heat-sinks and toying with the idea of housing it all
in an ABS case! (is that mad?)

I'd like to control the current below an amp ideally down to a few tens of milliamps but definitely a few
hundreds of milliamps if at all possible.

I'm off out for the evening and most of the days till Friday so will converse soon, probably Friday!

 

[manie]

Hi SAborn

I notice there are NO base current Resistors. Is this correct ? Also, which pot controls what ?
The 5k controls V or A's and the 500R pot controls what ? For 10A output extended periods of time what size heatsink is required ??
Also, R8=1K, for what range is this set ?

Thanks for your time
Manie

 

[SAborn]

Manie,

After looking at this thread from time to time i gather you wanted a supply for extended 10 amp use and not a bench supply with limited high amp use, so would not think this circuit would suit you now.

To answer your questions....... correct no base resistors needed.......Pots? a quick look at the data sheet would explain that, 5K for voltage and 500R for current, the pot values shown are just a indication and may need changing to better give a setting range targeted, example is i use a 500R pot in the place of R9 (360R) as a fine trim to voltage ajustments.

R8 on my schematic here shows 120R (maybe i changed it in the posted schematic) this voltage divider controls the output voltage on the inverting input (pin 4) against the set voltage on the non inverting input (pin 5) you might think of this function as a opamp.

As for the heat sink, i dont know what to rate that at, other than a bloody big one! for extended 10 amp operation, its not just the transistors running hot at 10 amp, but the output resistors too, as these will be smoking hot after a while at 10 amp, and is why i say its not recommended for extended high current use.

Mine has been in service for over 20 years now as a bench supply, and was the first bit of workshop equipment i built back then, but it dont sit in use at 10 amp very often, most use is 1 - 5 amp for general circuit work and 10 amp normally means magic smoke out and close to fire, hence why the current is normall dialed down to 1 amp unless its needing to be higher.

 

[Dicky Mint]

Hi SAborn

I've been looking at the data sheet for the LM732 and can't seem to find how to limit the maximum
output current to 2A, which is the new bench-mark I am aiming for.

I'm assuming its fairly easy to do by adjusting a resistor value but which one?

I'm also assuming that the leads that are connected to the 500R pot and the 330R resistor from the power resistors only
need to carry very low currents and hence only need to be low current wire?

Is this so or have I made an dangerous and potential explosive misconception?

I've got my hands on a wooden case which I intend to fit aluminium front and back panels to, one as a heat sink for the
2N3055's the other to mount the analogue meters and am wondering whether to fit a small fan with ventilation holes or
just to rely on the 2mm aluminium panel as a heat-sink as I'm running the 2N3055's at relatively low current?

 

[SAborn]

Dicky,

The 500R pot is what sets the current limit, and yes it is low current through the wires to it, i think i had used ribbon cable for this, for 2-5 amp i would only use 2 x 3055 transistors, but there is nothing wrong with using 4 transistors should you want.

The voltage applied to pin 3 is what sets the max current limit. (500R pot works as a voltage divider to pin 3)

For higher outputs there is no problems replacing the 5 watt resistors with 10 watt resistors, but for 2-5 amp range 5 watt resistors will be fine. (often 10 watt is easier to find than 5 watt)

I dont think you will need a fan if you have a reasonable size heat sink, you are only looking for low currents here, if pushing 10 amp then thats different, add the fan later if you fell its needed but i think you are safe without it.

 

[AllyCat]

I'd like to control the current below an amp ideally down to a few tens of milliamps but definitely a few hundreds of milliamps if at all possible.

Hi Dicky,

You're definitely not going to achieve that with the present configuration. The problem is that the current sense pins expect to see a differential of about 0.6 volts when the current limits. If that is to occur at 100mA then the current sense resistor needs to be 6 ohms, but then at 3 Amps it would drop 18 volts and dissipate over 50 watts! At present the potentiometer "hides" much of the current which is flowing in the current sense resistor(s); instead you want it to "pretend" that that there is more current than there really is.

If you're using a laptop power supply then I guess the current is going to limit at about 3 Amps anyway, so you needn't consider higher currents. If I were adapting the design to do what I believe you want, then I would try modifications as follows:

Delete two of the 2N3055s and ALL the 0.1 ohm resistors. Separate the two 0.47 ohm resitors at their "top" ends and connect one to each emitter (these are now going to be the current sense and current-sharing resitors). Remove all the connections to the pot wiper and its top end, and to IC pins 2 and 3. Connect the low end of the pot (Output+) to Pin 3 and the high end of the pot to Pin10 (BD139 base) - this is a higher voltage which will "fool" the current sense pin that there is more current flowing than really is.

Finally connect 3 resistors from the current sense (Pin 2), typically a 470 ohm to each of the 2N3055 emitters and 220 ohms to the pot wiper. Now, when the pot is at the top of its track the current sense pin will think that lots of current is flowing when there is little (if any), whilst at the lower end of the track the current is monitored normally.

Sorry only a theoretical design, but I have designed similar before. Even this looks quite "generous", you could probably delete yet another of the 2N3055s and/or move the lower end of the pot to the emitter of one of the 2N3055s (which will reduce the maximum current further).

Cheers, Alan.

 

[Dicky Mint]

Hi

I've decided to stick with the circuit diagram in post 14 and just run it at a max of 3 Amps.

I think I can limit the max current at 3A by fitting a 10k pot in place of R8 and adjusting,
but I must admit I don't understand how the circuit works so I may be barking up the
wrong tree completely!

I've designed a PCB and am going to mount the pots off board and connect them via
headers. The same type of header for the current sense and junction.

I'm going to mount all the power resistors close to the power transistors in old amp style.

I need a common collector cable and wondering the best way to attach that to the board?

I was thinking of just passing the stripped end through the board and soldering it to the pad
the same with the 0V and the 20V input but is there a better way?

I would like to run the whole set-up at 3 A max but retain the ability to, relatively easily, convert
the whole power supply to 10Amps max so I will use the original four power transistors.

I've included a pre-posed PCB design and if I could have some feedback on that, that would
be good?

Thank you for your time in advance I know it is a valuable commodity!

 

[Dippy]

What does "pre-posed" mean?

Maybe this is one for SABorn to check?

 

[Dicky Mint]

pre-posed means proposed

Inglish language was never my strong point

Rick

 

[AllyCat]

Hi,

It's unfortunate that the data sheet diagrams all show the pin numbers for the alternative package, but as far as I can see R8 does nothing to affect the current limiting. I think increasing it will just increase the output VOLTAGE, probably to the point where there's no regulation at all (i.e. Vout is almost equal to Vin).

The current limiter simply operates when the voltage between pins 2 and 3 reaches about 0.7 volts (but this varies considerably with the chip temperature). So, with the wiper of R11 at the LOWER end of its track, current limiting will occur with 0.7 volts across the 2 x 0.47 ohm resistors, or about 3 Amps (I = V / R = 0.7 / 0.235).

As you move the wiper up the track there needs to be MORE voltage across the 0.47 ohm resistors (before the IC sees 0.7 volts) so the current limit RISES. About 2/3 of the way up the track the current would limit at about 10 Amps, and further on it might be entering black smoke territory. So IMHO the originally designed configuration isn't particularly "safe".

I can't see anything obviously wrong with your PCB layout, but I'm not sure what you mean by putting the pots on headers. If that involves any form of them being "plugged in", then if they fall out you might get rather unpredictable or even destructive results.

Cheers, Alan.

 

[Dicky Mint]

Hi

I'm a bit confused at present by the detailed operation of the circuit.

I do take your point about the pin headers for the pots could be an
accident waiting to happen!

I'll use PCB pins or just solder the bared wires through the PCB to
the pads.

Unless there is a better way I've missed?

Rick

 

[AllyCat]

Hi Rick,

To be absolutely sure, perhaps I should point out that there appear to be two points on the circuit diagram that look like they might be "crossovers" but are actually 4 wires joining at a point (many years ago I was trained NEVER to do this but to "stagger" the meeting point of one of the wires). These points are to the left of R2 and the left of R8.

I can't think how to explain the current limiter any better. However, the Voltage Regulator control circuit (the output from pin 12) basically uses a comparator with inputs on pins 4 and 5. The control circuit attempts to make the voltages at these two inputs equal.

Pin 6 outputs a reference voltage which appears to be nominally 7.15 volts (from the data sheet). The lower end of R9 is connected to zero volts so the wiper of R10 can select a voltage of between about 0.5 volts and 7.15 volts. R7 and R8 divide down the Output voltage to about 3/4 (as drawn in the diagram) so I would expect the Output voltage to regulate over the range of about 0.65 to 10 volts. This presumably explains the comment in the margin that R8 can be changed from 10k down to 120 ohms (actually I'd expect it to need to be even lower for a 40 volts regulated output).

For a maximum output of about 25 volts (to allow for some tolerances) I would NOT reduce R8 to as low as 120 ohms, but leave R8 at 1k and increase R7 to about 2k2 or 2k7. I wouldn't use another pot as it could be confusing having two pots which both change the Output voltage.

I'm rather surprised how relatively poor the original design appears to be, maybe it was intended for a completely different purpose. I'm afraid you may have to try it out at low powers first. I hope you have a multimeter?

Cheers, Alan.