inverting op amp question

Thread Starter

MrWoo

Joined Jun 18, 2007
11
Hi.

I have a question about how to choose the correct op amp. I am just learning and tinkering, so bear with.

My current project has the need to drive a 12v fan at, 12volts. However, I have a signal that is PWM @ 25kHz at -.3v. I am thinking to drive the 12v fan with a transistor.

With this in mind, I have been thinking of an inverting op amp circuit. Simulation shows this to be working. However, now to build real world material, I am in question as to which op amp to use.

Currently I believe in simulation I am using an LM1458 with 2 resistors to keep the output @ 12v. Seems like a pretty simple circuit. I tried an LM324 but I don't believe the gain is enough as I was only able to achieve 10.6v @ output.

What would the considerations be for an op amp for this where the max input is 12v and the frequency is pretty low @ 25khz?

Thank you for any help,
MrWoo
 

bloguetronica

Joined Apr 27, 2007
1,541
... I tried an LM324 but I don't believe the gain is enough as I was only able to achieve 10.6v @ output.
The problem with the LM324 is not gain. The LM324 outpuc swings from 0V to Vcc - 1.5V. Since Vcc is 12V, you know the rest.
If you don't need your output to be analog, you can use a comparator such as LM339 or LM393 (they are essencially the same, one is quad and other is dual).
 

Thread Starter

MrWoo

Joined Jun 18, 2007
11
Hmm. While I don't implicitly understand all things electronic, well, I don't implicity understand all things electronic. But the more I mess with it the more I learn.

I have attached the very simple design. I am just beginning to understand the many varied uses of an op amp, and for that matter how to actually understand what the datasheets are trying so hard to tell me.

For now I am trying to understand some basics like the current application.

I know the input signal @ -.3v is PWMing the way I want, I want to try to drive a fan with it at the same rate. This may or may not be a proper thing to do but it is certainly good learning material.

I am under the impression that the correct op can boost the voltage to output 12v @ the same PWM freq, and that the correct transistor or fet or what have you, will be capable of switching at the same freq.

Sorry for the lack of know how, but listening intently none the less.

MrWoo
 

Attachments

Spoggles

Joined Dec 2, 2005
67
Hello Mr Woo

Here (below) is an excellent discussion of your issue. Especially with older technology such as the LM1458 (I was troubleshooting boards with them in the late 70's) it is not possible to use input voltages that are out of the power supply limits. (Rails) For the 1458 this limit should be 1.5v inside either limit. ie your input voltage must be above 1.5v from your low power supply 'rail' and not greater than 1.5v from your hi power supply rail.

Also, these devices 'like' to use 'split' power supplies. (+V gnd -V).

Your other issue is one of power...I do not think that you are going to be successful in running a fan using this device.

Hope this helps

http://www.analogzone.com/acqt0322.pdf


Spoggles
 

techroomt

Joined May 19, 2004
198
it appears the lm1458 can handle up to +/_15vdc supply, so that would work fine for achieving a +12v Vout. it also appears there's plenty of open loop or closed loop voltage gain, you'll be looking at Av = vout/vin, 12v/.3v= 40. what i couldn't put my hands on was the slew rate to see if the 0 - 12v will work at 25khz.

interesting project, mr woo. i am not sure a dc fan motor will like pwm - let me know.
 

Distort10n

Joined Dec 25, 2006
429
it appears the lm1458 can handle up to +/_15vdc supply, so that would work fine for achieving a +12v Vout. it also appears there's plenty of open loop or closed loop voltage gain, you'll be looking at Av = vout/vin, 12v/.3v= 40. what i couldn't put my hands on was the slew rate to see if the 0 - 12v will work at 25khz.

interesting project, mr woo. i am not sure a dc fan motor will like pwm - let me know.

It depends on what his signal gain is. Full Power Bandwidth (FPB) = SR/(2*pi*Av*Vp).
 

nanovate

Joined May 7, 2007
666
The LM1458 has a unity gain bandwidth of about 1MHz so you'd be better off going with a faster rail-rail output opamp or use a supply rail that will allow you the 12V output. Maybe the TLC082 which is also a dual opamp it has a SR of 16V/us and output current of 50mA. Do you really need to drive the output transistor w/12V? Could you use something like 5V?
 

nomurphy

Joined Aug 8, 2005
567
Provided you have a fan that will accept the 25KHz signal, then you must amplify the signal by a factor of approximately -10 (= ~+3V) and then drive a FET connected to +12V thru the fan.

Unfortunately, the circuit you attached will not go to gound, it will output +2V to +10V (or thereabouts). You need a minus supply for the output to go to ground.

If you use a +/-5v supply (instead of the +12V), and change the input resistor to ~3.3K (Av > -10), then the output will be sufficient to drive a FET gate.

Take the output of the opamp thru another 3.3K resistor to the gate (pin-1) of an NTD3055 FET. Connect the FET Source (pin-3) to GND -- assuming there maybe +/- fan connections -- connect the (-) Fan to the FET drain (pins-2/4), and the (+) Fan to +12V.

This will provide a +12V, 25 KHz, signal to the fan.

Note: The FET gate can easily tolerate any negative voltage from the op amp.
 

Thread Starter

MrWoo

Joined Jun 18, 2007
11
Erm, been sidetracked for a bit. Back on it now.

Thank you for the replies!

Give me a day or two to play with it.

Seems there is enough lingo in these replies that I must now do some more learning to see what they mean exactly.

As far as pwm'ing @ 25khz, my fans can indeed do that. My board right now is pwm'ing my fans, both 80 & 120's @ 25Khz. I can adjust my frequency on this mobo from 10Hz to 30kHz, and noticed the difference both in audible levels of the "kick" and in the led lights flashing, as well as just how low a duty cycle each frequency can keep the fans running at. So, I am curious to investigate further.

I put a scope on my board, and found that indeed tis a -.3 volt pwm signal. Now, as this thread indicates, to find a method to boost that up to 12v. Or for that matter, should anyone have a simple method, using any + voltage to drive any FET or whatever that might drive a 12v fan.

I am certainly not proud enough to hold to a certain design. As a matter of fact, possible within reason the more the better. To comprehend it that is.

Maybe to say that applying a common 4pin molex supply of either 12 or 5v should be kept, as this is an easily accessible source. I am not sure about using negative supplies.

Tonight I will try to get my uber bread board tuned in to provide -.3v & a pwm signal close to 25khz, if that is possible. Then I will try some of the OP's that I have on hand and see what I can produce.

Again, thank you for your time to respond.
MrWoo
 

Thread Starter

MrWoo

Joined Jun 18, 2007
11
Here is something maybe. This one outputs to 11.3v with a single supply OP. I will obviously drive a FET with it.

As per the NoMurphy design, I am afraid I do not completely understand it all. At least, are you saying to use both + and - 5v? Would that be on the VCC + & - on the OP? Or +5 on the VCC+ and -5 on Input +?

And what of the OP? Are you referring to the 1458 in the schematic?

I am sorry that I don't get it all. While I don't particuarly see a problem with a - voltage, as my breadboard has that available, I won't find that in a computer case I don't believe without tapping maybe one of the ATX lines. Maybe I am wrong about that tho.

As for the NTD3055, why that one? I have been playing with IRF530's and some different TIP's, in different scenarios. Can you explain why that one was your choice?

Some further study on the 4 pin PWM fan headers reveals that there is a spec that says that the "control" pin (the magical 4th one) is to supply max low logic of .8v (Vil). And further that there is a pull-up resistor in the PWM fans that, for standard is to 3.3v, with a max of 5.5v. So it would appear that the output of the 4th pin is correct, and that normally a PWM fan would be converting to 3.3-5.5v and some IC in it then actuates the 12v pulses. At least it seems that way.

I rescoped that again with a fan on the 3pins, thinking maybe the "control" voltage changed with a load, but it is very steady @ -.3v.

I have attached a very simple TLC081 circuit. Any suggestions?

Again, thank you for your time.

MrWoo
 

Attachments

nomurphy

Joined Aug 8, 2005
567
I chose the NTD3055 because it has a maximum gate threshold voltage of 2V, which means it can be driven from an op amp using +/- 5V, it is also a relatively high current/wattage device that would MOST likely cover whatever it is you're trying to do.

The IRF530 has a gate threshold voltage of 2V-4V, and it is unlikely that you will be able to completely turn it off using the LM324, which is not going to go all the way to ground, but maybe +1.5V ...this makes the design unreliable. However, using a rail-to-rail opamp would improve the reliability of turning the FET off.

Try the attached with your breadboard, and see if it does what you want.
 

Attachments

Thread Starter

MrWoo

Joined Jun 18, 2007
11
Thank you NoMurphy. I ordered some of those up so won't be able to try it out till next week.

Some further research reveals the following.

Specs for the 4pin PWM fan header state that the max VIL on the "control" pin to be .8v. Like I have stated I have scoped my board to be at -.3volt. The specs also state that the signal should be pulled-up to 3-5.5v, with 3.3v being the standard. This is done within the circuitry of a PWM fan. Further study has enlightened me to the opto-coupler and it's inherent isolation capabilities. Other posts elsewhere suggest a pull-up resistor and an opto-coupler buffer.

Now, whilst I am surely at infant stages, I grasp stuff fairly quick. However, electric theory is a rather large subject to digest.

Here I attach a schematic I made that utilizes an LM324, a pull-up resistor, an opto-coupler, a FET, and a digital meter used to represent a fan.

While EWB does not provide every component known, I do have some that may work.

For starters, I am uncertain as to whether or not this circuit follows any "laws" or not. It seems to work with the parts displayed, but simulation as I have found out is most definately not real life true. So, here are some questions on it.

1. For a pull-up resistor, would it be in the correct placement, or for that matter is it needed since I am using the OP?

2. The OP itself, I have now in a list the LM324, the 1458, the TL081/2, the TLC081/2. I think my concerns should be perhaps the PWM frequency so the OP can operate fast enough?

3. The opto, which seems straight forward enough, I am sure there are "do's and don'ts", but they are proving to be more elusive in finding good examples or tutorials about.

So, the real world parts I am thinking to use would be, as you state, the NTD3055 FET, perhaps a TL or TLC OP, and I have been looking at the 6N137 opto. I understand that the voltages etc that the circuit attached are too low to actually work, but I will assume that with the real parts I were to use, listed prior, that it would work. Ahem, well, being naive is good for optimism anyway :)

Again, thank you so much for the feedback.

MrWoo
 

Attachments

Top