question about active low pass filter, (will it work, if the Vin of the Op-amp, also has noise?? )

[settra]

hello guys... my problem is as follows: my car alternator, produces 12v DC, but ALSO produces a low frequency (100 - 1000)hz AC... i have a line, that has up to 20A-30A on it, which i want to filter... completely block all AC..
So i was thinking to build an active low pass filter with a buffer op amp , as follows:

so, the normal line would go to the non- inverting input , and i suppose i would get my filter 12v from the outpout??
my question is this:: If the Vin of the Op-amp (not shown above) comes from the same sources, as the signal in the non-inverting input, so it also has the alternator noise, will the above work?? or should i do something else??

(P.S i have rulled out the possibility of Ground loop or EMS, so take as a facrt, that the AC is produced by the Alternator)

 

[MrAl]

Hi,

You could always add a small amount of passive filtering to the op amp power source.

 

[audioguru]

You do not have an active lowpass filter. Instead you have a simple passive lowpass filter that has a low output current. Since the RC filter is so simple then if it filters out 50Hz it will take a long time to charge. The formula in your circuit reduces the frequency only a small amount (-3dB).

Most opamps have a maximum output voltage that is 1V to 3V less than their minimum power supply voltage except a "rail-to-rail" opamp that has a very low current load. Your 12V has 50Hz to 100Hz ripple that causes the voltage to go up and down. Even if you used a rail-to-rail opamp then its output will drop each time the power supply voltage drops.

What will the filtered 12V be used for?

 

[MikeMl]

...(P.S i have rulled out the possibility of Ground loop or EMS, so take as a facrt, that the AC is produced by the Alternator)

No, you have just stated the obvious, that the alternator produces ripple. However, the reason that the ripple effects your circuit is most likely because of a ground loop in the car's chassis. "Filtering" or "bypassing" will not do anything. Lets discuss how your circuit is intalled in the car.

I deal with this issue all the time in installing two-way radios in cars and aircraft. I have rewired many aircraft to get the alternator whine out of the aircraft's audio system...

 

[settra]

no, see?? that is why i ruled it out. becouse i DO NOT have a ground loop. if you disconnect everything, then the alternator still produces AC. just take it as an AC voltage source... and it is by far not ripple. ripple is high freq. mine is at the most, 1000hz...

audioguru.. the outpout will be used to power my carpc, (and i hope that enother one like that, my AMP) , so it has to be at least 20A....
i dont understand what you sed... since the op amp is wired as buffer, wont the output be, exactly the same voltage as the non inverting input?? (so with a low pass filter , in the input, you give a pure DC , and in the output you have a pure DC...)
that is my theory, but that is why i ask, becouse i dont have much Xp on Opamps...


[edit] just to be clear, on the schemantic of my first post, i suppose that the original Op-amp i have in my mind will be like this:

(although i dont know if there is something with 3 terminals...)

 

[MikeMl]

no, see?? that is why i ruled it out. becouse i DO NOT have a ground loop. if you disconnect everything, then the alternator still produces AC. just take it as an AC voltage source... and it is by far not ripple. ripple is high freq. mine is at the most, 1000hz...

99.99% probability is that you DO HAVE A GROUND LOOP!

All vehicle alternators put out full-wave rectified AC (ripple) at a frequency from 500Hz up to 5000Hz, depending on engine speed. The car battery "filters" the ripple component from the alternator, leaving only about 100mV of ripple riding on the 14Vdc power distribution in the car.

Most add-on electronics will tolerate that 100mV of ripple coming on its DC input without a problem, for example, my two meter VHF transceiver has no detectable alternator whine coming from the speaker, even though it is being powered from 14V dc with 100mV of ripple riding on it. Power to the radio comes from the fuse panel, the DC ground to the radio is near where it is mounted to the dashboard.

The speaker audio is totally whine-free until I ground the VHF antenna to the car body, say at the front fender. Now, the radio has two "grounds"; one is the black lead of its power cable, which is connected to the car frame on or near the dashboard. The other is at the base of the VHF antenna, through the shield of the RG58 antenna cable going from the rear of the radio to the car body.

The "ground-loop" happens because the Ohmic resistance of the coax-cable-shield and the #10AWG Black power lead have a lower resistance than the car body between the dashboard and the front fender of the car. All of the car's other electrical loads (lights, motor, fans, and even the battery charging current) use the car frame as the return path to the negative pole of the battery. Those currents (which can be much larger than the current going to the radio itself) cause an IR voltage drop between the two points where my radio is connected to the car's chassis. Now, currents that formerly would have flowed along the car chassis see a lower resistance through the wiring to my radio, and now flow through the radio pcb. That is why now I hear alternator whine in the radio speaker!

The solution is to not allow the second DC connection to vehicle chassis which is physically disjoint from the first. This is called single point grounding, and is the primary means of eliminating alternator whine from any electronics added to a car (or airplane).

Filtering will not eliminate the ground loop. If fact, adding a large bypass capacitor (especially the 1 FARAD capacitors used by the car audio enthusiasts) at the input of the radio will make the problem worse. Putting an inductive choke in series with +14V wire going to the radio may help, because it blocks some the ground loop current.

 

[settra]

look. i know that many people, hear "noise" and instantly turn to ground loop. as if it was the god of electronics. but this is not always the reason... i told you, that my Noise, is of very low frequency, and that , even with nothing at all conected at the alternator (except the ignition system ofc), the noise still pressists ( here, i will have to admit, that i have not checked the ignition system for ground loop , although, it has not been changed for years now, so i dont think its the source.)..

think of it this way: the alternator uses diodes, to rectify the AC to dc... so the diodes "block" the AC. but if a diode is dead or something like that, then the AC will go out, riding the DC line. no ground loops or anything....

that is why i want to build the filter... besides, my OP was about the filter, not the source of the noise (although , thoughts on that, are appriciated ofc)

 

[ericgibbs]

hi,
How much filtered power , Volts, Amps are you requiring.?
E

 

[settra]

the outpout of whatever filter i would use, has to be : 12v (+,- 2 ) and 20A (+,- 5 )

 

[ericgibbs]

the outpout of whatever filter i would use, has to be : 12v (+,- 2 ) and 20A (+,- 5 )

Thats going to be very difficult with a passive filter/regulator.
Have you considered a DC to DC converter.?

 

[settra]

well, i cant really find anything (not that i have searched much) . BUT the carpc, is supposed to be powered by a PSU that is dc-dc converter.... although it dosent filter anything ( maybe cause the noise is of very low frequency..)

edit: MikeMI. my system has two batteries, via a battery isolator. All my multimidia, (carpc, amplifier) are grounded on exact same spot (2nd battery negative pole ). and that battery then grounds to the chasis .. so the only things not grounded on the same spot with the carpc, dosent really have to do anything with it (car systems )

 

[ericgibbs]

Is a 2nd 12V battery an option.? connected to the main battery with a high power reverse blocking diode.?

 

[settra]

I do have a second battery as i sed before.... But how would this help??? Anyway, my question was about that specific filter, and how to make it work...

 

[audioguru]

Car audio amplifiers have the output power specified when the alternator is over-charging the battery at 14.4V. If you make a 20A voltage regulator then with a 13.8V input its output might be as low as 10.8V and the amplifier output power will be very low.

Car amplifiers have enough filtering so that an ordinary car power supply does not cause alternator whine when there is no ground loop.

 

[MikeMl]

look. i know that many people, hear "noise" and instantly turn to ground loop. as if it was the god of electronics...

That is because 95% of the time a ground loop is the cause of noise...

think of it this way: the alternator uses diodes, to rectify the AC to dc... so the diodes "block" the AC. but if a diode is dead or something like that, then the AC will go out, riding the DC line. no ground loops or anything....

So if you have a rectifier out in your alternator, fix your alternator. Don't band aid it by trying to put ineffective filters somewhere else.

that is why i want to build the filter... besides, my OP was about the filter, not the source of the noise (although , thoughts on that, are appriciated ofc)

I have been down this road hundreds of times. Filters almost never work!!!@!

 

[JimB]

Settra said:

that is why i want to build the filter... besides, my OP was about the filter, not the source of the noise (although , thoughts on that, are appriciated ofc)

the outpout of whatever filter i would use, has to be : 12v (+,- 2 ) and 20A (+,- 5 )

JimB says:
That filter does not stand a snowflake in hells chance of working.

You are filtering a signal (the supply to your equipment) as the input to an amplifier, and then using the output of the amplifier to power your equipment.
The power for the amplifier is coming from the same source as the power you are amplifying.
At this point my head hurts, can you not see the absurdity of what you are trying to do?

I do not want to get into an involved discussion on ground loops (I may lose!), but if you are adamant that you do not have a ground loop problem and that you want to use a filter, just use a simple L-C filter.
A nice big inductor with thick wire on an iron core, in series with the supply to the equipment and a capacitor from the equipment side of the inductor to ground and see what happens.

I will say it again, your proposed filter in your first post will not work.

JimB

 

[settra]

So if you have a rectifier out in your alternator, fix your alternator. Don't band aid it by trying to put ineffective filters somewhere else.

so, there is a change that its a bad alternator rectifier?? if so, will the alternator produce a noise as i describe (or would it still have to do with a ground loo)??
as for the ground loops, how can a ground loop be a problem, since everything, on my carpc/amplifier, ground on the exact same spot??

JimB. i expected the filter would not work, that is why i posted , to begin with..

 

[audioguru]

From your filter that will not filter, you want an output current of 20A. But the maximum output current from an opamp is 0.05A if you are lucky and then its output voltage loss is 4V.

 

[MikeMl]

so, there is a change that its a bad alternator rectifier?? if so, will the alternator produce a noise as i describe (or would it still have to do with a ground loo)??

With enging running, the ripple on your 14V main power bus when all six rectiftiers in the alternator are good and when the battery is in good shape will be <~100mV. With one rectifier out, the alternator will still appear to charge (but poorly), and the ripple voltage will increase to >~500mV. If you have a ground loop, and you hearing the good alternator faintly, when a rectifier fails, the alternator whine becomes much louder (>+10db).

as for the ground loops, how can a ground loop be a problem, since everything, on my carpc/amplifier, ground on the exact same spot??

To which I ask?
Where is the amplifier mounted? Where is it grounded?
Where is the head-end mounted? Where is it grounded?
Where is the AM/FM antenna mounted?
Are the speakers grounded?
Do you have any other audio sources (CD,IPOD)? How are they powered?
Anything else connected to the amplifier or head-end?

 

[settra]

so. the isolator is a diode type, (Will change it with relay soon)
amp is bolted to the backseats (no metal there)
Carpc, is bolted, on an aluminium case, wich is grounded, BUT removing that case, and having the pc un-grounded to the case, dosent change anything...
usb sound card -> RCA signal to the amplifier (huge noise, that changes with RPM, If no ground loop isolators on the RCA)
the LCD gets power from the pc. it has vga signal. no external ground (no metal parts of the LCD touch the chassi)
car pc and amp, are grounded to the 2nd battery, wich is then grounded to the chassis..

car pc, is powered either by : an inverter, and then a 220V PSU. or by a 12vDC psu (m4-atx) . no matter what, the noise is the same...

grounded speakers?? what do you mean??
no head unit, no AM/FM (although the antena is still there. it just dosent connect anywhere)...

some extra info: if the AMP gets it signal from my smartphone. it does not have any whine
the LCD has some history.... it a 15.6" home desktop screen.
when i used AC inverter, wich powered BOTH my carpc, AND my LCD. i had no noise.. Then i changed to a DC psu for the carpc, and the noise came for the first time on the LCD. thinking it was the invetrers problem, i changed the LCD's power supply, to get DC, and now it is powered from the PC psu. not only it did not solve the problem. but now, no matter if i have the DC psu, OR the DC->AC->DC method, it gets noise.