filter for USB line.

[settra]

hey. i was wandering, would it be possible to "Filter" a USB cable so no AC, passes through it, using a low pass impedence filtter between +5 and GND??

the reason i wanna do this, is becouse, i will be using a USB graphics card, and my pc produces allot of noise in the form of AC (you can see it on the 12v rails, that also have some small AC).

also, would it be possible, to cut the +5 and GND from the cable, and directly power it from enother source???
thanks in advancne!!

 

[MikeMl]

No reason you cannot power a device intended to be powered from USB from your own 5V supply. I have a USB backup disk-drive that worked unreliably when powered through the USB, but works just fine while powered off a modular linear 5V 2A power supply.

Any scheme to attempt to filter the USB 5V further will likely degrade the regulation somewhat. Any series inductor will have some DC resistance. You could just try adding an additional bypass (filter) capacitor at the distal end of USB. Find a cap with the lowest possible ESR.

 

[Nigel Goodwin]

There's no need whatsoever to attempt to 'filter' USB - and the ripple you're 'measuring' is most probably stray pickup from the scope, don't worry about it.

The USB graphics card is presumably intended to be directly connected and powered from a USB socket?, why do you not want to do that?.

 

[settra]

There's no need whatsoever to attempt to 'filter' USB - and the ripple you're 'measuring' is most probably stray pickup from the scope, don't worry about it.

believe me its not... i have a usb sound card that gives signal to the amplifier. even with nothing playing, you can clearly hear the noise on the speakers (and it is not a ground loop, because i have ground loop isolators in the signal lines, so, since it still passes , it must be AC)

at the graphics card i now have (normal PCI-e) , there is the same noise... allot of it. i can see it on the LCD. that is why i want to have a USB graphics card, because i think i might be more successful at filtering it's line.

Any scheme to attempt to filter the USB 5V further will likely degrade the regulation somewhat. Any series inductor will have some DC resistance. You could just try adding an additional bypass (filter) capacitor at the distal end of USB. Find a cap with the lowest possible ESR. .

since i will be using a regulator for 12v to 5v, i could use an inductor BEFORE the regulator... and also ofc bypass capacitors. the capacitors is a sure think they should have the lowest possible capacitance?? (and what about the inductor?? )

finally, i suppose there cant be a filtering in the D+ and D- lines, right??

 

[Nigel Goodwin]

I would suggest you probably have a SERIOUS problem with your computer, I suggest you try and sort that out.

It could be high ESR capacitors in it's PSU, causing excessive ripple and probably high voltages as well.

Or simply try another PC PSU, they are cheap enough - it would be crazy to try and solve internal problems externally, and your PC is likely to be damaged or destroyed as it gets worse.

 

[settra]

well, its a car pc, so nothing is for sure... the PSU is a dc-dc converter, and its supposed to be good. so i dont know if enother PSU would really solve the problem... also the particular problem , might not be from that at all. it could be a ground loop or something. thats why i sticked to just the "usb" part , in my OP. if i power the usb from a diferent source (Enother battary, not conected to the car at all), then it could solve some ground loop issues....

 

[KeepItSimpleStupid]

In a car anything is possible.

You do have to first figure out if the interference is there when the engine is running and/or the AC/heating system is on.

Isolating the USB is possible. There you can power from another source or in internal isolated source.
Here is one: http://keterex.com/kxusb150.php

I have one from an overseas company where you had to supply the other 5V source.

Here **broken link removed** is one that allows both bus owered and external power.

 

[Nigel Goodwin]

well, its a car pc, so nothing is for sure... the PSU is a dc-dc converter, and its supposed to be good. so i dont know if enother PSU would really solve the problem... also the particular problem , might not be from that at all. it could be a ground loop or something. thats why i sticked to just the "usb" part , in my OP. if i power the usb from a diferent source (Enother battary, not conected to the car at all), then it could solve some ground loop issues....

Use a scope on the outputs of the PSU - make sure there's no appreciable ripple on the outputs.

Your USB problems aren't normal - USB is EXTREMELY reliable and trouble free, you need to be curing the problem, not trying to bodge things to try and cure it's effects.

 

[settra]

how much ampere would a usb vga card drow?? would AA battaries, with a voltage regulator , be enough??.. also, if i was to power with battaries, and the voltage droped below 5v, what would happpen to the Usb to vga card??

 

[settra]

solving LCD Hum with USB vga card with filter ??

hey. my position is this :
i am running a car pc. the PSU is very noisy. (it has AC running on its +12, 5 etc rails).

my LCD, is powered by 12 and 5v , from the PSU, and it has allot of Huming (noise). the noise rises , when i rev the engine, so i guess it's AC produced by the alternator.

if i power the LCD from a completly diferent power source (Wall outlet), the noise is still there.
so, i guess it cant be a ground loop, correct??(or can it still be a ground loop between the signal wires??)

the reason i am asking, was becouse i am planing on buying an external USB graphics card, and completly cutting the 5v supply from the PC , and give 5v from a different PSU. and i want to see if i sould use an isolated DC-DC converter, or just a normal voltage regulator , with allot of filters would work...

thanks allot in advance!!

 

[Nigel Goodwin]

hey. my position is this :
i am running a car pc. the PSU is very noisy. (it has AC running on its +12, 5 etc rails).

my LCD, is powered by 12 and 5v , from the PSU, and it has allot of Huming (noise). the noise rises , when i rev the engine, so i guess it's AC produced by the alternator.

if i power the LCD from a completly diferent power source (Wall outlet), the noise is still there.
so, i guess it cant be a ground loop, correct??(or can it still be a ground loop between the signal wires??)

the reason i am asking, was becouse i am planing on buying an external USB graphics card, and completly cutting the 5v supply from the PC , and give 5v from a different PSU. and i want to see if i sould use an isolated DC-DC converter, or just a normal voltage regulator , with allot of filters would work...

Please don't post multiple threads asking the same question - the answer is still the same, cure the cause, don't try and bodge the symptoms.

Assuming the PSU is home made?, post the schematic , including the filtering feeding it from the car supply.

 

[settra]

the last post is about ground loop. the first post was about USB. it is not the same post.
but anyway, the PSU is not home made. its this : https://www.mini-box.com/M4-ATX

 

[Nigel Goodwin]

It's exactly the same question, about exactly the same problem.

So what filtering have you got feeding the supply, and how is it connected (directly to the battery, or via some other feed).

 

[settra]

its directly to the battery... and no filtering before or after the PSU atm!

 

[Nigel Goodwin]

its directly to the battery... and no filtering before or after the PSU atm!

You really need filtering in the feed to the PSU, a high current choke in series with the +ve lead, and a capacitor to chassis.

If you've getting alternator interference, a capacitor on the alternator would be good as well.

 

[settra]

yes but even if i get read of the ac comming from the Alternator, the PSU will still produce noise!!

when you say choke, what exactly you mean?? a serries inductor??
also, what capacitance should the capacitor be??? the smaller the better??

i bealive the noise should be about 20hz- 20kz, since i can hear it with a speaker.

 

[settra]

well, now that i see the calculations, if my noise is around 1000 to 10000hz, i will need pretty big capacitors, not small, am i correct??

 

[Nigel Goodwin]

You need an inductor, and capacitors, designed for interference suppression in cars - the capacitors bolt to the chassis and have a single lead out of the end.

If the PSU itself is generating interference then it's faulty.

How does it perform if you don't start the engine? - when there's obviously no interference caused by the car.

 

[settra]

there is still noise (proly from the PSU), but way smaller

 

[Nigel Goodwin]

there is still noise (proly from the PSU), but way smaller

Have you checked the outputs on an oscilloscope?.