Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Supercaps and DC offset
New Posts  All Forums:Forum Nav:

Supercaps and DC offset

post #1 of 26
Thread Starter 

I used some supercaps (3F) as DC blocking caps to mod my ipod but after the mod I measured some dc offsets from the line out dock. Initially it was measured around 150 mV on both channels, but recently I measured it again the dc offsets seemed having grown to 250 mV on both channels. I am fairly sure the dc offsets are related to the caps as I have a pair of mods with the same caps, and both mods were measured with almost identical dc offsets.

 

I am more puzzled by the growth in dc offsets than the dc offset itself.

 

So my questions are ...

 

1) Are supercaps prone to produce dc offsets? (FYI, my other mods with film caps produced no DC offsets.)

2) Why do the dc offsets go up with time?

 

I have been using these mods with the supercaps for a while and have not noticed anything unusual (e.g., no distortions). The amps and iem's seem doing just fine.

 

Any explanations?

 

Thanks

post #2 of 26

I'm not sure I understand, where exactly did you place these huge caps? Did you remember to add a high-value resistors to ground after the cap?

post #3 of 26
Thread Starter 
Quote:
Originally Posted by KT88 View Post
 

I'm not sure I understand, where exactly did you place these huge caps? Did you remember to add a high-value resistors to ground after the cap?

 

The caps are not that huge, which are small enough to fit inside the fat back plate. No resistors were added to ground after the caps. The mods with film caps also have no resistors, but no dc offsets were measured.

post #4 of 26

A. 3F is huge. There's really no need for it, but if you like it better, sure :)

B. Two wrongs don't make a right - just because you got away with it last time doesn't mean there's no need for a resistor to ground after the caps. Have a look at some capacitive coupling at the output of amplifiers, there's always a resistors to ground. If you don't put it there the DC voltage at the output is undefined (floating). Putting a load on the output will pull it down, but that is bad practice since the voltage can be high when you first plug it in, and this will discharge over the load.

C. The reason you might have missed it with the smaller caps is because they were... well, smaller :) They discharged much faster, and the input impedance of your DMM probably helped. With 3F this time constant is so high it will take the DMM ages to discharge. Just add a resistor, something large enough so you don't mind about it compared to the load (100's of Kohm's), it'll take care of it.


Edited by KT88 - 9/15/13 at 11:45am
post #5 of 26
Thread Starter 
Quote:
Originally Posted by KT88 View Post
 

A. 3F is huge. There's really no need for it, but if you like it better, sure :)

B. Two wrongs don't make a right - just because you got away with it last time doesn't mean there's no need for a resistor to ground after the caps. Have a look at some capacitive coupling at the output of amplifiers, there's always a resistors to ground. If you don't put it there the DC voltage at the output is undefined (floating). Putting a load on the output will pull it down, but that is bad practice since the voltage can be high when you first plug it in, and this will discharge over the load.

C. The reason you might have missed it with the smaller caps is because they were... well, smaller :) They discharged much faster, and the input impedance of your DMM probably helped. With 3F this time constant is so high it will take the DMM ages to discharge. Just add a resistor, something large enough so you don't mind about it compared to the load (100's of Kohm's), it'll take care of it.

Thanks for the suggestions. I will try on my next experiment to put in the resistors (hopefully I can find the space to put them). Can you explain how exactly the resistor are implemented? Currently I wired from the DAC to the caps then to the lod pins (the output channels pins). Sorry forgive my ignorance as I have no electrical background in theory or training.

post #6 of 26

By "how its implemented" you are asking how to connect it?

 

One end of the resistor goes to the output of the cap (the same point that goes to the LDO), the other should go to ground which is the same pin that goes to the ground pin on the LDO. I can't say what pin number it is because I don't know the pinout of this LDO.

Obviously you'll need a resistor on each channel.

post #7 of 26
Thread Starter 

Yes I was asking about how to connect the resistors.

 

Now if the resistor is connected to the cap and ground pin, does that mean at the same time I also keep the wire from the the cap to the pin? In another words, the cap is connected to both the resistor and the LOD output channel pin from the same lead, correct?

post #8 of 26

You have DC offset because large caps have relatively high leakage currents. 

In their intended application (typically running the clock on something like a VCR or microwave oven so that you don't have to reset it every time the power goes out, alternately for supplying power to computer memory) the leakage current is insignificant. In this application, it is a very real problem. 

 

Figure out the smallest value you can use, and get the highest quality cap you can that physically fits the space at that value. 3F is way too large. 

post #9 of 26
Thread Starter 
Quote:
Originally Posted by nikongod View Post
 

You have DC offset because large caps have relatively high leakage currents. 

In their intended application (typically running the clock on something like a VCR or microwave oven so that you don't have to reset it every time the power goes out, alternately for supplying power to computer memory) the leakage current is insignificant. In this application, it is a very real problem. 

 

Figure out the smallest value you can use, and get the highest quality cap you can that physically fits the space at that value. 3F is way too large. 

 

Thanks for the explanations. Yes I also have much smaller value (4.7 uf) caps which do not have any dc offsets. The only reason I am using the 3f caps is for experiment. I am just wondering what's going to happen if I continue using the ipod with the presence of dc offsets. So far I have not noticed anything causing concerns. Of course, if there is a way to eliminate the dc offsets (as KT88 suggested using resistors) I would like to try that on these supercaps.

post #10 of 26

Using these resistors will drop the DC down. Obviously if there's a very large leakage current as nikongod as suggested you might have to use lower value resistors to pull it down sufficiently. I'm not sure what the leakage current of these caps are.

 

Again, you could use this for the experiment if you'd like. But as I have mentioned above, and as nilongod noted as well, these caps are way too large. It'll probably be better to use lower capacitance value with high quality caps instead.

post #11 of 26
Thread Starter 
Quote:
Originally Posted by DMinor View Post
 

Yes I was asking about how to connect the resistors.

 

Now if the resistor is connected to the cap and ground pin, does that mean at the same time I also keep the wire from the the cap to the pin? In another words, the cap is connected to both the resistor and the LOD output channel pin from the same lead, correct?

 

KT88, can you confirm if the resistors should be installed this way? Thanks

post #12 of 26

yes, just as you said. The resistor is in parallel with output (which is between the out pin and ground).

post #13 of 26
Thread Starter 

Any thoughts on why the dc offsets have grown?

 

Another experiment I am thinking about is to use a LOD with dc blocking caps. it may sound crazy with caps in both the LOD interconnect and in ipod. 

post #14 of 26

A. Could be that the leakage current is too high as nikongod has suggested. What resistance did you use?

B. Could be that it'll take lots of time for it to drop - that's 3F! give it a few minutes and measure again, did it change at all?

C. No, dont use capacitive coupling for the capacitive coupling. That's like calling the movers to move the other movers :) Just use a single capacitor of good quality and be done with it.

post #15 of 26
Thread Starter 
Quote:
Originally Posted by KT88 View Post
 

A. Could be that the leakage current is too high as nikongod has suggested. What resistance did you use?

B. Could be that it'll take lots of time for it to drop - that's 3F! give it a few minutes and measure again, did it change at all?

C. No, dont use capacitive coupling for the capacitive coupling. That's like calling the movers to move the other movers :) Just use a single capacitor of good quality and be done with it.

 

The ESR is 61 mOhm for the 3F cap. Yeah I will measure again the dc offsets after running it for a few minutes (from my memory I probably have done it).

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Supercaps and DC offset