[vikingboy]

Hey guys, I need to check something because Im not sure Im understanding right here and I dont want to shell out $$$ before getting my facts right.

Some basics first to check my understanding of the theory and electronics math behind this subject.

Replacing the caps in the IPOD with the diyMOD enables us to improve the quality of the audio ultimately going to our headphones. The ones on the curcuit board are small and "poor" quality thus degrading the sound quality more than larger ones.

Mounting large caps outside of the case is possible before feeding the signal into our headphone amps. The Frankenstein IPOD thread over at http://www.head-fi.org/forums/f15/fr...d-here-254531/ shows some crazy large caps outside the case. What puzzles me is why use such large uF caps?

If the headphone impedance is 10k ohms, I find that the -3db point (essentially the lowest frequency that will pass te cap) is as follows per uF of cap.

EDIT: Using the formula =1/(2*PI()*AmpImp*(0.00001*DacOutCap))

AmpImp(Ω) - DAC Output Capacitor (μF) - Lowest Frequency (Hz)
100000.115.91549431
100000.27.957747155
100000.35.30516477
100000.43.978873577
100000.53.183098862
100000.62.652582385
100000.72.273642044
100000.81.989436789
100000.91.768388257
1000011.591549431
100001.11.446863119
100001.21.326291192
100001.31.224268793
100001.41.136821022
100001.51.061032954
100001.60.994718394
100001.70.936205548
100001.80.884194128
100001.90.837657595
1000020.795774715
100002.10.757880681
100002.20.72343156
100002.30.691978013
100002.40.663145596
100002.50.636619772
100002.60.612134397
100002.70.589462752
100002.80.568410511
100002.90.548810149
1000030.530516477

Using a 6.x uF cap seems like total overkill.

Assuming the frequency response of hearing is circa 20hz and the Shure earphones go down to 18Hz (taken fromthe shure site)

Technical Specifications:

* Speaker Type: Triple TruAcoustic MicroSpeakers
* Sensitivity (1mW): 119 dB SPL/mW
* Impedance (1kHz): 36
* Frequency Range: 18Hz – 19kHz
* Cable Length/with Extension: 18 in. / 45cm (54 in. / 136cm)

Why not use a smaller, and cheaper cap, something like a 0.1uF would facilitate a -3db at 15hz.

What Im curious about is do I totally misunderstand this and if not, what is the curve from 0db to the 03db point like, how steep is it, essentially meaning how large a cap do I have to use to ensure at my headphones lowest freq range Im within 0db range still.

I hope those are sensible questions, if not....redicule away my friends!

thanks in adv
Ian - aka 'n00b'

 

[joneeboi]

I'll try to keep the insults to a minimum.

Although a smaller capacitance will suffice frequency-wise, the higher value capacitors typically exhibit higher quality (see Tangent's article Input Capacitors for Headphone Amps). Verily, those smaller caps are bound to work and you're welcome to use them at your discretion. In EFN's case, it's more an issue of he wanted those caps than he needed them to achieve a certain FR. At that point, you're really splitting hairs, but some find that hairs need to be split to fully enjoy their audio. I've flipped between 47uF and 22uF and found the former to perform better in the bass region than the 22uF, but that's using my amp and my headphones; YMMV.

To see the frequency rolloff graph and for more information, check out Wolfson's Paper on The Role of the Headphone Coupling Capacitor. I think it'll answer a lot of your questions.

 

[vikingboy]

Hi Joneeboi,

Thank you very much for them links, they answer my questions perfectly. Its all so clear now.......

Makes me wonder why there arent more hghi uF, low V caps available?

I notice the lager value caps when rated at 300V+ are large enouh to be not exactly portable.

Does anyone make caps capable of 200+uF at approx 2v? The BG 47uF seems to be the closest-to-perfect little cap for our use.

Thanks for keeping insults to a minimum BTW!

Ian

 

[FallenAngel]

Quote:

Originally Posted by vikingboy /img/forum/go_quote.gif Hey guys, I need to check something because Im not sure Im understanding right here and I dont want to shell out $$$ before getting my facts right. Some basics first to check my understanding of the theory and electronics math behind this subject. Replacing the caps in the IPOD with the diyMOD enables us to improve the quality of the audio ultimately going to our headphones. The ones on the curcuit board are small and "poor" quality thus degrading the sound quality more than larger ones. Mounting large caps outside of the case is possible before feeding the signal into our headphone amps. The Frankenstein IPOD thread over at http://www.head-fi.org/forums/f15/fr...d-here-254531/ shows some crazy large caps outside the case. What puzzles me is why use such large uF caps? If the headphone impedance is 10k ohms, I find that the -3db point (essentially the lowest frequency that will pass te cap) is as follows per uF of cap. EDIT: Using the formula =1/(2*PI()*AmpImp*(0.00001*DacOutCap)) AmpImp(Ω) - DAC Output Capacitor (μF) - Lowest Frequency (Hz) 100000.115.91549431 100000.27.957747155 100000.35.30516477 100000.43.978873577 100000.53.183098862 100000.62.652582385 100000.72.273642044 100000.81.989436789 100000.91.768388257 1000011.591549431 100001.11.446863119 100001.21.326291192 100001.31.224268793 100001.41.136821022 100001.51.061032954 100001.60.994718394 100001.70.936205548 100001.80.884194128 100001.90.837657595 1000020.795774715 100002.10.757880681 100002.20.72343156 100002.30.691978013 100002.40.663145596 100002.50.636619772 100002.60.612134397 100002.70.589462752 100002.80.568410511 100002.90.548810149 1000030.530516477 Using a 6.x uF cap seems like total overkill. Assuming the frequency response of hearing is circa 20hz and the Shure earphones go down to 18Hz (taken fromthe shure site) Technical Specifications: * Speaker Type: Triple TruAcoustic MicroSpeakers * Sensitivity (1mW): 119 dB SPL/mW * Impedance (1kHz): 36 * Frequency Range: 18Hz – 19kHz * Cable Length/with Extension: 18 in. / 45cm (54 in. / 136cm) Why not use a smaller, and cheaper cap, something like a 0.1uF would facilitate a -3db at 15hz. What Im curious about is do I totally misunderstand this and if not, what is the curve from 0db to the 03db point like, how steep is it, essentially meaning how large a cap do I have to use to ensure at my headphones lowest freq range Im within 0db range still. I hope those are sensible questions, if not....redicule away my friends! thanks in adv Ian - aka 'n00b'

Actually, those numbers are a little off by 10x, 0.1uF cap with 100K pot will have a -3db of 15.9Hz, with a 10K pot as in portable amps, you're looking at 159Hz!

 

[vikingboy]

I used the excel spreadhseet linked from this thread somewhere to calculate those values, is it wrong then!?

=1/(2*PI()*AmpImp*(0.00001*DacOutCap))

EDIT: refering to this link Input Capacitors for Headphone Amps it would appear that Excel sheet I used is wrong! ***..... :grumpy:

The link is on the first page of this thread....

9.4.0 Black Gate Capacitors:

Look for NX Hi-Q 6.3V 22uF or 47uF. The more capacitance, the smaller the value of the corner frequency and the bigger the packaging. 47uF is consensus.
f = 1/(2*pi*C*R), where C is capacitance, and R is the input impedance of the amplifier
Use jEriCOh's simple Excel corner frequency calculator for convenience <<<<<<<<<<<--------------------

 

[FallenAngel]

Quote:

Originally Posted by vikingboy /img/forum/go_quote.gif I used the excel spreadhseet linked from this thread somewhere to calculate those values, is it wrong then!? =1/(2*PI()*AmpImp*(0.00001*DacOutCap)) EDIT: refering to this link Input Capacitors for Headphone Amps it would appear that Excel sheet I used is wrong! ***..... :grumpy: The link is on the first page of this thread.... 9.4.0 Black Gate Capacitors: Look for NX Hi-Q 6.3V 22uF or 47uF. The more capacitance, the smaller the value of the corner frequency and the bigger the packaging. 47uF is consensus. f = 1/(2*pi*C*R), where C is capacitance, and R is the input impedance of the amplifier Use jEriCOh's simple Excel corner frequency calculator for convenience <<<<<<<<<<<--------------------

Yep, it does appear to be wrong, easily fixed though, just add another 0 to the calculation

Basic things to remember:
10K Pot : 4.7uF
50K Pot : 1uF
100K Pot : 0.47uF

Those are "best case scenario" values, anything larger will have questionable yield.

 

[gueri_fr]

Hi!
gueri_fr is back!
I opened a broken ipod 5G to see inside before doing the mod on MY ipod. Ooops! I broken the lcd screen. I will need to pay more attention next time Waouh! caps and inductors are very very small. Soldering & desoldering will not be easy!
I think I will use an external jack connector instead of routing the wires from the DAC to the dock connector. So I just need to route the DAC signal out of the ipod (with a hole on the case off course). One question: Can I solder the wires without removing the two caps near the DAC ?

gueri_fr

 

[joneeboi]

Certainement. N'oubliez pas de bien attacher vos cables.

 

[gueri_fr]

Merci pour tes conseils joneeboi

Now, I have to buy blackgate caps and wires. To be continued...

gueri_fr

 

[joneeboi]

If you're going to go with completely external caps, why not go for some different caps? If size isn't a factor, I don't think you need to stick with BGs. With the iPod 4G, the NX Hi-Q is almost the automatic choice for an internal mod because it balances SQ and size. In your case, the sky (and/or your wallet) is the limit.

In other news, I've discovered yet another decent supplier of replacement iPod parts, iRepair.ca or iRepair.us. I can only really recommend their cheap front panels and back plates as the price on everything else is pretty jacked up. As always, my main recommendation for a wide selection of cheap and varying-quality parts is eBay, but for instance, the 3G back panel I was looking for some time ago only had one listing on eBay (which I did eventually buy) and was out of stock at iFixit: iPod, iBook, & PowerBook Parts and Accessories. Suppose you feel that you don't like the drilled hole in the back of your iPod but you can't reverse it. Try iRepair if eBay and the other listings on the OP don't work out.

 

[DrizzitT]

So, after finally having the time to attempt to trouble shoot the my Mini...

I find that there is STILL a DC offset no matter what I do.

From the LOD its 30 mV and after my amp (yes, I know running DC through my amp is BAD), its ~ 15 mV

I'm pretty sure I'm doing this right (I get one channel sound, cleanish), but am just worried about the DC offset. Anyone have any idea what could be happening?

DC Line out= 1.5 V
Output Offset (after caps)= 30 mV

I know the Caps are supposed to kill DC offset.... but for some reason they aren't. Not sure why. Since I'm assuming the output from the line out is clean, something wrong in the Dock?

 

[joneeboi]

Pics are always helpful. Try rolling different caps into the DC blocking position to see if you just have some bad caps. What kind are you using? Did you melt the caps during the soldering process? My best guess would be just to get a different set of caps. If anything, 15mV is tolerable for headphones, so you're still pretty safe using your amp.

 

[DrizzitT]

Bought the dock from Steven Kelby so I can't check the caps.

...

Interesting thing is that the output from the amp is actually reading only 12 mV... Might just be discharging issue?

Edit: Think it was discharging (was playing music to test if channels were correct). Seems like one is near dead on and the other is ~ 2 mV off (which was offset at the cap SMD pads). Looks good. :-D Thanks a lot guys!

 

[joneeboi]

Sounds good (I hope

). Glad to hear it. Enjoy your diyMod!

 

[iQEM]

dang, my left channel are loosen up...must fix the solder again, after along2 time not using it via line out...