The SkeletonDAC - Page 13

Quote:

Originally Posted by tomb 

Quote:

Originally Posted by Frodo 

Hi, I just wondered, is the Skeleton DAC safe to use with the Millet Starving Student Hybrid? I read that the SSMH had destroyed a few Bantam / Alien DACs.

 

http://www.head-fi.org/t/445851/starving-student-blows-up-dacs

Use with caution.  All of those other DACs had output capacitors, too.  The common symptom was a DAC that had the chip directly output (through capacitors) to the SSMH.  IOW, there is no buffer, opamp, or other active device between the DAC chip's output pins and connection to the SSMH.  Since the SkeletonDAC falls into that category, caution is advised.  It's possible that since the PCM2704/5 has a built-in headphone amplifier, that it offers more protection than the PCM2702, which formed the basis for the Alien and Bantam.  We don't know that for sure, though.

 

You should also be aware that the issue was really traced to the CISCO switching power supply used in the original Starving Student.  The CISCO power supply seems to have a floating ground and a different potential voltage can develop around the RCA jacks.  If you happen to plug in one of the DAC's mentioned and touch the outside of the jacks with the center pin of the RCA interconnects, a discharge will occur that will fry the DAC chip - output capacitors or not.

 

If you don't use that power supply, then you shouldn't have a problem.

Wouldn't that floating ground be a possible issue for nearly any source?

Quote:

Originally Posted by Frodo 

I just wondered if anyone is using bypass caps, and is there any particular value in it?

 

It's going to be a while before I get new output caps and wondered if using the Wima caps that came with the Skeleton DAC kit  in parallel with the Nicon 220uf caps I am using at the moment would be a worthwhile mod. I've read that bypass caps can improve high frequency response.
 

Actually, I tried a bypass combination once based on Dsavitsk's recommendations in his excellent "Notes on Output Coupling Capacitors."  It was back a few years ago and the DAC was the AlienDAC, but the same principles apply.  If memory serves, I used a couple of Muse ES's at 4.7uf, and bypassed them with a couple of 0.22uf Sonicap Gen II's.  It sounded very good ... until I blew it up with a Starving Student amp.  Sorry - I guess that's too close to home in this situation.

IOW, yes - I think the right combination of bypassing might yield excellent results if you have the room.  However, what I used was definitely one of the specific combinations that Dsavitsk mentioned in his articles.  The ES's are known for excellent bass, while the Sonicap Gen II's were known for sparkle.  Unfortunately, I think Soniccraft quit making the Gen II's.  Anyway, it's worth experimenting and Wima's are usually a great bypass caps in most situations  Roederstein box film caps are also a great bypassing cap.  Just remember that combinations may vary widely.  Bypassing is almost witchcraft and voodoo in terms of predicting the results of various combinations.

Quote:

Originally Posted by scootsit 

Wouldn't that floating ground be a possible issue for nearly any source?

There seems to be some benefit of protection when a DAC has an output buffer or opamp.  IMHO.  I don't keep up with some of the latest FIO/UDAC/Nuforce stuff, but a few years ago, DIY was about the only place where you had output signals taken directly off of the DAC chip.  That's what the original AlienDAC did and the BantamDAC after it.  Some famous mods to CD and DVD players did the same thing.  Someone would identify where the DAC chip was on the DVD player and solder some boutique film caps between the DAC chip output pins and the RCA jacks of the DVD player.  Then they'd claim the modded sound was more "pure" and better.  (Maybe it was?)  The SkeletonDAC does the same thing, it's just that the new chips actually have a rudimentary headphone amplifier encapsulated in the PCM DAC chip (PCM2704, PCM2705).  So, that may provide some additional protection because we never noticed frying DACs when the DAC had some additional circuitry beyond the DAC's outputs.

Frying the DACs in the scenarios that were mentioned is a very illustrative comparison betwen "theoretical infallibility" and the real world.  There's probably not a person on Head-Fi who would admit that a capacitor would pass DC current, but nevertheless, it does - under the right conditions.  After all, a lethal transient could simply be "interpreted" as the first part of a low frequency sine wave.  I don't know - maybe that explanation is invalid, but frying the DAC through a DC discharge at the outputs is a fact and was repeatable (for many people unfortunately).  I actually heard the static discharge through my headphones on a couple of occasions when a DAC was fried.

Quote:

Originally Posted by tomb 
There's probably not a person on Head-Fi who would admit that a capacitor would pass DC current

That's a funny way to say it. As you know, a coupling capacitor passes "dc current" while its charging to the potential difference of the 2 points being coupled.

Quote:

Originally Posted by DingoSmuggler 

That's a funny way to say it. As you know, a coupling capacitor passes "dc current" while its charging to the potential difference of the 2 points being coupled.

Yes of course, but I'm talking about well-charged caps - I think.   One might say the film caps could've discharged in the length of time it took to connect/disconnect to a Starving Student, but this happened with electrolytics, too.  You would hope they'd hold a charge a little longer.  I suppose as another possible explanation, the zapping could've happened while the caps were in a discharge state.  It was so intermittent/unpredictable, however, it was hard to test - expensive, too.

Quote:

Originally Posted by tomb 

[...] There's probably not a person on Head-Fi who would admit that a capacitor would pass DC current, but nevertheless, it does - under the right conditions.  After all, a lethal transient could simply be "interpreted" as the first part of a low frequency sine wave.  I don't know - maybe that explanation is invalid, but frying the DAC through a DC discharge at the outputs is a fact and was repeatable (for many people unfortunately).  I actually heard the static discharge through my headphones on a couple of occasions when a DAC was fried.

DC voltage is simply dV/dt = 0. Any transient would, by definition, not have dV/dt = 0. So a transient is not DC.

The decoupling capacitors block DC, but the moment dV/dt != 0, it will allow current trough. We use this to filter audio signal out of DC bias, since audio signal is everything but dV/dt = 0. The thing is, capacitors work both ways. If there's a change in voltage at the output of the circuit, the capacitor will allow this difference in voltage to flow back trough the circuit. Since the other side of the cap is the DAC chip, this transient discharges trough the chip. If it's not built to handle such current, it fries.

In this case, the static discharge when connecting the DAC to the amplifier causes the voltage at the output of the capacitor to rise (dV/dt > 0). This causes the capacitor to conduct in the opposite direction it's intended to, which applies a voltage to the output of the DAC chip. The reason it fries the chip is probably as simple as it not being tough enough to handle voltages being applied to it's output.

The damage occurs on the rising edge of the transient, as this is when the capacitor conducts.

Quote:

Originally Posted by tomb 

Actually, I tried a bypass combination once based on Dsavitsk's recommendations in his excellent "Notes on Output Coupling Capacitors."  It was back a few years ago and the DAC was the AlienDAC, but the same principles apply.  If memory serves, I used a couple of Muse ES's at 4.7uf, and bypassed them with a couple of 0.22uf Sonicap Gen II's.  It sounded very good ... until I blew it up with a Starving Student amp.  Sorry - I guess that's too close to home in this situation.

 

IOW, yes - I think the right combination of bypassing might yield excellent results if you have the room.  However, what I used was definitely one of the specific combinations that Dsavitsk mentioned in his articles.  The ES's are known for excellent bass, while the Sonicap Gen II's were known for sparkle.  Unfortunately, I think Soniccraft quit making the Gen II's.  Anyway, it's worth experimenting and Wima's are usually a great bypass caps in most situations  Roederstein box film caps are also a great bypassing cap.  Just remember that combinations may vary widely.  Bypassing is almost witchcraft and voodoo in terms of predicting the results of various combinations.

Thanks for the info, as you recommended I'm now using the Wilma's as bypasses and I'm pleased to report that it does seem to be a nice improvement , highs are much brighter and clearer (it also makes me happy that the Wilmas' are being put to good use). I have to make a trip into Tokyo after Christmas so will look out some of the caps you've mentioned. The shipping from Mouser is very expensive here.

Really sorry to hear about your Alien, I hope it was repairable. I've got my Skeleton plugged into my Starving student and it sounds excellent with thankfully no problems yet, I'm really really hoping it stays that way. I'm probably going to leave it more or less permanently connected.

I've also built an Alien, I'm going to pair that up with an Objective 2 eventually, so the link to Dasvitsk's article is very useful.

Quote:

Originally Posted by KimLaroux 

 

DC voltage is simply dV/dt = 0. Any transient would, by definition, not have dV/dt = 0. So a transient is not DC.

 

The decoupling capacitors block DC, but the moment dV/dt != 0, it will allow current trough. We use this to filter audio signal out of DC bias, since audio signal is everything but dV/dt = 0. The thing is, capacitors work both ways. If there's a change in voltage at the output of the circuit, the capacitor will allow this difference in voltage to flow back trough the circuit. Since the other side of the cap is the DAC chip, this transient discharges trough the chip. If it's not built to handle such current, it fries.

 

In this case, the static discharge when connecting the DAC to the amplifier causes the voltage at the output of the capacitor to rise (dV/dt > 0). This causes the capacitor to conduct in the opposite direction it's intended to, which applies a voltage to the output of the DAC chip. The reason it fries the chip is probably as simple as it not being tough enough to handle voltages being applied to it's output.

 

The damage occurs on the rising edge of the transient, as this is when the capacitor conducts.

Well, yes - I think this is an accurate explanation of what happens.  I didn't want to digress into diffierential calculus in my post, but you've got it.  It doesn't happen on more sophisticated DACs because there's enough circuitry on the output to dissipate/absorb that transient before it hits the DAC chip's output pins.

Regardless, some people tend to think that DC is DC, period, and just re-gurgitate the principle that capacitors block DC and that's supposed to cover it.  I just meant to convey the idea that not everyone's opinion of what constitutes DC is absolute.

It's probably the difference between a calculus-based Physics curriculum and one that's not calculus-based.

Hi Guys,

Is the PCB available in the UK?  I've looked at Beezar and it looks like it's $10 shipping to the UK, a bit expensive for me.

Due to that I thought I'd have a go at trying to proto-type it on a protoboard with through-hole components - but now I've read the whole thread I can see why I've not got it to to be detected and just spent two frustrating days trying it to no avail.  Wish I'd noticed that earlier!

I'm assuming it's due to the through hole components being the problem as stated.  I've put the PCM2704 on a SSOP-DIP converter board - is it easy to fry the chip in mounting/reflowing? I was a bit worried about the heat from keep dragging and reflowing to get it right.  I'm pretty new to DIYing.

Anyway, looks like I'll have to start properly and try to get the board somehow, if anyone knows a better route than paying $10 for postage please let me know!  

Quote:

Originally Posted by vaugi 

Hi Guys,

 

Is the PCB available in the UK?  I've looked at Beezar and it looks like it's $10 shipping to the UK, a bit expensive for me.

 

Due to that I thought I'd have a go at trying to proto-type it on a protoboard with through-hole components - but now I've read the whole thread I can see why I've not got it to to be detected and just spent two frustrating days trying it to no avail.  Wish I'd noticed that earlier!

 

I'm assuming it's due to the through hole components being the problem as stated.  I've put the PCM2704 on a SSOP-DIP converter board - is it easy to fry the chip in mounting/reflowing? I was a bit worried about the heat from keep dragging and reflowing to get it right.  I'm pretty new to DIYing.

 

Anyway, looks like I'll have to start properly and try to get the board somehow, if anyone knows a better route than paying $10 for postage please let me know!  

That's been corrected.  True shipping to the UK is $7.30.  That's with 75 cents of Beezar handling.  I've been having some trouble with the shipping module at Beezar because of the new USPS pricing.  First Class International really took a hit - things almost doubled.  However, there were another couple of dollars that were not supposed to be in that calculation because the weights were off.

Not counting my packaging, printing and gas costs to the Post Office (the 75 cents), the basic rate is still $6.55.  You are not going to find cheaper than that ... anywhere.

I refunded $2 to another customer who ordered some light weight stuff a couple of days ago.

P.S. This goes without saying, but even at that cheapest shipping rate, the full SkeletonDAC kit is your best value.  The shipping is still $7.30.

Many thanks for the reply TomB, and it's fine I fully understand the pricing, in fact I should have bought when I first looked a few months ago when it was about $5-6 delivered. Postal prices in the UK have shot up too over the past year.

I'll have another play with the protoboard and if I can't find a problem I'll bite the bullet and purchase.