DIY Line Out Cable - Works!

@14124all wrote:

The problem seems to be with input impedance on devices that people connect to. They seem to vary quite a bit. Car stereos and boomboxes are probably all over the place. Home receivers should be quite constant at around 10-20k ohm. axon gets good level when connected to a receiver, but many people get low level when connecting to car stereos using the Fuze headphone out.

I always thought it was an impedance issue as well, but someone posted here that I was incorrect (maybe neutron bob? can’t remember).  It’s defintely the case with both my home and car receivers that the Fuze connected via headphone gives a much lower level then a CD.

 

@14124all wrote:

The View cables I bought on ebay are the ones that roll up on a recoiler (got 2 for $7 shipped). I did not open the connector end because I could see no needed contacts in the 30 pin connector. The Sansa cable snaps apart. Other cables may be molded or heat staked and may break when seperated. You have to glue them back together.

It seems like the one I have has a full connector, but I guess the housing must be glued.  Maybe I can separate it with an exacto without doing too much damage.

I guess it’ll be worth sacrificing the cable for experimentation since it’s useless to me anyway.  Then assuming we get it figured out properly, I think I’ll order some of those connectors from that site I posted and some high quality audio cables from monoprice and try to make a really good cable.

I stand corrected Skinjob :smiley:

I crawled around through the “storage” room and dug out and  hooked up my CD changer again, and compared the same track… and there certainly is a difference. Granted, I’ve never used MP3 Gain or anything of the sort,but  I’ll still use the Fuze and Clip most of the time to preserve the CD’s though.

@14124all wrote:

axon, These are only assumptions.

 

I don’t think the lineout and headphone jack are the same signal. …

 

… I think what is killing us here is that the headphone out signal in the Fuze is overly good. It is very clean all the way up to full volume.

I’m not sure what you mean?   :) 

In message 47, sansafix said,

"At full scale,  the HP out will be the same level as line out assuming you use the Volume “High” setting."

As sansafix said above and I have found, the volume of the fuze and the dock are the same.  

If they had different impedance levels, would their respective volumes be the same when plugged into the same source?

For the record, in my office set up, the fuze by itself or with the dock does not produce a signal as loud as the CD player.   

Another interesting point relates to what Calv said.  I get the same “short circuit” 60 cycle buzz plugging the fuze into the dock as I get when I plug in the headphone jack, so what ever internal resistive measures are taken at the level of the headphone jack (not to fry the ICs), they are likewise employed at the level of the line out for the same reason… So if I follow what you’re saying, if the resistance is applied by the dock and not by the fuze, that could be why your line out cable is louder and why the fuze requires a voltage to signal that the dock is in place before it outputs the unprotected line level.

axon

@14124all - did you notice this fix from the new FW changelog today:

• Standard line out level is not used in Dock mode 

Maybe this fixes the overdrive problem with your cable?  Or is this just referring to the change to a fixed output?  Let us know!

Message Edited by Skinjob on 12-12-2008 06:52 PM

Doh!!! Found that cable mod is giving very small if any increase in level! Nearly all of my testing has been on a Klipsch Promedia. It has a mini input cable and a aux in jack that can mix inputs. I assumed the input impedance was the same on both, it’s not! Since the mod cable was a mini plug, I plugged into the aux input which has a higher impedance. I was plugging the Dock into the mini input cable which has a lower impedance. When plugging either into the same input, they are nearly the same. Looks like the cable jump will be a convenience item that will not help the car stereos. We will probably have to live with what we have here.

If axon is correct and I think he means the lineout and headphone out share the same signal, than the Fuze is not utilizing the AMS chip lineout pins. If it did, they would be independant signals (externally form the chip due to the dedicated outputs).

Oh well, you have “lineout” cable if you want it, but it looks like you can get nearly the same results out of the headphone jack.

That’s kind of depressing if line-out is not accessible, though I don’t see why they wouldn’t simply allow the use of the pre-amped output from the AS3525.  The AS3525 also has a third output, already amped for external speakers, though I am sure this output is not enabled in the Fuze.  Thanks so much for all your research into this, hopefully there is a way to access a pre-amped signal from the Fuze, I would hope that SanDisk would have taken advantage of this great chip on what is supposed to be their flagship player.  

Personally, I think both preout and ampout are used, but I can’t prove it. Tracing the circuits on the board is nearly impossible due to it being multilayer. The hint to both being used was when sansafix said the lineout should provide a better SNR. If the lineout is a tap off of the ampout this probably would not be true.

Line out is separate from the Headphone output.  It provides an improvement in Audio quality .

Thank you sansafix! Well, I thought so! That means the Docks and the modded cable are correct lineout at .7v rms p-p. This is giving us the best available signal the Fuze has.

I got on this forum looking for line-out cable info and this is my first post. Inspired by 14124all’s work (thanks!), I’ve been performing some experiments on my Fuze, using a voltmeter, power supply, and oscilloscope.  FW version 01.01.11A.

First, the Sandisk cable connector: there are two little rectangular holes on the back (non-logo) side, and two hooks

extending into the holes from the front side.  The hooks hold the halves of the shell together.  If you can get pieces of

shim material or similar to fit snugly down into the holes, they should press the hooks out of the way, enabling you to

separate the halves without breaking off the hooks.  The hooks on my connector broke.  The solder on the PCB is RoHS-compliant

and did not stick to my 30-gauge test wires well, so I wicked it out with fine solder braid and applied some good old-

fashioned tin-lead stuff.

Now for my electrical measurements.  Pin 7: the line output switches on with 1.82 V and higher on this pin (I tested up to

3.3 V).  The input resistance of pin 7 is 9.1 kohms with Fuze power turned on; higher with power off.  Line out AC

amplitude is not affected by pin 7 voltage, other than being switched on and off.

Headphone output: 0 V DC offset.  Output impedance of the headphone amplifier is approximately 20 ohms.  Level into

a high-impedance load, with volume setting on max and “system setting” = High, is the same as that of the line output. 

But of course as others have said, to get this level you must crank up the volume setting, and remember to turn it back

down when you plug your headphones back in.

Line out: 1.44 V DC offset.  I saw a maximum AC peak-peak output amplitude of almost 2 V, which

corresponds to 0.7 V rms (by the way, what do you guys mean by rms p-p?  Is this the same as rms AC?).  Output

impedance is very low, on the order of 1 ohm or so (I didn’t load it down enough to measure really accurately out of fear

of blowing the amp).  Output level is constant with volume settings above 1/4, and not affected by the “system setting”. 

This definitely is not the same amplifier as what drives the headphones, even though the max amplitudes are the same. 

Equalizer: line out is affected by the equalizer.  I found that when using “Custom EQ”, if I increased all the frequencies

equally, the amplitude went up but flatness became worse.  With 0 dB settings, flatness was 0.5 dB from 50 Hz to 6 kHz,

and 1.5 dB out to 20 Hz and 16 kHz (the limits of my test tones).  With 12 dB settings, peaks occurred at 400 Hz, 1.25

kHz, and 4 kHz, with 4 dB difference between the peaks and the adjacent valleys.  Even though the 0 dB setting

produces a lower level, the fidelity is better.

When I drive my car stereo with the Fuze’ line out, the volume is somewhat lower than that of a CD, but not significantly. 

In my application, it is better to change the volume on the stereo than to mess with the equalizer settings.

USB 5V: if I plug the USB cord into a PC, both the headphone output and the line out are turned off.  But if 5 V is applied

through the USB cable while the data lines are left unconnected, the Fuze outputs music via either output, while it

consumes 300-350 mA and indicates it is charging.  This means a single car cable can be used to provide line out to a

stereo, and charge the Fuze at the same time.

Some Very Cool Informantion waino :slight_smile:

thanks…

axon

@waino wrote:

by the way, what do you guys mean by rms p-p?

I believe “p-p” = “peak to peak”

So is there anyone here that makes custom LOD for the Fuze to connect to portable headphone amplifiers like the ones people make for iPods over on Head-Fi forums?  Not really sure where to find the dock materials and such, but I’m sure someone has.

Any info would be appreciated!  Thanks!

My understanding is that you don’t really need a docking cable because Sansafix said the line out is exactly the same as the headphone jack at 100%.

This was also verified by waino’s post above.

axon

Message Edited by axon01 on 12-17-2008 02:39 AM

I don’t get why Sansafix would say that the lineout was separate from the headphone jack and provided an improvement in audio quality then…  :cry:

@zonto wrote:
I don’t get why Sansafix would say that the lineout was separate from the headphone jack and provided an improvement in audio quality then…  :cry:

well, that one is easy: sandisk want’s to sell accessories like docking stations. So people at least have to believe that they improve audio quality, because they use line out.

Besides that, I also think that line out should not be influenced by eq settings. I don’t consider that a bug, but a fauxpas that reveals that line out passes the same signal processing “units” as headphone out does and for that reason can’t provide better quality anyway.

The non (or barely) existing impedance in line out should be a WARNING for everyone making their own cables to not underestimate the risk of a short circuit, especially when dealing with phone jacks. Please use at least a small resistor.

I said the line out signal level is the same as max headphone out.  However,  the Line out signal path is in fact separate from the heaphone path and has better performance.

Hey 1412

I have a thought for you…  do you think it’s possible to make a Fuze to ipod adapter? 

Just enough rewiring to be able to use some of the ipod docks and peripherals that are on the market.

axon

That would be glorious.  I mean if we have this site here then wouldn’t it be possible to take the information that’s been posted about the pins and make something similar to the Ipod LODs that are so prevalent over on head-fi?  There should just be a R, L, and ground and then that internal connection right?

I should point out that Line Out’s very low impedance was determined by driving a resistive load tied to a 1.44 V DC power supply, so as not to introduce any DC currents.  The smallest resistor I tested with was 33 ohms.

I have no idea what would happen if you shorted the output to ground or even tried to drive 33 ohms to ground.  You might well blow something up all right!