Page 9 of 23

Re: Toshiba MOSFETs with a twist

Posted: Wed Jul 26, 2017 4:49 pm
by nige2000
The akg k700 had two GND wires
That was mergered in the plug
Made it very simple

Re: Toshiba MOSFETs with a twist

Posted: Wed Jul 26, 2017 5:05 pm
by nige2000
jkeny wrote: Wed Jul 26, 2017 3:13 pm
nige2000 wrote: Wed Jul 26, 2017 1:59 pm is the superlux wired with two gnds one per side or is it common?
The common ground is disconnected & each side gets it's own ground - wouldn't be differential otherwise
yea but its a negative signal that replaces gnd?

Re: Toshiba MOSFETs with a twist

Posted: Wed Jul 26, 2017 5:20 pm
by jkeny
nige2000 wrote: Wed Jul 26, 2017 5:05 pm
jkeny wrote: Wed Jul 26, 2017 3:13 pm
nige2000 wrote: Wed Jul 26, 2017 1:59 pm is the superlux wired with two gnds one per side or is it common?
The common ground is disconnected & each side gets it's own ground - wouldn't be differential otherwise
yea but its a negative signal that replaces gnd?
I think this is where terminology confuses the issue - ground is just a return path for the signal - in a differential signalling system there is a signal on one line & it's identical signal on the other line but 180degree out of phase - the summation of these two signals is zero current so there is no return path. Some people say that the return path for each line is the opposite line but I can't see this.

My understanding does hit a problem if the signals are not exact opposites of each other & don't fully cancel each other out - so there's a small portion of each signal operating in single ended mode - where does the return current (difference signal) then return to it's source?

In a headphone situation, which is just a coil & diaphragm, I'm not sure what happens with any imbalance between differential signals ?

Re: Toshiba MOSFETs with a twist

Posted: Wed Jul 26, 2017 7:21 pm
by nige2000
jkeny wrote: Wed Jul 26, 2017 5:20 pm
nige2000 wrote: Wed Jul 26, 2017 5:05 pm
jkeny wrote: Wed Jul 26, 2017 3:13 pm
The common ground is disconnected & each side gets it's own ground - wouldn't be differential otherwise
yea but its a negative signal that replaces gnd?
I think this is where terminology confuses the issue - ground is just a return path for the signal - in a differential signalling system there is a signal on one line & it's identical signal on the other line but 180degree out of phase - the summation of these two signals is zero current so there is no return path. Some people say that the return path for each line is the opposite line but I can't see this.

My understanding does hit a problem if the signals are not exact opposites of each other & don't fully cancel each other out - so there's a small portion of each signal operating in single ended mode - where does the return current (difference signal) then return to it's source?

In a headphone situation, which is just a coil & diaphragm, I'm not sure what happens with any imbalance between differential signals ?
i think of it as push pull on the speaker coil,
the positive pulls it positive and the negative pulls it negative, they are the return path for each other
maybe more distortion? but maybe the extra current makes up for it subjectively?
maybe its called a bridged differential amp?
lots of confusion out there

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 1:20 am
by abraxalito
jkeny wrote: Wed Jul 26, 2017 5:20 pm My understanding does hit a problem if the signals are not exact opposites of each other & don't fully cancel each other out - so there's a small portion of each signal operating in single ended mode - where does the return current (difference signal) then return to it's source?
When the signals aren't exact opposites of each other there will be a common-mode component of the voltage. The headphone wire acts as an antenna but will be extremely inefficient at propagating audio frequencies into the ether. So the common-mode current associated with that CM voltage will be extremely small, not worthy of consideration.
In a headphone situation, which is just a coil & diaphragm, I'm not sure what happens with any imbalance between differential signals ?
Nothing - the current all returns to the source via the return cable. Another way of saying this is that the CM impedance is extremely high (Gigohms and beyond).

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 2:30 am
by jkeny
Thanks Richard
I seem to have killed a DAC chip so have to build another before I can report on differential sound with headbuff

Bit of a bugger as it's 32 pin VQFN package 0.5 mm pitch :(

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 2:58 am
by abraxalito
Ouch indeed for 0.5mm! I daren't go near even 0.65mm. So is it being dead the reason it wasn't talking to you over I2C?

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 3:19 am
by jkeny
abraxalito wrote: Thu Jul 27, 2017 2:58 am Ouch indeed for 0.5mm! I daren't go near even 0.65mm. So is it being dead the reason it wasn't talking to you over I2C?
I don't know if that's why I2C didn't work as I was getting sound out of it when I changed the strapping to hardware control instead of I2C control but it was also acting a bit odd - overheating at times & bad noise at other times so it was probably damaged & eventually failed

I think I probably damaged it due to overheating when trying to ensure all pins were properly soldered

I have soldered a second one but without overheating so hopefully I will have success with this - I use a Schmartboard for mounting this with some solder paste + hot gun

Why I wanted I2C control was to be able to access some of the nice features of this DAC like different gain settings (-6dB, -3dB, 0dB, 3dB, 6dB, 9dB, 12dB, 15dB) plus a high or low voltage output control (2.1V or 1V) as well as volume control & a new filter with no pre-ringing. With this level of control & 4.2Vrms differential outputs, it would seem a perfect DAC for headphone use?

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 3:35 am
by abraxalito
jkeny wrote: Wed Jul 26, 2017 3:12 pm
Wish I had a willing assistant like that :)
Ha - Wendy's found two sellers who are up for rewiring a fresh pair of Superluxes with a 4pin mini XLR at the end. Price is around 150rmb for the mod, the 'cans are 212rmb. A balanced pair for about 360rmb looks reasonable to me.

Re: Toshiba MOSFETs with a twist

Posted: Thu Jul 27, 2017 3:38 am
by jkeny
abraxalito wrote: Thu Jul 27, 2017 3:35 am
jkeny wrote: Wed Jul 26, 2017 3:12 pm
Wish I had a willing assistant like that :)
Ha - Wendy's found two sellers who are up for rewiring a fresh pair of Superluxes with a 4pin mini XLR at the end. Price is around 150rmb for the mod, the 'cans are 212rmb. A balanced pair for about 360rmb looks reasonable to me.
Yep, a balanced Superlux for about €40 is great price
You'll also need some 4 pin mini-xlr panel mounted sockets for connecting them into