I guess you means these ?
Free Shipping on A123 Racing Systems Nanophosphate Lithium Ion Packs.
I am sure they are fine and will play for 50 hours on one charge.
But aren't they a bit pricy at 135USD ?
In any case you need need 4 sets of batteries for two channel balanced, as they all swing different voltages.
I have zero knowledge about anything tube.
Paul (needsp) would be the man to ask, I guess.
Patrick
Free Shipping on A123 Racing Systems Nanophosphate Lithium Ion Packs.
I am sure they are fine and will play for 50 hours on one charge.
But aren't they a bit pricy at 135USD ?
In any case you need need 4 sets of batteries for two channel balanced, as they all swing different voltages.
I have zero knowledge about anything tube.
Paul (needsp) would be the man to ask, I guess.
Patrick
> I did not use the famous heatsinks and got the output stable at less than 2mV.
The famous heat sinks keep the temperature down for the 2SK369Vs running at 20mA for higher reliability.
They also keep the DC offset at the input close to zero, and not at the output.
Kirchoff's law takes care of the output.
But I am glad you like the music.
Cheers,
Patrick
The famous heat sinks keep the temperature down for the 2SK369Vs running at 20mA for higher reliability.
They also keep the DC offset at the input close to zero, and not at the output.
Kirchoff's law takes care of the output.
But I am glad you like the music.
Cheers,
Patrick
Got it...
Maybe I should look this side to solve my buzz issue.... I'll meditate that.
The 2SK170BL are merely warm without the sinks.
> Maybe I should look this side to solve my buzz issue.... I'll meditate that.
Read Paul's feedback on his noise problem, and also the pinkfish thread.
It has been solved before.
The art of Sen - pink fish media
(Disclaimer : I have no involvement in the pinkfish activities.)
> The 2SK170BL are merely warm without the sinks.
That is because you spread the heat into 3 FETs instead of one 2SK369V.
So you might get away without using heat sinks.
In our own measurements, we think that you will need to use 4x 2SK170BL to replace two 2SK369V for the same transconductance.
But then maybe a few ohms extra Zin is not so important.
Afterall, you are happy with the sound ? That's all matters.
As to copies of our work, our position remains the same.
The circuit is public property. Anyone can copy.
And we are genuinely pleased that you take the trouble to make a custom PCB for your own use.
But we as a matter of principle will not endorse anyone else's PCB layout other than our own.
Hope you will understand.
All these discussions will be irrelevant once we release our ES9018 integrated solution.
This will have everything on board, including regulators or passive power supply filters, XO's,
IV circuit, battery connections , SPDIF AND USB input for 32/384.
Will take time. Good things cannot be rushed.
Cheers,
Patrick
Read Paul's feedback on his noise problem, and also the pinkfish thread.
It has been solved before.
The art of Sen - pink fish media
(Disclaimer : I have no involvement in the pinkfish activities.)
> The 2SK170BL are merely warm without the sinks.
That is because you spread the heat into 3 FETs instead of one 2SK369V.
So you might get away without using heat sinks.
In our own measurements, we think that you will need to use 4x 2SK170BL to replace two 2SK369V for the same transconductance.
But then maybe a few ohms extra Zin is not so important.
Afterall, you are happy with the sound ? That's all matters.
As to copies of our work, our position remains the same.
The circuit is public property. Anyone can copy.
And we are genuinely pleased that you take the trouble to make a custom PCB for your own use.
But we as a matter of principle will not endorse anyone else's PCB layout other than our own.
Hope you will understand.
All these discussions will be irrelevant once we release our ES9018 integrated solution.
This will have everything on board, including regulators or passive power supply filters, XO's,
IV circuit, battery connections , SPDIF AND USB input for 32/384.
Will take time. Good things cannot be rushed.
Cheers,
Patrick
Last edited:
<snip>
All these discussions will be irrelevant once we release our ES9018 integrated solution.
This will have everything on board, including regulators or passive power supply filters, XO's,
IV circuit, battery connections , SPDIF AND USB input for 32/384.
Will take time. Good things cannot be rushed.
Cheers,
Patrick
I'm extremely interested, please do keep us up to day on any progress..
@Fred.
Great work! I think placing the SEN/CEN I/V below the Buffalo II/III will improve very significantly the implementation of the SEN/CEN circuit with the very popular TPA variant of the ES9018 DAC. Mine is a terrible wire clutter...
I am convinced that a SEN/CEN/ES9018 solution from Patrick and co. will be very good. Just a petty that many, less skilled DIY'ers, who have already spend a significant amount of money on the Buffalo have no "easy" way of trying out the SEN/CEN circuit.
Really a shame, as the SEN is a very significant better I/V than what most are currently using.
Cheers,
Nic
Great work! I think placing the SEN/CEN I/V below the Buffalo II/III will improve very significantly the implementation of the SEN/CEN circuit with the very popular TPA variant of the ES9018 DAC. Mine is a terrible wire clutter...
I am convinced that a SEN/CEN/ES9018 solution from Patrick and co. will be very good. Just a petty that many, less skilled DIY'ers, who have already spend a significant amount of money on the Buffalo have no "easy" way of trying out the SEN/CEN circuit.
Really a shame, as the SEN is a very significant better I/V than what most are currently using.
Cheers,
Nic
> I think placing the SEN/CEN I/V below the Buffalo II/III will improve very significantly the implementation of .....
It is the most terrible place you can put an analog circuit with a power supply that swings with the signal.
Guaranteed to pick up all the garbage from the digital circuitry running at 98MHz.
I do not understand why a few twisted wires for the current out and 2 resistors to make up a Vref is such a difficult task.
A convenient PCB layout does not necessary make a good layout in performance terms.
Patrick
It is the most terrible place you can put an analog circuit with a power supply that swings with the signal.
Guaranteed to pick up all the garbage from the digital circuitry running at 98MHz.
I do not understand why a few twisted wires for the current out and 2 resistors to make up a Vref is such a difficult task.
A convenient PCB layout does not necessary make a good layout in performance terms.
Patrick
Wow Patrick Thanks lot !
I did not know there was an "offspring" thread of the SEN discussion on Pink Fish... and an initiative close to mine.... probably could have saved me a few hours.... anyway...
Thanks again!
Fred
Hi Blitz,
It's not about the hassle of a group buy (check my GB for the DIY Curve Tracer to see I'm not afraid of that !.... that one would be piece of cake) it's about respect of intellectual property as EUVL expressed it. There is a commercial product for the SEN/CEN and I do not want to interfere with that.
Fred
> There is a commercial product for the SEN/CEN .....
For us it is never about financial gain.
Everyone in XEN earns a very comfortable living financially.
Anything we do in DIYA will never ever pay for our hours.
We charge you a nominal sum for PCBs as everyone else around.
And the heat sinks for the JFEts have been essentially the same price for years.
So they cover our cost and allow us to pay the person who handles all the orders some pocket money.
Nothing more.
It is about that we want you to build the circuit as we think you should.
Just read the last 10 posts or so and you can understand why.
Of course a purpose built integrated solution is the most convenient.
To do it properly, we have to start at the every front end.
i.e. our own 9018 implementation.
The devil is in the details, as they say.
Patrick
For us it is never about financial gain.
Everyone in XEN earns a very comfortable living financially.
Anything we do in DIYA will never ever pay for our hours.
We charge you a nominal sum for PCBs as everyone else around.
And the heat sinks for the JFEts have been essentially the same price for years.
So they cover our cost and allow us to pay the person who handles all the orders some pocket money.
Nothing more.
It is about that we want you to build the circuit as we think you should.
Just read the last 10 posts or so and you can understand why.
Of course a purpose built integrated solution is the most convenient.
To do it properly, we have to start at the every front end.
i.e. our own 9018 implementation.
The devil is in the details, as they say.
Patrick
Patrick,
No need to justify yourself. I was not being negative and I was certainly not judging. You have spent hours developping your products and there is no shame asking for a small premium to cover your costs and a fraction of your time (anyone who has ever been involved in a Group Buy will know it is hardly ever for money!). Please take no offense for the word "commercial" which I used for want of a better word...
Fred
No need to justify yourself. I was not being negative and I was certainly not judging. You have spent hours developping your products and there is no shame asking for a small premium to cover your costs and a fraction of your time (anyone who has ever been involved in a Group Buy will know it is hardly ever for money!). Please take no offense for the word "commercial" which I used for want of a better word...
Fred
Hi
I too am a bit bemused why the standard response of many, faced with integrating SEN with Buffalo, is to design- or await production of- an alternative pcb for SEN which plugs under the Buffalo board, like Twisted Pear's Legato or IVY IV stages! It's easy to hard wire the two boards together, as I showed here-
http://www.diyaudio.com/forums/digi...inimalistic-iv-converter-100.html#post3093382
I would have preferred mounting the SEN boards on either side of Buffalo, to minimise hf pickup of the type Patrick describes, but chose not to do so to make best use of the space in my case, and bcause such an arrangement would have required me to remove and re-orientate several of the Paul Hynes regulators to make space (see picture). I hoped the 35 mm gap between the boards, and, particularly, the ground plane on the Buffalo board would ensure good results, and so it has proved. But it's a compromise- not best practice!
Lazybutt- see the above link which also describes my initial buzz problem and how I resolved it
Blitz- the size of the coupling caps in SEN are determined by the value if the IV resistor used and the desired - 3bB low frequency point. (The IV R values are chosen to give the desired output voltage with a particular DAC output current). In practice the value needed will necessitate the use of electrolytics, but, as Patrick has explained, they operate with a constant voltage across them, and are therefore sonically benign. (Opinions on how benign differ- Patrick thinks completely, I’m still unsure- but certainly good enough to allow SEN to outperform all other IV stages I've tried by miles!) The input of the preceding stage, which can be DC linked as it's a cathode biassed triode with a grounded grid, would only affect the value of the caps required if its input impedance was of the same order as the IV resistor- not so in this case.
Paul N
I too am a bit bemused why the standard response of many, faced with integrating SEN with Buffalo, is to design- or await production of- an alternative pcb for SEN which plugs under the Buffalo board, like Twisted Pear's Legato or IVY IV stages! It's easy to hard wire the two boards together, as I showed here-
http://www.diyaudio.com/forums/digi...inimalistic-iv-converter-100.html#post3093382
I would have preferred mounting the SEN boards on either side of Buffalo, to minimise hf pickup of the type Patrick describes, but chose not to do so to make best use of the space in my case, and bcause such an arrangement would have required me to remove and re-orientate several of the Paul Hynes regulators to make space (see picture). I hoped the 35 mm gap between the boards, and, particularly, the ground plane on the Buffalo board would ensure good results, and so it has proved. But it's a compromise- not best practice!
Lazybutt- see the above link which also describes my initial buzz problem and how I resolved it
Blitz- the size of the coupling caps in SEN are determined by the value if the IV resistor used and the desired - 3bB low frequency point. (The IV R values are chosen to give the desired output voltage with a particular DAC output current). In practice the value needed will necessitate the use of electrolytics, but, as Patrick has explained, they operate with a constant voltage across them, and are therefore sonically benign. (Opinions on how benign differ- Patrick thinks completely, I’m still unsure- but certainly good enough to allow SEN to outperform all other IV stages I've tried by miles!) The input of the preceding stage, which can be DC linked as it's a cathode biassed triode with a grounded grid, would only affect the value of the caps required if its input impedance was of the same order as the IV resistor- not so in this case.
Paul N
Paul,
Did you see this ?
http://www.diyaudio.com/forums/anal...ced-single-ended-converter-2.html#post3159210
Cheers,
Patrick
Did you see this ?
http://www.diyaudio.com/forums/anal...ced-single-ended-converter-2.html#post3159210
Cheers,
Patrick
Hi
I made a small modification to the main regulator I use with the Buffalo board which feeds my SEN boards. (It's a Paul Hynes PR3 series regulator- I installed a common mode choke between the raw dc and the regulator). This involved some dis-assembly and reassembly of the whole DAC case and circuitry.An unwelcome consequence of this was a return of the apparent instability/rf issues I had when I first constructed the unit. With certain input connections, and particularly on switch on, there is an audible buzz on the outputs, which can be modulated by moving the wiring from the battery supplies or the wires between boards. Holding my hand near the circuitry abolishes it, but the buzz returns as I move away.
I strongly suspect that this problem is a result of locating the digital board directly above the SEN boards, as I and several others have (copying the arrangement from that used with Twisted Pear IV stages). See pictures here-
http://www.diyaudio.com/forums/digi...inimalistic-iv-converter-100.html#post3093382
I think I'm going to have to rearrange things withe SEN boards either side and in the same plane as the Buffalo board
Paul N
I made a small modification to the main regulator I use with the Buffalo board which feeds my SEN boards. (It's a Paul Hynes PR3 series regulator- I installed a common mode choke between the raw dc and the regulator). This involved some dis-assembly and reassembly of the whole DAC case and circuitry.An unwelcome consequence of this was a return of the apparent instability/rf issues I had when I first constructed the unit. With certain input connections, and particularly on switch on, there is an audible buzz on the outputs, which can be modulated by moving the wiring from the battery supplies or the wires between boards. Holding my hand near the circuitry abolishes it, but the buzz returns as I move away.
I strongly suspect that this problem is a result of locating the digital board directly above the SEN boards, as I and several others have (copying the arrangement from that used with Twisted Pear IV stages). See pictures here-
http://www.diyaudio.com/forums/digi...inimalistic-iv-converter-100.html#post3093382
I think I'm going to have to rearrange things withe SEN boards either side and in the same plane as the Buffalo board
Paul N
From your description I don't understand why you blame the noise on the stacking (which is the same before and after). To me it sounds more likely that it is the wiring that creates the problem, since the noise go and come when you touch it
Maybe no wire connections would be better
As long as you continue to cite me for thinking that a well done stacking will work better - I will continue to point out the inconsistency of your conclusions
Hi Nic
I was quoting Patrick’s opinion- I forgot he was reacting to you! Though as a researcher, you should welcome even an inadvertent citation
Actually, touching the wires this time doesn’t help (and in the original scenario, touching other parts of the circuit helped too). IMO it’s the whole assembly of boards, components and wiring that is prone to rf pickup and hf oscillation. I’ve had considerable experience of this before with alternative IV stages for DIY Paradise’s Monica DAC. A common grid IV stage using a high Gm triode was a particular nightmare. There, as now, what precisely is going on we’ll never know- even attaching probes to scope the circuit can affect its “state of tune”- all we can do is adopt good practice, i.e. gate/grid stoppers (which Patrick of course used) and good layout to reduce the interaction of rf fields from the DAC board with the IV stage and the interconnecting wiring. Putting the SEN boards to the side of the DAC board should help with this. If this doesn’t work, we either give up or try small component or wiring reorientations in the hope that this will curtail the problem. In the original case with SEN I stuck some ungrounded copper foil to the top of the battery boxes- the capacitive coupling to the foil apparently detuned or damped the circuitry. In the present case, holding my hand close to the circuit- or just standing close to the DAC!- stops the buzz. But as I back away it restarts....
I agree that optimally short, perpendicular connections between the stacked boards would be better than my present arrangement, but non-stacked should be better still. For an optimal no wires setup we’ll have to wait for XEN Audio’s product!
I’ll change the layout, shorten the connections to the batteries and report back. I can prove nothing though!
Paul N
I was quoting Patrick’s opinion- I forgot he was reacting to you! Though as a researcher, you should welcome even an inadvertent citation
Actually, touching the wires this time doesn’t help (and in the original scenario, touching other parts of the circuit helped too). IMO it’s the whole assembly of boards, components and wiring that is prone to rf pickup and hf oscillation. I’ve had considerable experience of this before with alternative IV stages for DIY Paradise’s Monica DAC. A common grid IV stage using a high Gm triode was a particular nightmare. There, as now, what precisely is going on we’ll never know- even attaching probes to scope the circuit can affect its “state of tune”- all we can do is adopt good practice, i.e. gate/grid stoppers (which Patrick of course used) and good layout to reduce the interaction of rf fields from the DAC board with the IV stage and the interconnecting wiring. Putting the SEN boards to the side of the DAC board should help with this. If this doesn’t work, we either give up or try small component or wiring reorientations in the hope that this will curtail the problem. In the original case with SEN I stuck some ungrounded copper foil to the top of the battery boxes- the capacitive coupling to the foil apparently detuned or damped the circuitry. In the present case, holding my hand close to the circuit- or just standing close to the DAC!- stops the buzz. But as I back away it restarts....
I agree that optimally short, perpendicular connections between the stacked boards would be better than my present arrangement, but non-stacked should be better still. For an optimal no wires setup we’ll have to wait for XEN Audio’s product!
I’ll change the layout, shorten the connections to the batteries and report back. I can prove nothing though!
Paul N
Paul,
I know Patrick well enough to listen carefully to his advice - so I do.
I have my SEN boards mounted diagonally above the buffalo (+ an extra voltage divider PCB) with numerous wires seeing the full noise radiation of the DAC chip. Possibly the worst possible implementation
But I have no noise issues whatsoever. Why?
I don't know, but my strong suspicion is on the Buffalo circuit - not the SEN. I am feeding the 5 onboard regulators (TPA and PH shunt types) from 5 independent pre-regulators (simple LM317 types, independent TX secondaries) to eliminate ground plane currents between the PSUs. Maybe this is it.
Anyway, in keeping with the advice of Patrick I think that if I was able to design PCBs I would place the SEN circuit, the voltage divider and a panel mount XLR socket on a single PCB, well away from the DAC. In this way I would only need to worry about shielded high quality connections for the current signal and Vref.
Who will do the layout for me
Cheers,
Nic
P.S. Of course all this will become obsolete with Patrick's ES9018 solution, that we all look forward to see (and hear the price)
I know Patrick well enough to listen carefully to his advice - so I do.
I have my SEN boards mounted diagonally above the buffalo (+ an extra voltage divider PCB) with numerous wires seeing the full noise radiation of the DAC chip. Possibly the worst possible implementation
But I have no noise issues whatsoever. Why?
I don't know, but my strong suspicion is on the Buffalo circuit - not the SEN. I am feeding the 5 onboard regulators (TPA and PH shunt types) from 5 independent pre-regulators (simple LM317 types, independent TX secondaries) to eliminate ground plane currents between the PSUs. Maybe this is it.
Anyway, in keeping with the advice of Patrick I think that if I was able to design PCBs I would place the SEN circuit, the voltage divider and a panel mount XLR socket on a single PCB, well away from the DAC. In this way I would only need to worry about shielded high quality connections for the current signal and Vref.
Who will do the layout for me
Cheers,
Nic
P.S. Of course all this will become obsolete with Patrick's ES9018 solution, that we all look forward to see (and hear the price
)