You are welcome.
Please note that :
a) There is now an offer for just the SEN V18 only, and not CEN/SEN V18 combined.
b) The CEN/SEN V18 is of course still available.
c) The JFET set is optimised for the SEN V18 circuit and hence may not be perfect quads in that sense.
I would encourage those who have the CEN V18 to try it out.
It is not a single ended but push-pull circuit, so will have less even harmonics, especially if you are not using both balanced outputs of the DAC.
Patrick
Please note that :
a) There is now an offer for just the SEN V18 only, and not CEN/SEN V18 combined.
b) The CEN/SEN V18 is of course still available.
c) The JFET set is optimised for the SEN V18 circuit and hence may not be perfect quads in that sense.
I would encourage those who have the CEN V18 to try it out.
It is not a single ended but push-pull circuit, so will have less even harmonics, especially if you are not using both balanced outputs of the DAC.
Patrick
If you use 4 pairs of 2SK170BL/2SJ74BL (8 FETs) per PCB (one balanced channel), it will be a direct drop-in replacement.
As a bonus you half the DC current consumption.
It has never been a pain with any of our projects, has it ?
Your only "pain" is to get P-N matched 2SK170/2SJ74 pairs.
And you need different heatsinks as in SEN, of course.
Patrick
As a bonus you half the DC current consumption.
It has never been a pain with any of our projects, has it ?
Your only "pain" is to get P-N matched 2SK170/2SJ74 pairs.
And you need different heatsinks as in SEN, of course.
Patrick
Last edited:
Patrick the Sen only package is suddenly a very attractive deal but I have to admit to getting a little confused with V18, SE, push-pull, there is now a new variant?.
There are a bunch (seven and rising) of us over in pinkfish with Buffalo-II and -III in Stereo mode, dual mono mode (two Buffalos one Legato 8 dacs per channel, converter are 4mA each in balanced mode )
The dual mono can easily be reduced in the number of converters per channel, one through eight..
We all have SE equipment
Could you perhaps find time to do a summery for requirements suitable to our application and bearing in mind we are very amateur guys and often, well almost always struggle to get there.
Many thanks
Tony
There are a bunch (seven and rising) of us over in pinkfish with Buffalo-II and -III in Stereo mode, dual mono mode (two Buffalos one Legato 8 dacs per channel, converter are 4mA each in balanced mode )
The dual mono can easily be reduced in the number of converters per channel, one through eight..
We all have SE equipment
Could you perhaps find time to do a summery for requirements suitable to our application and bearing in mind we are very amateur guys and often, well almost always struggle to get there.
Many thanks
Tony
Last edited:
I think I will soon have the raw material to do my own matching and picking of SJ74 so there should be no problem also trying out the CEN I already have the CEN PCBs and sinks from the evaluation pack so that will not be a problem.
I remember Davide reporting that both the SEN and CEN were very good - and very different, but that was not with the ES9018.
Having the choice should I pick pairs with as high Idss as possible?
Can't wait to hear both of them them with the F5X.
Cheers,
Nic
I already have the CEN PCBs and sinks from the evaluation pack so that will not be a problem.I remember Davide reporting that both the SEN and CEN were very good - and very different, but that was not with the ES9018.
Having the choice should I pick pairs with as high Idss as possible?
Can't wait to hear both of them them with the F5X.
Cheers,
Nic
> I have to admit to getting a little confused with V18, SE, push-pull, there is now a new variant?
If you take some time to read through the original GB thread for the SEN/ CEN IV Eval kit, you will know that the original design was for PCM1704 and the like. To satisfy public demand for a circuit to suit the ES9018 we designed and released a V18 later. 18 is chosen (alias 9018) such that we have enough numbers in between to update the PCM1704 boards. There have always been the CEN and the SEN variants since day one. Paul has already tried to explain to you on the other forum. Please read the article for more details. Another good basic reference is JFET Frontiers Part 1&2 by Erno Borbely. There might be pdf archives on the net. Please search.
The ES9018 has balanced output. But no one stops you from using one phase of the output for SE equipment only. If you want to use both phase of the output, then you need to do BAL->SE conversion. Paul used a transformer, which is one way. A semiconductor solution is probably more complicated (unless you are willing to have caps in the signal path). Each solution has its own sonic signature. So all a matter of taste.
If you find this difficult, then maybe one of you should just try to copy and paste what Nic or Paul has, before you start building 10's of these.
Or you wait fr our 9018 integrated solution.
Patrick
If you take some time to read through the original GB thread for the SEN/ CEN IV Eval kit, you will know that the original design was for PCM1704 and the like. To satisfy public demand for a circuit to suit the ES9018 we designed and released a V18 later. 18 is chosen (alias 9018) such that we have enough numbers in between to update the PCM1704 boards. There have always been the CEN and the SEN variants since day one. Paul has already tried to explain to you on the other forum. Please read the article for more details. Another good basic reference is JFET Frontiers Part 1&2 by Erno Borbely. There might be pdf archives on the net. Please search.
The ES9018 has balanced output. But no one stops you from using one phase of the output for SE equipment only. If you want to use both phase of the output, then you need to do BAL->SE conversion. Paul used a transformer, which is one way. A semiconductor solution is probably more complicated (unless you are willing to have caps in the signal path). Each solution has its own sonic signature. So all a matter of taste.
If you find this difficult, then maybe one of you should just try to copy and paste what Nic or Paul has, before you start building 10's of these.
Or you wait fr our 9018 integrated solution.
Patrick
Last edited:
Thanks for your patience Patrick, I think the penny has dropped.
Waiting for 9018 integrated solution seems ideal, I assume the integrated solution architecture is still in design stage so may be Cen or Sen based, and balanced output or would you consider including a semiconductor balanced to single ended conversion on the end.
Tony
Waiting for 9018 integrated solution seems ideal, I assume the integrated solution architecture is still in design stage so may be Cen or Sen based, and balanced output or would you consider including a semiconductor balanced to single ended conversion on the end.
Tony
Hi Patrick
Will your SEN-ES9018 integrated solution be a DIY amenable product, and/or a fully finished, retailable offering at a price to match its performance- i.e. high?
I changed the IV resistors in my SEN to 150R Charcrofts, and installed some larger coupling caps (470u Nichicon ES types) to match. It sounds- I think- less good! After I've checked over my work and found the presumed bad joint or snagged wire responsible- or after my paranoia settles and I realise everything is well after all!- I'll remove the foil on the battery box and reinvestigate the buzz for you (which is still absent despite the changes I made).
As I explained my CD transport's output is galvanically isolated, and the unit is an un-earthed, Class 2 design anyway. And so, as Buffalo +SEN feed a floating transformer primary, they are floating too, and I found it necessary to ground the DAC's ground to mains earth to remove some 100 Hz buzz on the output. (This buzz is not the same as the original one- such grounding left that unaffected)
Best wishes
Paul N
Will your SEN-ES9018 integrated solution be a DIY amenable product, and/or a fully finished, retailable offering at a price to match its performance- i.e. high?
I changed the IV resistors in my SEN to 150R Charcrofts, and installed some larger coupling caps (470u Nichicon ES types) to match. It sounds- I think- less good! After I've checked over my work and found the presumed bad joint or snagged wire responsible- or after my paranoia settles and I realise everything is well after all!- I'll remove the foil on the battery box and reinvestigate the buzz for you (which is still absent despite the changes I made).
As I explained my CD transport's output is galvanically isolated, and the unit is an un-earthed, Class 2 design anyway. And so, as Buffalo +SEN feed a floating transformer primary, they are floating too, and I found it necessary to ground the DAC's ground to mains earth to remove some 100 Hz buzz on the output. (This buzz is not the same as the original one- such grounding left that unaffected)
Best wishes
Paul N
Paul,
If your question is whether we shall have a poll here for what the public wants before we design, the answer is no.
I believe it art, and you cannot make art by democratic voting.
The first sentence I quoted in the F5X thread is that there are 100 ways to skin a cat.
I share with you my way, and you can decide whether you like it or not.
We are not here to win the popularity contest.
So as long as those who like our work really enjoy what we share, that is good enough.
And apparently, there are quite a few who do, and not just the F5X gang.
Having said that, where there are options that does not affect the sonic results, we are open for suggestions.
There have been many occasions like that in the F5X GB. Nic will testify to that.
We do not design to cost.
We don't intend to compete with Cambridge Audio for value for money.
We design to what we think is good audio engineering.
But at the same time, we are not known for blindly applying teflon caps, Black Gate NX, naked Vishays, ....., etc.
We think we choose our components sensibly.
Are they expensive ? Considering what you get, I don't think so. In fact, not all all.
But that is a judgement that other will have to make.
There are much cheaper ways to do DIY.
There are enough ebay dealers offering cheap PCBs.
One day, we expect some of our work will get copied.
So we concentrate on delivering a level of quality that makes ebayer not even bother to try.
Because top quality IS expensive, and bargain dealers won't want to spend the effort to achieve such quality levels.
This is our goal and our philosophy, as we have also explained in our website.
Patrick
If your question is whether we shall have a poll here for what the public wants before we design, the answer is no.
I believe it art, and you cannot make art by democratic voting.
The first sentence I quoted in the F5X thread is that there are 100 ways to skin a cat.
I share with you my way, and you can decide whether you like it or not.
We are not here to win the popularity contest.
So as long as those who like our work really enjoy what we share, that is good enough.
And apparently, there are quite a few who do, and not just the F5X gang.
Having said that, where there are options that does not affect the sonic results, we are open for suggestions.
There have been many occasions like that in the F5X GB. Nic will testify to that.
We do not design to cost.
We don't intend to compete with Cambridge Audio for value for money.
We design to what we think is good audio engineering.
But at the same time, we are not known for blindly applying teflon caps, Black Gate NX, naked Vishays, ....., etc.
We think we choose our components sensibly.
Are they expensive ? Considering what you get, I don't think so. In fact, not all all.
But that is a judgement that other will have to make.
There are much cheaper ways to do DIY.
There are enough ebay dealers offering cheap PCBs.
One day, we expect some of our work will get copied.
So we concentrate on delivering a level of quality that makes ebayer not even bother to try.
Because top quality IS expensive, and bargain dealers won't want to spend the effort to achieve such quality levels.
This is our goal and our philosophy, as we have also explained in our website.
Patrick
As to your buzzing problem, we built a new module today to try to duplicate your problem without success.
We suspect that it is a grounding issue with your particular setup.
Sometimes there is no universal solution for this. And you already have one which seems to work.
We do not believe it to be oscillation, which is at much higher frequencies.
To give you an example, my own preamp is battery powered.
Left and right channel both have floating Gnd, but fully balanced.
They are connected to a CD player with balanced output where the Gnd of both channels are connected to mains earth.
So in theory no chance of mains pick-up.
I still had a buzz which was solved simply by earthing the preamp case.
The preamp circuit Gnds are still floating (relative to mains earth) though.
And that is a circuit using fixed rails, not floating supplies.
Patrick
We suspect that it is a grounding issue with your particular setup.
Sometimes there is no universal solution for this. And you already have one which seems to work.
We do not believe it to be oscillation, which is at much higher frequencies.
To give you an example, my own preamp is battery powered.
Left and right channel both have floating Gnd, but fully balanced.
They are connected to a CD player with balanced output where the Gnd of both channels are connected to mains earth.
So in theory no chance of mains pick-up.
I still had a buzz which was solved simply by earthing the preamp case.
The preamp circuit Gnds are still floating (relative to mains earth) though.
And that is a circuit using fixed rails, not floating supplies.
Patrick
Last edited:
Hi Patrick
That was most definitely not my question! My question was will your finalised ESS9018-SEN unit be available as a completed, assembly pcb subassembly- like Twisted Pear's Buffalo, or as a finalised, encased fully finalised unit for retail, or both? My question was prompted by your suggestion that some of those planning to build using the buffalo and SEN boards might want to wait for your finished product instead. I was just trying to clarify what they might expect from you, so they can best decide how to proceed.
For what it's worth I think the first buzz I reported is the result of rf pickup by the floating circuitry- which may be exciting some oscillation- and demodulation of af noise carried by the rf. Though I may ultimately fail to prove this due the limitations of my test equipment. The solution to this sort of problem- in my experience- is appropriate grounding and shielding (as you say), good layout, and the use of appropriate stoppers resistors and, sometimes, ferrites. Even measuring the circuit can affect its state of tune however, and cures tend to be whatever works in practice (like my- ungrounded- copper foil) to ensure rock like stability.
I was not implying that your circuit is intrinsically prone to oscillation or pickup. I suspect I could build my unit with a slighly different layout many times over and not experience any problems- I just got unlucky
Hope that's clear, and hopefully, uncontroversial
Best wishes
Paul N
That was most definitely not my question! My question was will your finalised ESS9018-SEN unit be available as a completed, assembly pcb subassembly- like Twisted Pear's Buffalo, or as a finalised, encased fully finalised unit for retail, or both? My question was prompted by your suggestion that some of those planning to build using the buffalo and SEN boards might want to wait for your finished product instead. I was just trying to clarify what they might expect from you, so they can best decide how to proceed.
For what it's worth I think the first buzz I reported is the result of rf pickup by the floating circuitry- which may be exciting some oscillation- and demodulation of af noise carried by the rf. Though I may ultimately fail to prove this due the limitations of my test equipment. The solution to this sort of problem- in my experience- is appropriate grounding and shielding (as you say), good layout, and the use of appropriate stoppers resistors and, sometimes, ferrites. Even measuring the circuit can affect its state of tune however, and cures tend to be whatever works in practice (like my- ungrounded- copper foil) to ensure rock like stability.
I was not implying that your circuit is intrinsically prone to oscillation or pickup. I suspect I could build my unit with a slighly different layout many times over and not experience any problems- I just got unlucky
Hope that's clear, and hopefully, uncontroversial
Best wishes
Paul N
Not at all Paul. I was not worried about accusations re your buzz problem. We know how well SEN works.
I merely want to understand why, and whether we have overlooked things. A ground case is in any case good practice.
As to what form our ES9018 solution will take ?
In its first version (we don't want you to wait too long), it will probably be a single PCB which accommodates the 9018, XO's (maybe with options for Dual XO), most likely SEN IV (simply because of lack of availability of P-JFETs), power supply or supply connections, regulators (if any), chargers, etc. As in the F5X solution, we might also offer a case to suit, so that you can (but don't have to) build a complete DAC with SPDIF input according to our design.
We chose such a solution because I just release how big the communication gap is when mailing with Tony. Either we shall end up spending hundreds of hours in explaining, or we simply avoid all these by providing a turn-key solution.
We might (not decided yet by any means) also make provisions for I2S or DSD input. But we do believe that you cannot design these things on their own and just plug them together later. They will work, for sure, but whether this is the optimal solution is open to debate. I am a strong believer of "one clock drives all". And that makes life difficult to do I2S or DSD unless we do the front end as well. And that in turn is a lot of effort that will take time if we want to do it well. And we do want to do it well, much better than average.
Patrick
I merely want to understand why, and whether we have overlooked things. A ground case is in any case good practice.
As to what form our ES9018 solution will take ?
In its first version (we don't want you to wait too long), it will probably be a single PCB which accommodates the 9018, XO's (maybe with options for Dual XO), most likely SEN IV (simply because of lack of availability of P-JFETs), power supply or supply connections, regulators (if any), chargers, etc. As in the F5X solution, we might also offer a case to suit, so that you can (but don't have to) build a complete DAC with SPDIF input according to our design.
We chose such a solution because I just release how big the communication gap is when mailing with Tony. Either we shall end up spending hundreds of hours in explaining, or we simply avoid all these by providing a turn-key solution.
We might (not decided yet by any means) also make provisions for I2S or DSD input. But we do believe that you cannot design these things on their own and just plug them together later. They will work, for sure, but whether this is the optimal solution is open to debate. I am a strong believer of "one clock drives all". And that makes life difficult to do I2S or DSD unless we do the front end as well. And that in turn is a lot of effort that will take time if we want to do it well. And we do want to do it well, much better than average.
Patrick
BTW I know there are some discussions about single ended output with SEN 9018.
As you quite rightly pointed out, I almost exclusively build balanced circuits, so this is also the natural way to use SEN or CEN.
I am not aware of any elegant, simple, high performance BAL to SE circuits.
Transformer is one simple and elegant solution, but not necessarily everyone's cup of tea.
The last thing I want to do is to use an INA137 or the like (from Analog Device and from THAT) to do the conversion.
So the single ended solution from us would most likely be just to let you use the positive phase of the balanced output.
If a lot of people hate us because of that, then maybe we shall be soft-hearted enough to make space provision of an add-on module with an INA137.
But we shall NOT supply the module, for sure. It violates the principle of XEN.
Patrick
As you quite rightly pointed out, I almost exclusively build balanced circuits, so this is also the natural way to use SEN or CEN.
I am not aware of any elegant, simple, high performance BAL to SE circuits.
Transformer is one simple and elegant solution, but not necessarily everyone's cup of tea.
The last thing I want to do is to use an INA137 or the like (from Analog Device and from THAT) to do the conversion.
So the single ended solution from us would most likely be just to let you use the positive phase of the balanced output.
If a lot of people hate us because of that, then maybe we shall be soft-hearted enough to make space provision of an add-on module with an INA137.
But we shall NOT supply the module, for sure. It violates the principle of XEN.
Patrick
Last edited:
Hi Patrick
Thanks for the clarification. So a "DIY" solution, as I define it- good!
One thing. I've tried the ESS9018/Buffalo II board with three other IV stages than SEN and, in all cases, found that taking just one phase of the balanced ouput for SE use resulted in a very noisy and distorted output signal. I'll try this with SEN soon, but I suspect it will be the same
I could do with an SE optiont too. I spent a long time recently upgrading some amplifiers for a friend, and need something to demonstrate them with!
Paul
Thanks for the clarification. So a "DIY" solution, as I define it- good!
One thing. I've tried the ESS9018/Buffalo II board with three other IV stages than SEN and, in all cases, found that taking just one phase of the balanced ouput for SE use resulted in a very noisy and distorted output signal. I'll try this with SEN soon, but I suspect it will be the same
I could do with an SE optiont too. I spent a long time recently upgrading some amplifiers for a friend, and need something to demonstrate them with!
Paul
Patrick,
Do you have a specific suggestion on how to use the SEN I/V-stage with single ended (SE) amps? Paul found a transformer solution, but I would like to hear if you might have a better one - apart from getting a balanced amp
What is the estimated ETA of the integrated XEN ES9018/SEN solution?
Will it (also) have SE output?
Cheers,
Nic
Do you have a specific suggestion on how to use the SEN I/V-stage with single ended (SE) amps? Paul found a transformer solution, but I would like to hear if you might have a better one - apart from getting a balanced amp
What is the estimated ETA of the integrated XEN ES9018/SEN solution?
Will it (also) have SE output?
Cheers,
Nic
