The advancement in semiconductor technology is anything but unbelievable.
A university central computer in the 80's still used punch cards and the thing occupied an entire building.
They probably had less computer power than a tablet of today, and certainly thousand times slower.
That in 30 years, and there seems to be no end in sight, as yet.
If only software efficiency can catch up with that.
Patrick
A university central computer in the 80's still used punch cards and the thing occupied an entire building.
They probably had less computer power than a tablet of today, and certainly thousand times slower.
That in 30 years, and there seems to be no end in sight, as yet.
If only software efficiency can catch up with that.
Patrick
I think the buffer, how it is made, in SMD, sounds better then my prototype Hawksford Buffer with Sowter coils. Maybe it is the compactness, maybe layout, maybe mechanical things we do not 100% understand.
I got certainly hocked on SMD. I wish i would be better equipped for micro work like this.
I got certainly hocked on SMD. I wish i would be better equipped for micro work like this.
The compactness and the layout makes a big difference IMHO.
That is also why our ES9022 module is deliberately so small and have caps on the reverse side.
Certainly not design for manufacture.
They take me on average 3 hours to make one, including testing !!
To do SMD is not difficult.
You need a ring light with a big magnifying glass in the middle (from e.g. Conrad), and a good soldering iron with a very sharp tip.
I bought an Oki Metcal recently. Love it to bits.
🙂
Patrick
That is also why our ES9022 module is deliberately so small and have caps on the reverse side.
Certainly not design for manufacture.
They take me on average 3 hours to make one, including testing !!
To do SMD is not difficult.
You need a ring light with a big magnifying glass in the middle (from e.g. Conrad), and a good soldering iron with a very sharp tip.
I bought an Oki Metcal recently. Love it to bits.
🙂
Patrick
It is more difficult than good old through hole parts soldering. Nothing will change that. Also new equipment is needed: magnifying glasses, special soldering tools, special tweezers etc etc. An important aspect is the quality of your vision too. I think the size of some parts is really not meant for detection by human eyes and manual soldering. With soldering TH parts I never got a headache contrary to SMD work.
Tube amps can be repaired with the same tools care are repaired with. Some parts have cola tin like proportions. For people used to such parts SMD is passing a threshold too much.
BTW the lamps with magnifying glasses from the german company Waldmann are very good. Also this Maxdetail device can be of great help:
http://www.eschenbach.com/ed678c56-...4d88a/products-telescopes-galilean-detail.htm
Last edited:
The Metcal is very good, certainly.
I am very short sighted, so working in the near field is possible for me without lens.
You will be amazed but Mishima San, the guy that builds the Lyra cartridges, works without lens.
I am just not that experienced. Where i come from, though hole rules but this little thingy changed my mind.
I am very short sighted, so working in the near field is possible for me without lens.
You will be amazed but Mishima San, the guy that builds the Lyra cartridges, works without lens.
I am just not that experienced. Where i come from, though hole rules but this little thingy changed my mind.
Even at work, it can be felt increasingly that through-hole components choices are disappearing.
They are more and more being replaced by their SMD counterparts.
SMD parts are cheaper to produce, and PCBs are also cheaper without drilled holes.
And every other cost, such as mechanical enclosure, etc. follows.
Only knobs and switches remain the same in size, as our fingers are not getting more delicate.
And electro-mechanical parts often dominate costs these days.
Patrick
They are more and more being replaced by their SMD counterparts.
SMD parts are cheaper to produce, and PCBs are also cheaper without drilled holes.
And every other cost, such as mechanical enclosure, etc. follows.
Only knobs and switches remain the same in size, as our fingers are not getting more delicate.
And electro-mechanical parts often dominate costs these days.
Patrick
I simply was afraid that through hole sounds better.
This must not be the case, i learned today.
This must not be the case, i learned today.
Let's agree that cost is the driving factor (not practical usefulness). Of course some qualities of SMD are better. Just compare a through hole built DAC with the same DAC in SMD form. The lack of lead wires is a good thing in most cases when high speed devices are concerned. I can not build any DAC without SMD parts anymore as some SMD parts function way better than their TH counterparts !
BTW I am curious how this filter changes sound character as I am already very satisfied with our ES9023 DAC in standard form. In fact I did/do not think it needs any extra's. Well, it sounds so good that I can live with it for quite some time from now on.
Will there be a group buy of this filter or the PCB in any form ?
BTW I am curious how this filter changes sound character as I am already very satisfied with our ES9023 DAC in standard form. In fact I did/do not think it needs any extra's. Well, it sounds so good that I can live with it for quite some time from now on.
Will there be a group buy of this filter or the PCB in any form ?
Last edited:
Yes, i could live with it too but the filter-buffer puts it into a higher class. The sound is much more fleshed out and real. OK, i have to drive 5m of cable.
The ES9022 Zout is about 240R, the buffer about 25R.
Even with 5m of cable I still consider 240R low enough.
So IMHO the JFET follower is not making the difference.
Patrick
Even with 5m of cable I still consider 240R low enough.
So IMHO the JFET follower is not making the difference.
Patrick
I have searched the forum but can't find the schematic for this project(if available). Can you show a link to it please?
Thank you.
I have a lot of experiences with buffers and a good one can improve the sound, even when it does not contain a filter. Why that is even in a bandwidth limited system like CD i can only speculate. I know it is a tradition here to back up subjective claims with measurements or hard evidence. I can not offer this at the moment. Just listen to the buffer. If you like the result, i am happy.
Part of it is of cause the coil that is in SERIES with the unwanted sound over 20kHz. I think this solution is better then dumping the dirt into ground. I have modified CD players since the early 80th when they came out and the coil solution gave always the better ( subjective ) result.
Part of it is of cause the coil that is in SERIES with the unwanted sound over 20kHz. I think this solution is better then dumping the dirt into ground. I have modified CD players since the early 80th when they came out and the coil solution gave always the better ( subjective ) result.
Ahhh I can see now. Looks quite elegant and simple. Does the stage actually need the input LPF to be that complex?
The ES9022 board we talk about here has only a cap to ground as anti aliasing filter.
I found the sound very good but it had a tiny bit of artifice i could not get away.
I concluded that a bit of filtering could be beneficial.
My filter-buffer adds some some damping of frequencies over 20kHz and works in tandem with the cap already fitted. In combination it is a filter with constant coup delay, so i tried not to worsen the transit response. At least as little as a could. The final values have then be set by listening so this filter contains my personality in part.
I found the sound very good but it had a tiny bit of artifice i could not get away.
I concluded that a bit of filtering could be beneficial.
My filter-buffer adds some some damping of frequencies over 20kHz and works in tandem with the cap already fitted. In combination it is a filter with constant coup delay, so i tried not to worsen the transit response. At least as little as a could. The final values have then be set by listening so this filter contains my personality in part.
We have quite a numbers of emails asking to be put on the waiting list for the JG Filter Buffer.
We appreciate your interest and support very much, but ask you for your patience to give us some time.
We need to sort out how best to make this available to you without overloading ourselves.
And we WILL NOT forget that it is Joachim's intellectual property in the end.
Regards,
Patrick
PS of course the schematics is published and no one can / will stop you from aking your own version.
We appreciate your interest and support very much, but ask you for your patience to give us some time.
We need to sort out how best to make this available to you without overloading ourselves.
And we WILL NOT forget that it is Joachim's intellectual property in the end.
Regards,
Patrick
PS of course the schematics is published and no one can / will stop you from aking your own version.
The buffer can be build through hole with 1 x 2SK170BL and 3 x J310, J311 or J111, J112.
Other Fets work too if you understand the principle.
My MPP thread shows how. It is the buffer in the Paradise R2.
Other Fets work too if you understand the principle.
My MPP thread shows how. It is the buffer in the Paradise R2.
- Status
- Not open for further replies.