Took a little while to find the Kaneda schematics.
Here is the one from MJ Feb 2017 :
http://ph7dc.but.jp/dc_amp/no251mixIVC.html
It is actually a power opamp-based IV converter. They call it IVC in Japan.
Rails at +/-25V due to max rating for 2SK246/2SJ103.
Onsemi JFETs are even limited to 35V (+/-17.5V).
Patrick
Another Kaneda using SiC, this one for headphones :
http://champ.chips.jp/gazou5/audiopage45.html
Patrick
http://champ.chips.jp/gazou5/audiopage45.html
Patrick
I find it absolutely amazing that after 15 years, so many people are still chasing Semisouth SJEP120R100.
One can tell that many got disappointed after the last lottery :
https://www.diyaudio.com/community/threads/another-sjep-giveaway.394988/
Of course, other than that Nelson might still have plenty of them, there is only one known source.
And they are asking for 250USD a pair, with "matched date code", i.e NOT matched.
https://www.ebay.com/itm/202756261818
We went through our remaining inventory from the last GB 5 years ago :
https://www.diyaudio.com/community/...r100-curve-tracer-matched-pairs-quads.330916/
And found some "singles" which can still be grouped as Vgs-matched pairs (max 35mV, most < 20mV, measured at 55°C).
We are of course not as rich as Nelson to be able to afford giving them away.
We also do not want to see people getting them from us, with the sole purpose of selling them off later to make a quick profit.
So this is what we shall order, probably in 3~4 weeks time, in a new GB :
1 pair of SJEP120R100, Vgs matched to 40mV at 55°C
1 pair of 2SK3497, Vgs matched to 20mV at 55°C
Registered air mail from HK plus extra insurance
For 1800HKD per set, for 2 channels of J2 or F8.
We can only offer 13 sets in total, after which we shall also be depleted of our remaining "singles".
Not exactly a gift I know, but for serious J2/F8 builders, it is still a much better offer than available elsewhere.
When we are ready, we'll start a GB thread.
Patrick
One can tell that many got disappointed after the last lottery :
https://www.diyaudio.com/community/threads/another-sjep-giveaway.394988/
Of course, other than that Nelson might still have plenty of them, there is only one known source.
And they are asking for 250USD a pair, with "matched date code", i.e NOT matched.
https://www.ebay.com/itm/202756261818
We went through our remaining inventory from the last GB 5 years ago :
https://www.diyaudio.com/community/...r100-curve-tracer-matched-pairs-quads.330916/
And found some "singles" which can still be grouped as Vgs-matched pairs (max 35mV, most < 20mV, measured at 55°C).
We are of course not as rich as Nelson to be able to afford giving them away.
We also do not want to see people getting them from us, with the sole purpose of selling them off later to make a quick profit.
So this is what we shall order, probably in 3~4 weeks time, in a new GB :
1 pair of SJEP120R100, Vgs matched to 40mV at 55°C
1 pair of 2SK3497, Vgs matched to 20mV at 55°C
Registered air mail from HK plus extra insurance
For 1800HKD per set, for 2 channels of J2 or F8.
We can only offer 13 sets in total, after which we shall also be depleted of our remaining "singles".
Not exactly a gift I know, but for serious J2/F8 builders, it is still a much better offer than available elsewhere.
When we are ready, we'll start a GB thread.
Patrick
GB started, unfortunately only 8 sets left, not 13 sets.
https://www.diyaudio.com/community/...or-j2-vgs-matched-at-55-c.398229/post-7324167
Patrick
https://www.diyaudio.com/community/...or-j2-vgs-matched-at-55-c.398229/post-7324167
Patrick
The most recent Kaneda power amp published at the MJ magazine is the Kanada No. 279 of November 2021.
It was next to impossible to find the schematics.
This is actually traced back from a Vero board layout published at a JP DIY webpage :
http://ph7dc.but.jp/dc_amp/279-959-606type.html
What is interesting is the gain structure between the stages.
The SiC Mosfets from Rohm have rather low transconductace (1S or less).
Most of the open loop gain has therefore to come from the first two stages.
And there is quite a bit more NFB compared to, say, the First Watt J2.
With +/-15V rails, it is not delivering a lot of power either.
1st Stage gain ~ 18x
2nd Stage gain ~ 16x
O/P Stage gain ~ 12x
CLG ~ 25x
NFB ~ 140 (43dB)
Compensation is required for the high positive tempco of the SiC MOSFETs. a
The solution was an NTC mounted on the SiC to reduce the current through the bias resistors.
Patrick
It was next to impossible to find the schematics.
This is actually traced back from a Vero board layout published at a JP DIY webpage :
http://ph7dc.but.jp/dc_amp/279-959-606type.html
What is interesting is the gain structure between the stages.
The SiC Mosfets from Rohm have rather low transconductace (1S or less).
Most of the open loop gain has therefore to come from the first two stages.
And there is quite a bit more NFB compared to, say, the First Watt J2.
With +/-15V rails, it is not delivering a lot of power either.
1st Stage gain ~ 18x
2nd Stage gain ~ 16x
O/P Stage gain ~ 12x
CLG ~ 25x
NFB ~ 140 (43dB)
Compensation is required for the high positive tempco of the SiC MOSFETs. a
The solution was an NTC mounted on the SiC to reduce the current through the bias resistors.
Patrick
BTW he has actually published a book with all his earlier SiC based circuits in 2014.
ISBN 9784416114315, and there is ONE used copy on offer at Amazon.
There you can see how he compares SiC to other MOSFETs (e.g. UHC-MOS), and what measures are necessary in their applications.
As mentioned earlier, it is not (cannot be) 1:1 substitute for the Semisouth.
You need to tailor your design to use them properly.
And do not expect distortion much below 0.1%.
Patrick
ISBN 9784416114315, and there is ONE used copy on offer at Amazon.
There you can see how he compares SiC to other MOSFETs (e.g. UHC-MOS), and what measures are necessary in their applications.
As mentioned earlier, it is not (cannot be) 1:1 substitute for the Semisouth.
You need to tailor your design to use them properly.
And do not expect distortion much below 0.1%.
Patrick