Comparison of the VSSA with a modified one (dumped Diamond Input and servo):
-> www.esperado.fr - VSSA with Diamond input and DC servo
-> www.esperado.fr - VSSA with Diamond input and DC servo
😎
I always knew the diamond was not "bad" 😀.
Your results are close to my experiences.
In fact , I have to "trick" my VSSA design (no input cap , higher UG point)
to get 350V+ slew.
Out of the box , my NAD-BH gives 100V in .3uS (340V/uS).
Servo hooked ... running the EF3 (below).
I did get 700V , but this design is a compromise to get a cooler running,
better clipping IPS while retaining a CFA "heart" 🙂 .
PS - I'm testing new OPS's ... not all can relay (from IPS) these high slew rates to output. This has a lot to do with the end result (slew wise).
OS
Attachments
Last edited:
NEW -- OPS ... VFET/LFET Class A or AB !!
I almost have it figured.
I've already simulated "Moskido" Class A (Below 1).
A. Crude current feedback error correction.
B. is not "self settling/adjusting" , needs a Vbe for thermal comp.
C. It does error correct ... but not to the degree of (below attachment 2 +.asc).
Attachment 2 - - Voltage FB from the IRF's source. Voltage across the
output source resistors collapses the driver stage .... equilibrium
without a Vbe , this seems to work for bjt/vfet/and Lfet.
Non-linearities are also fed back.
(below 3) is the "jewel" , made for MOSFETS ... built/tested.
All 3 are copyrighted ... I must add.
Therefore , I will take the best attributes of each to create
a universal "SLEWFET" OPS.
1. must be fast 2. must be either class A or AB switchable.
3. foolproof thermal stability.
PS .. EF3's are great , but some want FETS and/or A
(no X-over distortion). This might be a universal solution with great results.
OS
I almost have it figured.
I've already simulated "Moskido" Class A (Below 1).
A. Crude current feedback error correction.
B. is not "self settling/adjusting" , needs a Vbe for thermal comp.
C. It does error correct ... but not to the degree of (below attachment 2 +.asc).
Attachment 2 - - Voltage FB from the IRF's source. Voltage across the
output source resistors collapses the driver stage .... equilibrium
without a Vbe , this seems to work for bjt/vfet/and Lfet.
Non-linearities are also fed back.
(below 3) is the "jewel" , made for MOSFETS ... built/tested.
All 3 are copyrighted ... I must add.
Therefore , I will take the best attributes of each to create
a universal "SLEWFET" OPS.
1. must be fast 2. must be either class A or AB switchable.
3. foolproof thermal stability.
PS .. EF3's are great , but some want FETS and/or A
(no X-over distortion). This might be a universal solution with great results.
OS
Attachments
It would be great if we would have option to use Toshiba Audio mosfets like
2SK1530, 2SJ201.😀
Several guys who used IRFP and Toshiba in a identical circuit clamed the Toshiba has superior sound to the IRFP or even to the modern ON semi BJT.
I do believe it would be interest on the Toshiba circuit or PC board to
My humble opinion.🙂
Greetings
2SK1530, 2SJ201.😀
Several guys who used IRFP and Toshiba in a identical circuit clamed the Toshiba has superior sound to the IRFP or even to the modern ON semi BJT.
I do believe it would be interest on the Toshiba circuit or PC board to
My humble opinion.🙂
Greetings
Very cool!
OS, just one thought... If you don't use "CFP-kind-of" topology with IRFPs (as MOSKIDO does), then what are those 0.1R resistors at the drains are for?
Cheers,
Valery
You can easily measure at the source (double 0.33R), although dividing by 0.165 is more complicated than dividing by 0.1 😛
It's a current limiter, but what for? In "CFP-like" topology they are used to align the currents between the pairs (if sources would be looking at the rails), but at the drain there will be no alignment...
They do not cost more than the good lateral mosfets.
Here I'm not talking about Chinese made copy 2SK1058 & 2SJ162- something more like Alfet, Exicon or other similar quality mosfet.
Unfortunately the so called Renesans or other fake 2SK1058 & 2SJ162 just the shadow quality wise to the original Japanese Hitachi.
I tested and that is MY OPINION.
May be you got different experience .
Id rather pay a bit more build something I know it has the potential.
Back to the topic.😀
I have been quite satisfied with the Renesas devices. I don't hear or see any issues with them at all. I didn't really think of them as fakes. They don't claim to be genuine Hitachi. They claim to be their own brand. They sound and measure great in my VSSA and even the little mini ops I just made. However, even the Renesas cost double what the IRFP devices cost.
Did you compared the Renesans with Hitachi 1 on 1
If not you don't really know what you are missing not to talk about other quality mosfet I mentioned.
Just go to the ESP mosfet amp and you can read more about Fake or copy mosfet.
I do not say Renesans is fake😕, I say many sellers sell under the name of Renesans fake mosfets unfortunately.
I do had the opportunity compare the original Renesans with the orig. Hitachi.
- Pleased do not forget I wrote that is my experience, may be you have better experience and I respect that.
I just mentioned the ESP because he had similar experience to me.
Let s go back to the topic, these was commented over and over already on the forum and other places.
I do not think OS will be happy if we start discussing here about components.
Greetings
If not you don't really know what you are missing not to talk about other quality mosfet I mentioned.
Just go to the ESP mosfet amp and you can read more about Fake or copy mosfet.
I do not say Renesans is fake😕, I say many sellers sell under the name of Renesans fake mosfets unfortunately.
I do had the opportunity compare the original Renesans with the orig. Hitachi.
- Pleased do not forget I wrote that is my experience, may be you have better experience and I respect that.
I just mentioned the ESP because he had similar experience to me.
Let s go back to the topic, these was commented over and over already on the forum and other places.
I do not think OS will be happy if we start discussing here about components.
Greetings
Last edited:
I'm not comparing with latfet, because 2SK1530, 2SJ201 are hexfet with low vgs. I'm comparing with irfp240/9240
Very cool!
OS, just one thought... If you don't use "CFP-kind-of" topology with IRFPs (as MOSKIDO does), then what are those 0.1R resistors at the drains are for?
Cheers,
Valery
Since the source(s) are all tied together and have a common resistor ,
how would you share current with multiple pairs ?
This will be a design consideration for a class switchable (A to AB) offering.
I'm trying for a V/LFET that will go from an AB biased state to a
"hard" 2A+ class A state with the flick of a switch.
If I make this a 3 pair contraption with the advanced HEC , I must
be able to monitor at least one pairs source R voltage drop as
feedback to set the class A current.
For an OPS , these are lofty goals (both Vfet/Lfet) and class switchable
operation - BUT ... i will produce.
OS
I sort of stuffed the Latfet OPS from post #4432 into the Spooky Amp just to see what would happen. I didn't do a real sanitary job but it runs.
N-Period=1
Fourier components of V(vin)
DC component:-3.79238e-011
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.000e+00 1.000e+00 -0.00° 0.00°
2 2.000e+03 1.473e-09 7.365e-10 -2.96° -2.96°
3 3.000e+03 2.208e-09 1.104e-09 -1.97° -1.97°
4 4.000e+03 2.943e-09 1.471e-09 -1.48° -1.48°
5 5.000e+03 3.678e-09 1.839e-09 -1.19° -1.19°
6 6.000e+03 4.413e-09 2.207e-09 -1.00° -1.00°
7 7.000e+03 5.148e-09 2.574e-09 -0.86° -0.86°
8 8.000e+03 5.882e-09 2.941e-09 -0.76° -0.76°
9 9.000e+03 6.617e-09 3.308e-09 -0.68° -0.68°
Total Harmonic Distortion: 0.000001%
N-Period=1
Fourier components of V(vout)
DC component:-0.109327
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 5.536e+01 1.000e+00 0.07° 0.00°
2 2.000e+03 9.334e-04 1.686e-05 103.41° 103.34°
3 3.000e+03 1.377e-03 2.487e-05 3.86° 3.79°
4 4.000e+03 5.998e-04 1.084e-05 -99.57° -99.63°
5 5.000e+03 5.272e-04 9.523e-06 2.86° 2.80°
6 6.000e+03 2.594e-04 4.686e-06 -104.43° -104.49°
7 7.000e+03 1.445e-04 2.611e-06 6.86° 6.80°
8 8.000e+03 1.777e-04 3.210e-06 -105.23° -105.29°
9 9.000e+03 8.764e-06 1.583e-07 39.19° 39.13°
N-Period=1
Fourier components of V(vin)
DC component:-3.79238e-011
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.000e+00 1.000e+00 -0.00° 0.00°
2 2.000e+03 1.473e-09 7.365e-10 -2.96° -2.96°
3 3.000e+03 2.208e-09 1.104e-09 -1.97° -1.97°
4 4.000e+03 2.943e-09 1.471e-09 -1.48° -1.48°
5 5.000e+03 3.678e-09 1.839e-09 -1.19° -1.19°
6 6.000e+03 4.413e-09 2.207e-09 -1.00° -1.00°
7 7.000e+03 5.148e-09 2.574e-09 -0.86° -0.86°
8 8.000e+03 5.882e-09 2.941e-09 -0.76° -0.76°
9 9.000e+03 6.617e-09 3.308e-09 -0.68° -0.68°
Total Harmonic Distortion: 0.000001%
N-Period=1
Fourier components of V(vout)
DC component:-0.109327
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 5.536e+01 1.000e+00 0.07° 0.00°
2 2.000e+03 9.334e-04 1.686e-05 103.41° 103.34°
3 3.000e+03 1.377e-03 2.487e-05 3.86° 3.79°
4 4.000e+03 5.998e-04 1.084e-05 -99.57° -99.63°
5 5.000e+03 5.272e-04 9.523e-06 2.86° 2.80°
6 6.000e+03 2.594e-04 4.686e-06 -104.43° -104.49°
7 7.000e+03 1.445e-04 2.611e-06 6.86° 6.80°
8 8.000e+03 1.777e-04 3.210e-06 -105.23° -105.29°
9 9.000e+03 8.764e-06 1.583e-07 39.19° 39.13°
Renesas took over the production from Hitachi, the Fets are made on the same machines, that used to make the Hitachi's, they stopped production after the Japan earthquake.
I think that Samelab/allfet and Exicon are the same Fets, just under a different Brands. ALL-fet builds an interesting fellow an N and P channel in the same housing.
I have made many amplifiers using the Exicon-Fets, they are excellent.
I think that Samelab/allfet and Exicon are the same Fets, just under a different Brands. ALL-fet builds an interesting fellow an N and P channel in the same housing.
I have made many amplifiers using the Exicon-Fets, they are excellent.
I see what you mean, but... easier to imagine with a normal EF3. Let say, you've got a 5-pair SlewMonster. Emitter resistors allow aligning the bias (and thus sharing the current) between the pairs. But what's gonna happen if you add some resistors to the collectors? Nothing positive 🙂 OK, collector "V" will be slightly different because of the difference in "I", but "I" distribution will not get any better.
Same thing here... 😉
Last edited: