The 2SK170 data sheet says:
Low noise: En = 0.95 nV/Hz1/2 (typ.)
(VDS = 10 V, ID = 1 mA, f = 1 kHz)
I am using these FEts in my preamp and it did not add any noise....
WEIRD!!!
😱
Low noise: En = 0.95 nV/Hz1/2 (typ.)
(VDS = 10 V, ID = 1 mA, f = 1 kHz)
I am using these FEts in my preamp and it did not add any noise....
WEIRD!!!
😱
Fortunately the ear is not particular sensitive to noise in the lower reaches. I build quite a lot of phonostages with 2SK170, 2SJ74 that are dead quiet subjectively.
Here is a measurement of a phonostage with 4 plus 4 2SK170, 2SJ74. The red track is my Lyra Olympos producing 0.3mV at 5cm/sec. The noise of the phonostage is comfortably lower then the noise it makes "when the stone hits the plastic (Salas)".
The circuit could be used for MM as well....capacitance plays a role...yet what you loose in one end you kind of gain in the other..when you switch in a lover impedance RIAA network..
For sure the low capacitance fets are to be preferred for MM usage...will run a test to see just how much the bandwidth is limited with 47 Kohm input impedance...
then I will layout a small test board...🙂
For sure the low capacitance fets are to be preferred for MM usage...will run a test to see just how much the bandwidth is limited with 47 Kohm input impedance...
then I will layout a small test board...🙂
Hi,
Wrong way.
Add source degeneration.
That cuts capacitance, gain and nonlinearity equally.
Of course, noise is up, but so is the signal level, so SNR is retained.
Ciao T
Wrong way.
Add source degeneration.
That cuts capacitance, gain and nonlinearity equally.
Of course, noise is up, but so is the signal level, so SNR is retained.
Ciao T
By using 3x60mS and lower GR noise fets than the 22mS 4xK170/370 on low MC, I have seen the same deflation trend under 500Hz that Scott measured on them individually. That one FFT I showed in this thread when discussing with Thorsten about the single box Tx harmonic noise susceptibility, reads -112dB at 100Hz relative to 1VRMS=0dBV. Rs=5R. Its with just a V1.2R shunt reg common to all stages and channels, don't know if it could go a bit lower with extra local Cmult or other filters. The 8+8 BJT Accuphase reads -115dB there if I remember correctly on a German magazine FFT you showed. Why I make this preamble? To report on what less noise in the mid to bass regions in the electronics may reflect sonically when striving much more than enough. What I received was less grunge and more distinction in bass players and bassy male vocalists, or in big wind instruments, and more easy to follow rhythm sections. That is not about the LP surface noise and its distance to the electronics noise floor I would guess, maybe more about the dynamic distinction that a quiter in the nether regions pre can muster.
Thorsten.. thanks....
But also a hell of a place to have a switch....My thought was to have the MM Riaa as a parallel circuit you then switch in when running higher signal levels....
for my personalized use only MC is needed...
Then there is the grand implementation question...how to make a good impedance matching adjustment...would really prefer to have a continuous possibility and to avoid any mechanical switching elements. Bulk metal foil trimmers spring to mind.. but a hassle to work with...and not easy to adjust without a screwdriver....and also not easy to know where you are without a meter.
hmmm.. requires some thinking I guess...
But also a hell of a place to have a switch....My thought was to have the MM Riaa as a parallel circuit you then switch in when running higher signal levels....
for my personalized use only MC is needed...
Then there is the grand implementation question...how to make a good impedance matching adjustment...would really prefer to have a continuous possibility and to avoid any mechanical switching elements. Bulk metal foil trimmers spring to mind.. but a hassle to work with...and not easy to adjust without a screwdriver....and also not easy to know where you are without a meter.
hmmm.. requires some thinking I guess...
Last edited:
If not making a commercial product the trimmers are OK. Now if you have Medvedev as a potential customer in mind (already owns a Vitus) you gotta make it showing Riaa curves on a 21inch TFT, changing lines via remore control.😀 Hey, you win a ''full package holiday'' in Sochi non the less.🙂
If you want lowest 1/f noise take bipolars ( the Accuphase has them in the MC section ) but that is not an issue here. The Vitus sadly has mediocre measurements so Medvedev may need a second stage when i also wants to listen to music and not only look at the values 🙂 MiiB, a non inductive wirewound could do and Mentor makes very nice scales and 1:6 gears for readout.
If making something commercial.. I do believe in KISS.....Like my good friend does with the Lyra Connesieur products...It needs to able to do what its supposed to do...in a simple and understandable way
I would include a bulk foil trimmer saying you get all values, no switches, KISS etc. but recommend only a Fluke to trim. So I spice up with prestige. That gets me to Sochi along with Bono.😉
Or go all in/out with a motor driven trimmer and off course taco calibrated viewable on a display.... A solution for the Fancy...
hhmmm fun and games...🙂
hhmmm fun and games...🙂
Tried to sim with the JBF862.. I I loose app 4dB of gain compared to the SK369... But for sure it can be adjusted in the Riaa values..
Anyway nice to have a good low capacitance replacement for the Toshiba fets..
Anyway nice to have a good low capacitance replacement for the Toshiba fets..
Indeed. Holidays granted!😀 *Russia will be THE high-end goldmine my crystal ball says, focus you audio pros.😉
Attachments
Hi dimitri,
Could you please explane why do you think Thorsten is wrong...as I think he is right (local feedback will do all that, I would say as a non EE)
MiiB, maybe this is a strange idea but you could switch the current source to a lower value for MM. That way you can keep the RIAA values. I asume that requires a muting relay.
why not have specialized circuits optimized for MM and MC..... I mean we are already spending a fortune on special impedance matching devices....and making it workable from a few ohms to 47K without any switching seems like a bad idea....... then we can have two parallel circuits and just decide which one to give the supply Voltage...
I remember the original idea was to find a stage that does both. Seems to be difficult in good quality.
- Status
- Not open for further replies.