ikoflexer said:
Thanks. This phono has been done with tube logic and comparison, right from the start. It will not let down some tube oriented guy.
I don't think that the Linear Systems LSK170B will let you down either because ''Jonners'' from England used them for our phono already with no problems. Position wise matching (left & right per stage) is the best strategy indeed. The strongest for Idss pair, better be reserved for 1st stage. Bypass your batteries with a 10.000uf 35V capacitor. Batteries are not that good although there is an urban legend that they are special. Good thing is that they are virtually floating so they don't give such a hard time with not so good grounding, and don't emit hum fields. When you are about to make a PSU, don't reinvent the wheel, go directly for the shunt. We will be happy to learn about your progress and help out.
Re: row psu
Someone must plan why and what values, knowing the particular raw PSU details. Such small resistors are probably there to cut initial rough edges on the charge cycle that translate into unwanted harmonics. They certainly can not create problems if you use them. If the capacitors are very good like Nichicon KG, and not that large in value, maybe doubling is not too efficient in cutting much self inductance and esr, because it may be low enough in the first place. Doubling can be very anti economic when using expensive capacitors...
nicoch46 said:
Someone must plan why and what values, knowing the particular raw PSU details. Such small resistors are probably there to cut initial rough edges on the charge cycle that translate into unwanted harmonics. They certainly can not create problems if you use them. If the capacitors are very good like Nichicon KG, and not that large in value, maybe doubling is not too efficient in cutting much self inductance and esr, because it may be low enough in the first place. Doubling can be very anti economic when using expensive capacitors...
Honestly, I have no idea who makes the jfets sold by MCM. In time I'll get some Toshiba originals, thank you guys for all the pointers.
We all know simulation isn't reality, but it's fun. I've attached a simulation file any one can run on the freely downloadable ltspice. I've included the regulator in the simulation, with a 1V p-p ripple, DC bias 30V going into the regulator, and out coming an impressing almost zero ripple at 24.7V. Also nice, a bit of distortion analysis with .027% THD, of which 99.91% is second harmonic.
Now, can anyone see why I'm a bit peeved that the jfets are on back order and I have to wait another two weeks to get them?
Actually it seems I can't attach this type of file, so it can be found here
Right click on the link and then save the file to a place where you can open it with ltspice if interested in playing with the schematic.
We all know simulation isn't reality, but it's fun. I've attached a simulation file any one can run on the freely downloadable ltspice. I've included the regulator in the simulation, with a 1V p-p ripple, DC bias 30V going into the regulator, and out coming an impressing almost zero ripple at 24.7V. Also nice, a bit of distortion analysis with .027% THD, of which 99.91% is second harmonic.
Now, can anyone see why I'm a bit peeved that the jfets are on back order and I have to wait another two weeks to get them?
Actually it seems I can't attach this type of file, so it can be found here
Right click on the link and then save the file to a place where you can open it with ltspice if interested in playing with the schematic.
get yr 2SK170's on ebay
Just do a search on ebay- that's where I've bought my 2SK170BL's. I get the 'matched' quad sets from Audiotek; usually around $4-5 including shipping. The 2 sets Ive bought were fine.
Idss, set 1-
8.18 8.13 8.09 8.00mA
Idss, Set 2-
7.72 7.40 7.25 7.09mA
Matched? Good enough for me. The first set was better. I put trimmers for my source R's to bias them all for ~8.2v on the drains. Coming from tubes, these cost peanuts to me!
-Kent
Just do a search on ebay- that's where I've bought my 2SK170BL's. I get the 'matched' quad sets from Audiotek; usually around $4-5 including shipping. The 2 sets Ive bought were fine.
Idss, set 1-
8.18 8.13 8.09 8.00mA
Idss, Set 2-
7.72 7.40 7.25 7.09mA
Matched? Good enough for me. The first set was better. I put trimmers for my source R's to bias them all for ~8.2v on the drains. Coming from tubes, these cost peanuts to me!
-Kent
ikoflexer:
Thanks for your effort. Our phono and shunt have been developed along 3 axis. Simulating, comparative listening, measuring. Everything that I have posted has been simulated on Circuit Maker 2000 and your results in LT Spice agree. Very nice for no loop feedback and a couple of FETS.
I have also made some real measurements versus a $10K phono reviewed as amongst world's top 3 phono stages. Actually I have listened to our RIAA versus some extremely expensive dream machines, driven by exotic carts and TTs at the same time in the same systems. I won't discuss brands or my subjective opinion. I will only say, be confident about our RIAA & PSU. Also a macho 2 box phono & double mono PSU build, painstakingly tuned with money no object parts is something I owe to myself. Will do at a point when I will have a better cart that can do it justice. It will cost some money, but being simple it will not skyrocket.
Black line is non cascoded, non buffered Salas. Blue line is $$$ one.
The 100Hz usual ripple peak is absent due to my shunt PSU on the black line. The commercial phono uses a solid state/tube double mono shunt hybrid PSU and it is fully symmetrical audio circuit wise.
Expect 6-10dB better 50Hz hum with the unshielded toroidal transformer out of my tiny RIAA box, and more careful loop measurement. Also my USB card has about -100dB noise floor between 10kHz-20kHz and 0.007% THD residual.
Distortion & noise:
Thanks for your effort. Our phono and shunt have been developed along 3 axis. Simulating, comparative listening, measuring. Everything that I have posted has been simulated on Circuit Maker 2000 and your results in LT Spice agree. Very nice for no loop feedback and a couple of FETS.
I have also made some real measurements versus a $10K phono reviewed as amongst world's top 3 phono stages. Actually I have listened to our RIAA versus some extremely expensive dream machines, driven by exotic carts and TTs at the same time in the same systems. I won't discuss brands or my subjective opinion. I will only say, be confident about our RIAA & PSU. Also a macho 2 box phono & double mono PSU build, painstakingly tuned with money no object parts is something I owe to myself. Will do at a point when I will have a better cart that can do it justice. It will cost some money, but being simple it will not skyrocket.Black line is non cascoded, non buffered Salas. Blue line is $$$ one.
The 100Hz usual ripple peak is absent due to my shunt PSU on the black line. The commercial phono uses a solid state/tube double mono shunt hybrid PSU and it is fully symmetrical audio circuit wise.
Expect 6-10dB better 50Hz hum with the unshielded toroidal transformer out of my tiny RIAA box, and more careful loop measurement. Also my USB card has about -100dB noise floor between 10kHz-20kHz and 0.007% THD residual.
Distortion & noise:
Attachments
Anyone needing 2sk's can have a look here:2sk170bl sale
I bought 100, and it's the real deal, Spencer is a stand up guy.
Best regards
Ebbe
I bought 100, and it's the real deal, Spencer is a stand up guy.
Best regards
Ebbe
Hi! I've been well. Still listening. I'm waiting on some parts to upgrade. As it is running into the power amp directly and no lack of gain (!!), I thought I'd try to take thr r-load back up to original 1K ohm. The result was an improvement on imaging. What I would like to do is put a pot there and find the optimum R value.
Another change was cables from the VPI to the RIAA. Originally, I was using some old Silver Sonic cable I cut up (I try to minimize RCA connectors where I can). Replaced it with two 24ga. 2-strand untwisted silver wire connectors I made up. Very susceptible to hum, I had to giggle them around to find a quiet 'zone' on my counter top. But I did I notice more 'fluidity' in the music; and I am not one to take in the audiophoolery that some cable companies benefit from. It could be that I am just using better connectors.
What is funny is I twisted them for kicks (to reduce some of the hum) and the fluidity went away....
I'm thinking of re-building it, too. My old-school breadboard layout looks cool, but a lot of lead wire is there; by making more compact, might it improve things?
Finally, I need to do some mods to my TP Buffalo so I can do an AB test with lp's to check for levels.
-Kent
Another change was cables from the VPI to the RIAA. Originally, I was using some old Silver Sonic cable I cut up (I try to minimize RCA connectors where I can). Replaced it with two 24ga. 2-strand untwisted silver wire connectors I made up. Very susceptible to hum, I had to giggle them around to find a quiet 'zone' on my counter top. But I did I notice more 'fluidity' in the music; and I am not one to take in the audiophoolery that some cable companies benefit from. It could be that I am just using better connectors.
What is funny is I twisted them for kicks (to reduce some of the hum) and the fluidity went away....
I'm thinking of re-building it, too. My old-school breadboard layout looks cool, but a lot of lead wire is there; by making more compact, might it improve things?
Finally, I need to do some mods to my TP Buffalo so I can do an AB test with lp's to check for levels.
-Kent
Hi Kent!
Whatever you do with cables or parts or Rload or VTA is going to be audible for better or worse. More compact layout is never bad as well. My suggestion is to always use coaxial shielded cable especially for such a low level input and so much gain. What I have found is that coax sounds more natural when it is thin in diameter, 2.5mm or under. See about Mogami. It has many thin ones. Better tune your tone with other parts like your caps and resistors but make it hum free.
Whatever you do with cables or parts or Rload or VTA is going to be audible for better or worse. More compact layout is never bad as well. My suggestion is to always use coaxial shielded cable especially for such a low level input and so much gain. What I have found is that coax sounds more natural when it is thin in diameter, 2.5mm or under. See about Mogami. It has many thin ones. Better tune your tone with other parts like your caps and resistors but make it hum free.
salas said:Thanks for your effort. Our phono and shunt have been developed along 3 axis. Simulating, comparative listening, measuring. Everything that I have posted has been simulated on Circuit Maker 2000 and your results in LT Spice agree. Very nice for no loop feedback and a couple of FETS.
Oh no, I thank you! Cool stuff. Something is a bit unclear though, you mention the non-cascoded, two jfet per channel version. Is that the version you recommend building? Sorry, but with so many changes to the schematic over the life of the thread it's a bit confusing which one is the one that you think is best to build.
For MM I recommend what you simulated. Both the gain and Miller capacitance are right. What you have seen measured in replies #611 & #612, is what Kent has for 0.5mV MC. He added a buffer later. That is non cascoded input, since MC does not interact with the high 2SK Miller capacitance, due to very low self inductance unlike MM. You will have even better signal to noise ratio with your MM and lower gain. Keep the buffer, it has been proved transparent and matches better in most systems. Both Jonners and Kent found better matching with what they drive, and could not detect it sonically. Over the life of the thread, it was not about changes, but about specific versions for specific applications. Its just live, we do it together. DIY, not strictly engineering. Subjective comments and tunings about parts, cabling, TTs, carts, etc, and ''sounds good'' or ''bad'' are also welcome and seriously evaluated in this thread.
salas said:
Well today waiting the pcb(row hole) for shunt I mods a little the psu phono-replace byv27-200 with stealth II
-put 0.1uf X2 in the secondy of toroid
-delete fuse on dc rail ( do not do!at last thermal brake on ac)
-replace power cable (computer) with old supra speaker cable
I still evaluate but for sure I don't lost time