Not that new after a few months but at least i know it doesn't burst into flames. Listened briefly to a cascoded CCS in place of the choke loads and it sounded pretty good, maybe even better than the chokes. Unfortunately don't have enough DN2540s at the moment for further investigations. Listened also to a transformer loaded stage using the LL1660/18mA but ended up prefering the chokes and caps. PS is mostly "borrowed" from Salas although CCS is different. Tried to use RCRC filtering instead of the LC but it sounded clearly worse. Even with minimal filtering there is less than 1mV ripple in the regulated output. T1 and T2 are the same R-core transformer while T3 is a separate EI. Separating brought a marked improvement. Currently the PS is shared between two channels, the tubes run at around 21mA and 20mA goes through the shunt. Heatsinks get really hot. Interestingly, if the voltage drop across the CCS is reduced from the current nearly 100v to about 70v the heat situation improves a lot but for some reason the sound seems to suffer. I have no explanation for this and will investigate further. Quick measurements showed exemplary square wave response and frequency range of at least 100kHz. Will measure distortion soon, mostly out of idle curiosity.
This is clearly the best sounding preamp i have ever built or listened to. It can probably be improved by using separate power supplies or at least regulators for each channel but my current case won't be able to accomodate.
I suspect the most important part of this design is the PS and especially the shunt reg. My most sincere thanks to Salas for making his design public.
Always interested in suggestions and improvements.
This is clearly the best sounding preamp i have ever built or listened to. It can probably be improved by using separate power supplies or at least regulators for each channel but my current case won't be able to accomodate.
I suspect the most important part of this design is the PS and especially the shunt reg. My most sincere thanks to Salas for making his design public.
Always interested in suggestions and improvements.
I just finished a simple 6N1P pre with a SV83 pentode as a CCS and never heard anything so magic.
The CCS transformed the amp. Everything about it, the bass was almost too much, the treble, voices were jumping out of my Klipsh's.
Using 1 yellow LED as bias at around 1.9 V.
As you said SY before there is a great synergy between the two.
But here there are 3 red's setting a bias at 6V.
Isn't this excessive for a pre tube?
Or am I setting my bias incorrectly?
T1 and T2 are on the same R-core. T3 is an ex wall brick which lost its original secondary and got a really thick new one for the heater voltage. The R-core has the required secondaries but the sound was much better with a separate transformer. It took me an entire evening to rewind it without disassembly so i was pretty strongly motivated.
It has been playing for a few months. I am still tempted to try out a CCS against the chokes. It will allow for a better wiring layout as well (the current mess is temporary).
It has been playing for a few months. I am still tempted to try out a CCS against the chokes. It will allow for a better wiring layout as well (the current mess is temporary).
Attachments
Thanks but allergic to both cream and CFs 
I have short cables to drive into a tube power amp. Output impedance in the region of 1k suits me just fine. Although the bass is often better through a CF, the loss of transparency elsewhere is just not acceptable to my ears. I i really wanted low output impedance i would try a nice amorphous parafeed step-down. Quite possibly as Mr Davenport has done it.
I have short cables to drive into a tube power amp. Output impedance in the region of 1k suits me just fine. Although the bass is often better through a CF, the loss of transparency elsewhere is just not acceptable to my ears. I i really wanted low output impedance i would try a nice amorphous parafeed step-down. Quite possibly as Mr Davenport has done it.
CCS and LED bias; that could almost be one of my designs...
You'll find that, if you do your grounding right and use a good CCS, the power supply becomes nearly irrelevant.
I have a 5842 pre amp (into Lundahl LL1630) with 1 green led under the cathode (2V). CCS fed with 0A2.
I tried a 1 uF good quality film cap parallel with the led and the sound inproved a lot. I don't know if this makes sense and what the best cap value is but it is easy to try this with the 6H30 preamp.
I haven't tried this. I will give this a go when I get back to my house.
I've seen many discussions on using bypass caps for LED's.
But sometimes you think "how can it get much better?" Then you make a small adjustment and sit there listening in amazement. Approaches on the border of obsession! Lol
But the SLCF doesn't share the faults of the CF, only its advantages.
You should lash one up one day just to prove me wrong
Shoog
Proving you wrong was still not on the list of my New Year resolutions; thanks for bringing it up
I have done this previously but didn't use a particularly sophisticated CCS - just a 2SK170. The 6H30 demands a minimum of 20mA to sound acceptable (ARC runs them at 5mA!) in common cathode and probably even more as a follower. My current PS simply cannot cope with this so it's not really a question of lashing one up tonight. I also prefer not to involve a negative supply rail due to the additional complications. Any idea of what is the optimal voltage across a cascoded CCS? My impressions so far are that 100v seems to work (sound) better than 70v which is a far cry from the 10-20v i have heard mentioned and would make a negative rail essential.
No graphs, but I spot-measured the impedance of some cheap red LEDs at 1k and 100kHz and saw no difference. Indeed, the cap values ARE terrifying before they have any effect! And of course, those caps will have high impedance at high frequencies...
I don't have reliable noise measurements yet but, following a suggestion from Pete Millett, I hope to have some soon. Anecdotally, I use LEDs in the cathodes of my phono amp (effectively at the input) and the noise level seems to be below the thermal noise of the cartridge.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.