Luminaria?

ra7

Member
Joined 2009
Paid Member
I just finished building a #26 and 01a preamp. I want to have a go at this project next. Couple of questions:

1. What is an easy way to get a negative adjustable bias on the grid/gate? Hoping someone has already done that and figured out a way. I was thinking some sort of adjustable DC supply from the heater windings of an Antek 120V trafo.
2. Can the negative bias be supplied from a 9V battery? How much current is drawn?
3. Has anyone experimented with a choke or CCS load yet?

We really need some better devices for audio. 100-year old tubes and unobtanium VFETs cannot be the solution.
 
Last edited:
Member
Joined 2012
Paid Member
I used to run a 26 preamp, but that was replaced with the Luminaria. No more tube noise, no more tubes wearing out.:)

My Luminaria bias supply is a simple basic LM337 regulated supply followed by a pot for bias voltage adjustment. The current draw is very low, less than 1mA. A low voltage transformer provides the power. The regulated voltage then feeds to a 2K resistor and 10K pot in series for bias voltage adjustment. I used a similar bias supply in my 2SK180 SE amplifier. 25W Single Ended Hammond 193V Choke Loaded 2SK180 L'Amp

Edit: I just check my Luminaria and Vbias=-16.5V. My bias supply is as described above except the transformer is 30VAC so the VDC is probably around 30V or so. I used what transformer I had. A bias supply of 20VDC probably is enough.

A battery would probably work for bias but I wouldn't risk roasting a VFET if the battery loses voltage.

I have not tried choke or CCS loading. A choke load is feasible. diyAudio member Pinholer used a Hammond 156C in the 2SK82 front end of his 2SK82-SSJ28 amplifier. 2SK82 -2SJ28 SE Amp

A CCS would require more V+ and also more heatsinking.
 
Last edited:
Member
Joined 2012
Paid Member
My window of time for editing expired. I mentioned that a bias supply of 20VDC is probably enough. I am taking that back. I think that it may be enough or may not be enough. So 30VDC that I have is on the safer side.

The high bias voltage probably makes a battery not practical.
 
Last edited:
Member
Joined 2010
Paid Member
I used a 24VAC transformer which, lightly loaded, produces over 36VDC. For fixed bias (no source resistor), Vgs is approximately -20V. I used a TL431 regulator in an adjustable power supply to provide a range of about -14V to -26V.

I have used the Hammond 156C as a choke load for the 2SK82 as a voltage amp stage in my 2SK82-2SJ28 amp. In this case, I used a different operating point than the LuminAria since the Hammond choke is only good for about 8mA. Also, it is self-biased with source resistors, instead of a negative bias supply. It works quite well and is capable of large voltage swing with low distortion.
 
Member
Joined 2012
Paid Member
Here are the modifications that I made to the circuit to accommodate fixed bias.

Re: CCS - I mentioned more V+ but that is the case only if compared to choke loading. The resistor loading already drops quite a bit of voltage so the voltage drop may be similar for CCS.
 

Attachments

  • Luminaria - fixed bias.png
    Luminaria - fixed bias.png
    137.9 KB · Views: 383
With it the 2sk82 so hard to find I found a Russian 2P926B might be a good substitute. It does however need a lot less -bias voltage. I did a very quick check just using a source resistor and a variable power supply on the couple I bought from this seller and they did seem ok and shiping took 60 days. I did find a cheaper source on the internet but my order from them never arrived so I had to let Paypal step in to get me a refund but the seller I am listing seems to be a good source.

Transistor silicon 2P926B (KP926B) USSR Lot of 1 pcs | eBay
 
Member
Joined 2012
Paid Member
2SJ28 Luminaria

I have had my Luminaria now for over three years and I am still enjoying it immensely.

I know 2SK82s are hard to find now but perhaps people still have 2SJ28s. So I was playing with LTSpice and modified the Luminaria circuit to accommodate the 2SJ28. It worked in the simulations, so I expect it to work in real life. Comment and critique if you see issues - I'm not an expert. :)

I hope that it will help others to enjoy the sounds of a Luminaria preamp, and also get some 2SJ28s out of Greedy Boyz's stashes. :)

One plus of the 2SJ28 is that for fixed bias, a separate bias supply is not required when overall power is provided by a V+ supply.

It would be a great preamp for the latest DIY Sony VFET Part 1 follower amp too!
 

Attachments

  • Luminaria 2SJ28 Source bias.png
    Luminaria 2SJ28 Source bias.png
    88.9 KB · Views: 280
  • Luminaria 2SJ28 fixed bias.png
    Luminaria 2SJ28 fixed bias.png
    94.9 KB · Views: 258
I was about to mention the $25 KP926b as an attractive substitute but noticed the ebay seller who had them has upped his price from $25 to $49 each. What Nelson needs to do is give us a nice DIY time machine to go back and scoop these parts up when they were cheep !

I bought a handfull at $25 over a year ago. You can still scoop up LU1014D and LD1014D at very reasonable prices.
 
Last edited:
I have had my Luminaria now for over three years and I am still enjoying it immensely.

I know 2SK82s are hard to find now but perhaps people still have 2SJ28s. So I was playing with LTSpice and modified the Luminaria circuit to accommodate the 2SJ28. It worked in the simulations, so I expect it to work in real life. Comment and critique if you see issues - I'm not an expert. :)

I hope that it will help others to enjoy the sounds of a Luminaria preamp, and also get some 2SJ28s out of Greedy Boyz's stashes. :)

One plus of the 2SJ28 is that for fixed bias, a separate bias supply is not required when overall power is provided by a V+ supply.

It would be a great preamp for the latest DIY Sony VFET Part 1 follower amp too!

Dear Ben,do we have a chance to use the -145V power supply for this 2SJ28.I have 4pces of this VFET.
I plan to use the sit follower as output

Thanks.