I built the newest B1 R2 buffer preamp from the schematic Mr Pass graciously recently posted. Easy to built on a perfboard in a few hours. I am using a Meanwell 24V switcher for now supplying the voltge. To me it appears to have some gain. Wonderful sounding with just an hour or two of listening with it.
Here is where I first heard about the new version. Post #17.
http://www.diyaudio.com/forums/pass-labs/295398-choosing-best-output-coupling-capacitors-b1-2.html
Here is where I first heard about the new version. Post #17.
http://www.diyaudio.com/forums/pass-labs/295398-choosing-best-output-coupling-capacitors-b1-2.html
Attachments
Last edited:
Still hundreds of original Toshiba 2SK170BL in stock that will appear in Swap Meet in a few weeks. I am still in doubt to sell them all at once or to match them.
Curious thing is I need a few 2SK170GR and some 2SJ74BL myself
Please don't send me PM about 2SK170BL but keep an eye on Swap Meet.
Curious thing is I need a few 2SK170GR and some 2SJ74BL myself
Please don't send me PM about 2SK170BL but keep an eye on Swap Meet.
You have a choice.
a.) sell them as unmeasured.
b.) Measure Idss and group them by selected Idss
c.) spot check Id at various Vgs and match similar "curves" to make matched pairs.
d.) Curve trace using automated equipment and match the curves.
For a B1 and a DCB1 you only need method b.
It is perfectly adequate for the CCS and the Follower to be selected for same Idss, since that is the current they operate at, ±the signal current.
I batched using method b. and then used method c. to find matched pairs.
For the 600 jFETs I had bought, stage b. was completed in a couple of afternoons.
Stage c. took about ten times longer, spread over about 2weeks.
If you value your time @ £0.00 per hour then you can provide a service to the community.
If you value your time @ £10 per hour then method c. comes out at ~ £3 per pair of jFETs.
After I had matched them up, nearly 300pairs, I found no matched triples and certainly no matched quads.
See my swap meet post and you'll see I serviced the community. Some day I will put a higher value on my time.
a.) sell them as unmeasured.
b.) Measure Idss and group them by selected Idss
c.) spot check Id at various Vgs and match similar "curves" to make matched pairs.
d.) Curve trace using automated equipment and match the curves.
For a B1 and a DCB1 you only need method b.
It is perfectly adequate for the CCS and the Follower to be selected for same Idss, since that is the current they operate at, ±the signal current.
I batched using method b. and then used method c. to find matched pairs.
For the 600 jFETs I had bought, stage b. was completed in a couple of afternoons.
Stage c. took about ten times longer, spread over about 2weeks.
If you value your time @ £0.00 per hour then you can provide a service to the community.
If you value your time @ £10 per hour then method c. comes out at ~ £3 per pair of jFETs.
After I had matched them up, nearly 300pairs, I found no matched triples and certainly no matched quads.
See my swap meet post and you'll see I serviced the community. Some day I will put a higher value on my time.
Last edited:
I bought them a few years ago. I am not sure of the source, sorry. I stockpiled a few sets but as any obsolete part I was skeptical of being originals. The ones I used were Toshiba but I would buy the LSK's now since the 74 has become available. This set seems to sound great so I am satisfied that they are not fake, if so very good fakes. I have a number of buffers and pres but I found this schematic and had the j-fets hanging around and decided to build something. The last few FW amps I built I bought and used LSK's to be sure of original good quality j-fets. I still have a few questionable Toshiba J-fets I may use to build something in the future.
I thought I would add I have bought the 2SK170 and 2SK74 from B&D Enterprises in the past and there reputation for selling original parts is excellent. If you want Toshiba they would be a good source to try. If you do not want to gamble buy LSK from our store.
Last edited:
I am inexperienced with electronics, busy with my B1 Build. I have two regulated PSU units, one supplying 15V and the other 24V. The buffer works well with both. I know 15V is below the recommended 18V spec and the 24V is at the high end of the spec. My caps are rated to 35V.
Which PSU will be best to use, 15V or 24V?
I am not sure how, if at all, voltage affects the sound?
Which PSU will be best to use, 15V or 24V?
I am not sure how, if at all, voltage affects the sound?
the spec for the B1 is 18Vdc. It will work with 15Vdc.
24Vdc is only 4V above the DCB1 PSU, so I'm inclined to say try it.
the maximum output voltage of these single ended Followers is ~ 1/3rd of the supply voltage, i.e. ~6Vac for 18Vdc when the load impedance is high.
But as a single ended follower approaches that maximum the distortion increases a lot. And gets worse as the load impedance drops.
The easy rule of thumb is:
Limit the maximum signal through the follower to <50% of the just starting to clip threshold into the load impedance normally used.
That effectively sets the B1 normal maximum output voltage to ~2.5Vac to 3Vac
Most digital sources have a maximum output <2.5Vac.
i.e. there is no effect on the sound when transients exceeding the average -20dB level need to be processed.
I think this is why many here approve the B1 and DCB1 as a universal Buffer.
24Vdc is only 4V above the DCB1 PSU, so I'm inclined to say try it.
the maximum output voltage of these single ended Followers is ~ 1/3rd of the supply voltage, i.e. ~6Vac for 18Vdc when the load impedance is high.
But as a single ended follower approaches that maximum the distortion increases a lot. And gets worse as the load impedance drops.
The easy rule of thumb is:
Limit the maximum signal through the follower to <50% of the just starting to clip threshold into the load impedance normally used.
That effectively sets the B1 normal maximum output voltage to ~2.5Vac to 3Vac
Most digital sources have a maximum output <2.5Vac.
i.e. there is no effect on the sound when transients exceeding the average -20dB level need to be processed.
I think this is why many here approve the B1 and DCB1 as a universal Buffer.