I believe the answer is yes: https://www.diyaudio.com/community/threads/diy-front-end-2022.394339/page-24#post-7284864
Thanks! I bought 100 of the Nichicon Muse bipolar 3.3uF caps so running out shouldn't be an issue.
Spreading out the legs of the transistors would make life easier during assembly. Good thought! Much apprecieated.Great Preamp on a Budget
This design by Nelson once again shows that he can design a great preamp that does not cost a great deal of money to build, and expensive parts are not needed for it to sound great.
To prove that point, no audiophile grade parts were used in this build. Electrolytic capacitors were standard types, rectifier diodes were 1N4007, the coupling caps were Panasonic polypropylene, but less than $2 each at Mouser. Resistors were no name metal film made in China that I bought from a local electronics store in packs of 10 for $0.70 a pack. Wire, screws, nuts, and standoffs were also sourced there, all made in China. Also purchased there was the made in China steel case with an aluminum front plate (thin) for only $30. Note that all of the prices are in Canadian dollars (CAN $1.00 = US $0.74 today).
I do have a resistor ladder volume control that has Holco resistors, but this was a left-over from over 20 years ago when I built it for a 26 tube preamp at that time.
I decided to go with a V+ power supply, aiming for 60V. The power supply is regulated by a zener shunt, followed by a mosfet voltage follower. I did the design and by keeping the board size to 100mm x 100mm, the JLCPCB price was $8.33, and that included shipping to my door, albeit slow shipping . It took about three weeks from ordering to receiving the pcbs, but at that price, I couldn't buy five pieces of copper clad perf board. So what a deal.
The 60V that I was aiming for ended up being 62V, and I was fine with that. The zeners I bought were specified at 1%, but they were high by 1.5%, and the Vgs of the mosfet I had was less than 4V, all adding up to a higher output voltage. I decided to go 60V+ instead of 50V for more voltage swing. My speakers are 103dB sensitive so I really don't need the higher output, but I wanted it in principle to better match my higher wattage THF-51S mu follower follower. The higher V+ should also reduce the distortion a bit too, although I wasn't looking to achieve that.
I am listening to the preamp right now, and it sounds great. It has clarity, but the sound is not sterile. As ra7 said in his post #594, the preamp is transparent.
In the next day or two, I will do some distortion measurements.
Thank you, Nelson, for this gift of a wonder design to diyers.
And everyone loves looking at pictures:
This design by Nelson once again shows that he can design a great preamp that does not cost a great deal of money to build, and expensive parts are not needed for it to sound great.
To prove that point, no audiophile grade parts were used in this build. Electrolytic capacitors were standard types, rectifier diodes were 1N4007, the coupling caps were Panasonic polypropylene, but less than $2 each at Mouser. Resistors were no name metal film made in China that I bought from a local electronics store in packs of 10 for $0.70 a pack. Wire, screws, nuts, and standoffs were also sourced there, all made in China. Also purchased there was the made in China steel case with an aluminum front plate (thin) for only $30. Note that all of the prices are in Canadian dollars (CAN $1.00 = US $0.74 today).
I do have a resistor ladder volume control that has Holco resistors, but this was a left-over from over 20 years ago when I built it for a 26 tube preamp at that time.
I decided to go with a V+ power supply, aiming for 60V. The power supply is regulated by a zener shunt, followed by a mosfet voltage follower. I did the design and by keeping the board size to 100mm x 100mm, the JLCPCB price was $8.33, and that included shipping to my door, albeit slow shipping . It took about three weeks from ordering to receiving the pcbs, but at that price, I couldn't buy five pieces of copper clad perf board. So what a deal.
The 60V that I was aiming for ended up being 62V, and I was fine with that. The zeners I bought were specified at 1%, but they were high by 1.5%, and the Vgs of the mosfet I had was less than 4V, all adding up to a higher output voltage. I decided to go 60V+ instead of 50V for more voltage swing. My speakers are 103dB sensitive so I really don't need the higher output, but I wanted it in principle to better match my higher wattage THF-51S mu follower follower. The higher V+ should also reduce the distortion a bit too, although I wasn't looking to achieve that.
I am listening to the preamp right now, and it sounds great. It has clarity, but the sound is not sterile. As ra7 said in his post #594, the preamp is transparent.
In the next day or two, I will do some distortion measurements.
Thank you, Nelson, for this gift of a wonder design to diyers.
And everyone loves looking at pictures:
Attachments
I think that to be so lucky and to be at the right time at the right place when someone offered a spare kit and snagged it (#596) is a talent! Never happened to me.
But I feel like it's coming....the wait will soon be over....then it will be spring in my audio world.
I am in the waiting list (email), so.... I am planning, if ever, to use this circuit (Bride of Son of Zen, BOSOZ) PSU for 40-50 V+ (BOSOZ PSU). Will this be good?
Thank you!
But I feel like it's coming....the wait will soon be over....then it will be spring in my audio world.
I am in the waiting list (email), so.... I am planning, if ever, to use this circuit (Bride of Son of Zen, BOSOZ) PSU for 40-50 V+ (BOSOZ PSU). Will this be good?
Thank you!
Just thinking out loud here... might be kind of nice to have a companion V+ or bipolar power supply project (maybe a high performance shunt?) to go along with the FE '22. Same form factor so they could be stacked. Having the regulation stage nice and close to the circuit being powered is normally recommended good practice. Just an idea..
Last edited:
Please feel free to contribute your circuit design and PCB manufacturing ("Gerber") files and parts list with Mouser part numbers. I'm sure plenty of diyAudio members would be very grateful to receive exactly the thing you describe.
Naw, I'm no accomplished and decorated EE. I'm nothing but a lowly repair tech that tries to fix (usually successful) old / worn out / abused locomotive electronics on a daily basis. And dabble with a little audio at home. Give me a PCB, some parts, and an iron... and I'll do my best human pick & place emulation.
At Patrick's request, I ran a quick experiment on the 2022 FE to investigate the feasibility of eliminating the capacitors C1, C2 and C3. We did this against a baseline build by observing the DC offset at R5 as everything warms up.
Conditions:
Next, I replaced R7 with a trimpot and nulled out the offset at steady state to within 5 mV. Now, the initial offset is around -300mV, moving to near 0 as things warm up, with about the same 300mV or so of thermal drift.
Conclusion, it doesn't look encouraging for eliminating the coupling capacitors C1, C2 and C3.
I think there is a possibility to eliminate C1 and C2 from the circuit (retaining C3) provided that:
Conditions:
- Built with the silicon and R7, R8 from the kit.
- C1 and C2 shorted.
- C3 not populated, offset measured at R5 to ground.
- The node at [-in] shorted to ground, as a single ended input.
- Everything else per Nelson's schematic except minor changes to R9 - R12, C4 and C5 to accommodate what I had on hand.
- All measurements done on a bench in free air.
- Supply at +/- 23 VDC
- Initially the offset measured at R5 is about 2.64 V, but moving upwards.
- It is 2.8 V within about a minute.
- After a couple of hours it is pretty steady at 2.91 V.
- V across R7 (332R): 1.46 V (4.4 mA)
- V across R6 (681R): 0,604 V (890 uA)
- V at gate of Q2: 0.265 V
Next, I replaced R7 with a trimpot and nulled out the offset at steady state to within 5 mV. Now, the initial offset is around -300mV, moving to near 0 as things warm up, with about the same 300mV or so of thermal drift.
Conclusion, it doesn't look encouraging for eliminating the coupling capacitors C1, C2 and C3.
I think there is a possibility to eliminate C1 and C2 from the circuit (retaining C3) provided that:
- You are willing to tweak R7 to get the offset into a reasonable range.
- You can live with the thermal drift.
Hi Greg,
The LuminAria was two preamps ago. When I switched my SIT power amps to follower configuration, the LuminAria did not have enough gain. So I replaced it with a Sony VFET 2SK79 preamp. It had much more gain and voltage swing, and it was very similar in sound to the Luminaria.
This DIY 2022 FE has much less distortion than either of the previous preamps and to my ears, the difference is similar to the difference between common source and common drain single ended SIT amps.
The DIY 2022 FE has less gain than the 2SK79 preamp though. Since I use my preamp to amplify the output of my TDA1541A dac that only has a 65R IV resistor at its output, the 10X gain is just barely enough, mainly because my speakers don't need a lot of voltage. I may play with the gain or increase the IV resistor value.
The LuminAria was two preamps ago. When I switched my SIT power amps to follower configuration, the LuminAria did not have enough gain. So I replaced it with a Sony VFET 2SK79 preamp. It had much more gain and voltage swing, and it was very similar in sound to the Luminaria.
This DIY 2022 FE has much less distortion than either of the previous preamps and to my ears, the difference is similar to the difference between common source and common drain single ended SIT amps.
The DIY 2022 FE has less gain than the 2SK79 preamp though. Since I use my preamp to amplify the output of my TDA1541A dac that only has a 65R IV resistor at its output, the 10X gain is just barely enough, mainly because my speakers don't need a lot of voltage. I may play with the gain or increase the IV resistor value.
Hi Rahul,
I have the bipolar BOSOZ PSU for a 30-40V circuit that I made when I was playing with the BA-3 FE. Caps are 50 V and I get about ~36V +- and can add zeners (and replace the trafo) to increase to say -+ 40V with the current caps. What voltage value did you shoot for?
Last edited:
I just read the Balanced Zen Line Stage article and yes, my power supply is basically the same. However my inspiration for the power supply was Michael Rothacher's LuminAria, which used the same type of power supply.
The board that I did for this preamp can be configured as two single voltage supplies or one bipolar voltage supply. For this build I chose to go with the two single supplies so that each channel has its own supply. The power transformer that I used has two 24V secondaries so I connected them in series to power both channels. So only the power transformer is shared by the two channels.
I am willing to share the Gerber files with diyAudio members if they want to use it for their DIY 2022 FE build.
Many thanks for the measurements.
The JFETs and the PNP have different tempco.
So DC drift is to be expected.
As you have measured, ~0.3V.
But the absolute DC level is higher than expected at almost 3V !!
Patrick
For this one, I would not go beyond total 70V plus to minus, whether bipolar or single-ended. I recall Papa saying the JFETs can withstand up to 70V, but not sure that will ever be used. My own supply wa +-27V and I wouldn't go beyond +-35V.
You can just change the zeners to get the the required voltage, if the raw supply is capable of > than the regulated voltage minus the Vgs, plus a few volts for staying within regulation.
Ben, you can make the zener follower adjustable by connecting a pot across the zener and taking the gate connection from the pot wiper. Maybe some more parts are needed for safety in case the wiper fails. I can draw it up if you like.
But the absolute DC level is higher than expected at almost 3V !!
Yes, it is almost exactly the offset at the gate of Q2 multiplied by 11 ((R2 + R4) / R2).
But you can reduce that by tweaking R7.
If C2 is put back in the circuit, the offset at R5 will be essentially the offset at the gate of Q2.