10V higher than max input signal is more than adequate for max performance.
Eg CD source of 2V rms (2.8V pk) lands you at around +/- 12 Volt rails.
Hmm. So does that mean if running from Balanced output on the source at 6.8Vrms instead of 2.2Vrms, that I need higher voltage PSU, like +/- 17 Volt rails in order to get best results, or am I misunderstanding the implied math in your statement?
Then again running balanced would be 2 B1 boards rather than one, so...
If you are making a balanced B1 then each half of circuit sees half that Voltage, so you don't need to adjust anything.
Besides that, 10V overhead is approaching overkill so a liitle under that spec will be fine.
If you have a lab powersupply laying around it doesn't hurt to experiment though.
Besides that, 10V overhead is approaching overkill so a liitle under that spec will be fine.
If you have a lab powersupply laying around it doesn't hurt to experiment though.
Last edited:
Makes sense to build: Stage 1 - one board as single ended - test of the preamp in my system as better than existing. Stage 2, rewire the one board as balanced one channel - proof of concept. Stage 3, build a second board for other channel - testing with existing 24v laptop PSU. Stage 4- build linear PSU to needs.
Some newbie questions during layout
- Original B1 says all resistors are ¼ watt. Same for B1 Rev2?
- Original B1 says the input impedance of the preamp is the same as that of the potentiometer. Same for B1 Rev2?
- If I understand correctly, the SK170 (N channel) wiring is: Gate-Signal In, Source-Signal Out, Drain-Voltage in. Is that right?
- If I understand correctly, the SJ74 (P channel) wiring to circuit is opposite: Gate-Signal Out, Source-Signal In, Drain-Voltage in. Is that right?
- It is said that for full passives, the wiring between potentiometer and in/out jacks should be as short as possible, often using a shaft extension. Is that also at issue for a buffer passive such as the B1 and B1 Rev2?
- Are R101/R102 between the potentiometer’s Earth and RCA Neg? My schematic literacy is a bit challenged and it looks like two terminals of the potentiometer are wired to join just before IN1. Is that true or does the switch side go to Pos and the resistor side go to Neg?
- If the answer to 6 is that both P101 In and Earth join just before the RCA Pos, I would think the negative voltage input to the board (D-) should NOT be connected to the RCA Negs because it is not Ground but an actual negative voltage. Is that right? If so, then the RCA Negs should get ground from chassis, or have direct connection to PSU Earth, or…?
Last edited:
OK, so let's forget most of the questions in the above post, as I've already found an answer or can experiment. But I do need help with these questions still.
Same pin outs when facing the part's writing- DGS
2SK170 Datasheet(PDF) - Toshiba Semiconductor
2SJ74_07 Datasheet(PDF) - Toshiba Semiconductor
-D is not ground. Just route back to a single ground point between +D and -D.
See ground symbol. Isolate the RCAs as usual with shoulder washers.
You can join all input grounds to a bus bar and send to PS ground or each ground to the PS ground with a dedicated wire for a "star-ground" config. Run output returns to same ground point with their own wires.
Vince
2SK170 Datasheet(PDF) - Toshiba Semiconductor
2SJ74_07 Datasheet(PDF) - Toshiba Semiconductor
-D is not ground. Just route back to a single ground point between +D and -D.
See ground symbol. Isolate the RCAs as usual with shoulder washers.
You can join all input grounds to a bus bar and send to PS ground or each ground to the PS ground with a dedicated wire for a "star-ground" config. Run output returns to same ground point with their own wires.
Vince
So I correct it to: both SK170 & SJ74 wiring is: Gate-Signal In (fr R103), Source-Signal Out (to P102), Drain-Voltage (+/-D).
OK, no AC input is required in this schematic! So… some connection of the chassis to true earth. Even if an LPSU is in the same enclosure, and the AC input socket has earth to chassis connection, perhaps more often signal ground is a star ground at a separate point on the chassis, in this case, where R104 & L104, RCA in, & RCA out are connected to the chassis at the same point. Better?
Thank you.
Last edited:
I don't understand what you are trying to say/do.
Separate mains Safety from audio requirements.
Your Mains powered equipment Chassis needs a Protective Earth (PE) wire connected directly and mechanically from the third wire (green/yellow) in the mains cable. Do this near the entry point through the Chassis wall/panel. Don't confuse this SAFETY requirment with anything audio.
Now consider your audio.
Do only what you need on the bench without any chassis to confuse the assembly. Get it working.
After that work out how you fold that working assembly into a size that will fit inside the Chassis.
Now make an extra Safety connection. Add a link from Main Audio Ground (MAG) to the Chassis. This connection is likely to be near the PSU. It does NOT go all the way back to the mains input cable.
Last edited:
Yeah, me thinking a bit sloppily again. Certainly Protective Earth is about safety, and Signal earth is about a circuit common. Since there are also shielding issues to be considered in the end product, I have a tendency to think about that at the same time and so was tending to merge all the discussion into one. As you pointed out that is a bit woolly and not helpful.
So it is determined by current draw. Then 1w or 2w should also be plenty.
Sent from my GT-I9505 using Tapatalk
Many circuits will tolerate +-10% for many of their components. Thus we used 5% carbon film resistors that drifted out to +-10% over the decades and the circuits still worked.
Some circuits work better/best when close tolerances are specified for a few components. Here +-1% is generally good enough.
In extreme cases one might specify +-0.1% and even tighter, but this is pretty rare in ordinary audio circuits.
Balanced impedance connections achieve very good attenuation of interference because they specify <0.1% for the resistors at either end of the balanced impedance connection and require tight tolerance on the cable manufacture.