All I see in Farnell for that price are Polyester caps (MKT) - are these good enough? MKT is not the best sounding as signal passing capacitors - polyprop are better. The role of the cap here is noise reduction, I think - there must be a further role which influences the sonics, otherwise electros would suffice!
I am not sure of what capacitor distortion and resultant filter bandwidth is going to work synergistically with what sreg catered audio circuit. This is a highly subjective area. As this is the Vref which is essentially the goal of the error loop, whatever noise filtering bandwidth, lead/wire inductance interaction, external noise pick up, conditions you create, are going to magnify and reflect. I prefer the shortest, closest solution, with enough low frequency decoupling. I guess that local decoupling experience from digital electronics and op amps is handful here. What about testing say 1000uF OSCON?
Certainly OSCONs are considered good for digital PS bypassing but not particularly for analog supplies - I've experienced this myself - don't know exactly why this phenomena - OSCONs have high leakage (twice to 3 times higher than Panny FMs), maybe this is one factor?
jkeny, search for MKP1839 at Farnell. See pic bellow (white cap).
For Vref I am using LED.
For Vref I am using LED.
Last edited:
I have just read from the other thread that the latest Salas v1 shuntreg has replaced the zener with a resistor and a trimpot, and added Q7.
Would the same Q7=BC550c, R8=3.3k work with +15V rail? And I suppose for the negative rail Q7=BC560c?
Would the same Q7=BC550c, R8=3.3k work with +15V rail? And I suppose for the negative rail Q7=BC560c?
As a general FYI, I have built a +15/-15 version of the V1 sss, which is on soak at the moment.
For Vref I used 8 red led's UK Farnell part number 1142517, Kingbright L-934ID.
I get +15.01Vdc out on the +ve rail, and -14.98Vdc on the -ve Spot On!
For Vref I used 8 red led's UK Farnell part number 1142517, Kingbright L-934ID.
I get +15.01Vdc out on the +ve rail, and -14.98Vdc on the -ve Spot On!
I haven't built a negative yet. Its for the 3 sensitivity Simplistic phonos. It will normally mirror I expect. Should be called a ''V1.1Var.'' if it is to be named. Stick with your Leds since you know your target Vout though. Lower Vref impedance, no tiny wiper to get potentionally noisy. You can just add the emitter follower later to see if it woks first, and what it subjectively does in your aplication and it will be a V1.1Led. It was better on the phono in my system, but I don't know how it will translate in other systems or for line level applications yet. Short of a Beta.
For those that are too lazy to build the -ve version. I wonder if this would be worthwhile trying? Virtual Ground Circuits
Those who want to best swallow the bitter pill and build the -ve version. Otherwise just use lm317/337 and be happy, or yes, virtual ground circuits... but unless you want something portable, there isn't much of an excuse to settle for a virtual ground. If you go to the trouble of building a good shunt regulator, why spoil it with an opamp base virtual ground?
Edit: it's just my opinion, of course. 🙂
Edit: it's just my opinion, of course. 🙂
Last edited:
Iko is right. It took me one year to just test a simple follower for the V1. So don't expect from me to test the -Ve mirror anytime soon since I don't have a wanting application right now. You are on your own.😀
Haven't tested.
OK OK. Forget about it. If you have to go to the trouble of making the virtual ground, you may as well go to the trouble of making the negative supply.
It's good to know about it, for sure, so thanks for mentioning it. Might be useful to someone.
I will give the emitter follower a go and just mirror the neg with BC560c. If it goes "Bang" then someone else can avoid the problem. I think this brings it closer to v2, as Iko found that adding the buffer extends the low impedance to higher frequencies.
I am not sure if the increase of the impedance of Vref to up to a few k ohms is very important, as the impedance of the bipolar transistor is still much higher than that of the Vref, or not?
From what I learnt from you guys the zener / LED is noisier so a resistor would be better. But interestingly, this is a reverse of the R6 mod, or not? The R6 mod was believed to provide better ripple rejections. With the resistor, would the Vref be as stable? same PSSR?
Of course, we are just looking for perfection. And this is what this hobby is about.
I am not sure if the increase of the impedance of Vref to up to a few k ohms is very important, as the impedance of the bipolar transistor is still much higher than that of the Vref, or not?
From what I learnt from you guys the zener / LED is noisier so a resistor would be better. But interestingly, this is a reverse of the R6 mod, or not? The R6 mod was believed to provide better ripple rejections. With the resistor, would the Vref be as stable? same PSSR?
Of course, we are just looking for perfection. And this is what this hobby is about.
- Status
- Not open for further replies.