OK, take fig 3. Remove RF2 completely. Add (or retain) the two 1 ohm resistors in parallel with the balance pot. Reduce the balance pot to 20 ohms. Adjust RF1 to get the right gain
john curl said:
No John, that would change the gain and the noise performance.
Actually, the split feedback networks could be in principle unified, but for certain reasons (I'll let you figure it out 🙂) that's not a good idea.
Edit: A balance pot is not required, this one has a servo that takes care of the possible offset. The feedback loop takes care of distortions (28dB available, so strictly matching the JFETs is not critical for the distortions requirements in a MC headamp).
john curl said:
John, I asked 'without changing the amount feedback?'. This also means, of course, without changing the gain.
Now that Syn08's preamp is now outside the pathological range of operation, perhaps we should talk about other noise influences, rather than just the input stage. I had hoped to do this weeks ago, but got diverted.
In this, I mean power supply noise, both internally generated and externally induced from the power line, etc.
We have gone through heavy shielding before, but that has to be noted here, in any case.
In this, I mean power supply noise, both internally generated and externally induced from the power line, etc.
We have gone through heavy shielding before, but that has to be noted here, in any case.
While this may now be obvious, the power line has become much more contaminated with medium and high frequency garbage over the years. Since we only want to convert the 50-60Hz signal into an acceptable DC with a small amount of ripple that we can easily regulate out with a typical IC regulator, we must be careful with this added garbage, and make sure that it doesn't get into our grounding system. Yes, the grounding system, as we usually add enough series regulation to knock out most of it.
With todays high efficiency lighting, light dimmers, computers, and all other components with switching supplies, there is even more garbage than before. Even a few years ago, we still had electric stoves, dimmers, electric motors, etc, and but now it is even worse.
There are still practical phono stages being sold that use batteries exclusively, in order to get around this problem. Fets take too much current to operate to do this easily, so we had better address the issues directly.
For example, one big problem is toroid power transformers. Why?
With todays high efficiency lighting, light dimmers, computers, and all other components with switching supplies, there is even more garbage than before. Even a few years ago, we still had electric stoves, dimmers, electric motors, etc, and but now it is even worse.
There are still practical phono stages being sold that use batteries exclusively, in order to get around this problem. Fets take too much current to operate to do this easily, so we had better address the issues directly.
For example, one big problem is toroid power transformers. Why?
YES, the coupling is much stronger capacitively than you would normally think. IF you have a cap meter, just measure the capacitance between a primary lead and a secondary lead. It might surprise you. Then, if you can, measure the capacitance in the same way with an 'old fashioned' split bobbin transformer of roughly the same power rating.
So when you measure that C, should the each end of the primary and secondary be connected together, ie short the input and output, or leave them open?
Power amps are a special case. It costs too much to NOT use a toroid transformer, but several companies go out of their way to avoid it, including Ayre and YBA. Big money, but an improvement. Other things can be done to get around the toroid problem as well.
When measuring capacitance, I don't think that you have to connect anything special. It is all relative, and Hi Z, rather than low Z where the transformer normally thinks it is operating.
I got rid of the toroids in my MC preamp and line amp, now using the Signal transformers you recommended. Any suggestions for power amps that use those big toroids?
Is there a reason why you can't or shouldn't apply series resistance and/or inductance between both rectifier bridge outputs all the way through the filter output? By that I mean in both the usual positive (or negative) connection as well the output normally connected to "ground." Balanced filtering, if you will, instead of unbalanced filtering.
The idea is that ground is only supposed to be an internal reference point, and hopefully a clean one at that. Why connect it to the noise coupled through the transformer if you can avoid it?
"Double insulation" so that you can remove the ground wire back to the mains connection is also a good idea. That removes at least one of the common mode return paths. (see Henry Ott's book...)
The idea is that ground is only supposed to be an internal reference point, and hopefully a clean one at that. Why connect it to the noise coupled through the transformer if you can avoid it?
"Double insulation" so that you can remove the ground wire back to the mains connection is also a good idea. That removes at least one of the common mode return paths. (see Henry Ott's book...)
- Status
- Not open for further replies.