Thanks 2quad, that should teach Ed to watch out with what he digresses with. His approach can also be very noisy, as well. Why?
Impossible to repair, Joachim, but 600 is better than 200, won't you agree? I'm pretty sure that I brought this up earlier.
Pavel, do not laugh. This one may get popular. And it is still developping. Doing a PCB in the old blueprint style whould be fun too.
Anyway, i am very surprised how much you can do wrong in a primitive circuit like this.
Good and very cheep pots : Ohmeg
Good and cheep input switch : Lorlin
Anyway, i am very surprised how much you can do wrong in a primitive circuit like this.
Good and very cheep pots : Ohmeg
Good and cheep input switch : Lorlin
Yes, i should work well at typical listening levels. But what is that ?
90 dB avarage with 100dB peaks at the listening seat ?
90 dB avarage with 100dB peaks at the listening seat ?
That is the point. It's better to keep the common mode voltage low. You do that by increasing stage gain.
There's other solutions available:
Change the stage topology
Place a variable resistor in series between the feedback voltage divider and the inverting of the opamp that nominally tracks the volume control pot
Bootstrap the supply rail that the substrate is attached to
Probably 4873 others, too. But they are all more complicated.
Joachim, you were right, I made a mistake in the phono gain. I calculated in 40dB rather than 60 dB for the offset. Yes, we could get up to 2V output offset, and typically .5V offset. With a cap in place, big deal. We can live with it.
This is a trivial amount of extra distortion compared to the phono cartridge, itself. Perspective is necessary.
In many reinforced sound systems people will complain it is not loud enough if the "A" fast average level is not at least 92db. Then other people will complain it is too loud when it is above 95db on the same scale. Of course my experience is with only a few percent of The U.S. population.
I expect peaks to be 10db louder for train stations, 20db for recorded music systems and 30db for a live orchestra. That is based on my personal preferences.
John,
When I publish a design or design tips I intend for the readers to use it as a starting point. I think the fun in advanced DIY is the exploration and learning. So my circuits will work so as not to leave unhappy newbies, but the clever folks will improve them.
Of course for even this basic preamp we could add a second inverse RIAA network to the output of the first stage to drive an active shield on the input signal. Five parts and almost no more cable capacitance losses.
My preamp uses switched resistors, but since I built it, better resistors have shown up in my research.
For this preamp I would use Xicon 1/2W metal film resistors, really cheap and probably the third best audio resistor I have measured. (Caveat Emptor, I have only measured small samples and at one value only!)
Last edited:
Yes, i understand the advantage of the 600 Ohm resistor. It does not load the Opamp too much in the treble and the RIIA caps are smaller and cheeper. I was too focussed on lowest noise but your example has shown that we talk about very little here.
The coupling cap can be made smaller if the 50 Hz pole is lowered to provide some compensation. E.g. pic below shows the error with two poles at 40hz and 25hz instead of one at 50hz. -6dB point is at about 10hz.
The amplitude response of the arm-cart combination has most of the time a rising trend under 20Hz anyway, so your idea whould work fine.
So we should set the cap to 25Hz - 3dB and the RIAA from 3180 usec to 3975 usec ?
So we should set the cap to 25Hz - 3dB and the RIAA from 3180 usec to 3975 usec ?
If you think so John, but then if you get too cocky I can do actual measurements!
Or perhaps not so drastic. A nice feature is it allows some wiggle-room with the 50hz pole, which might make it easier to optimize the 500hz and 2khz time constants with standard resistor and cap values, then choose the coupling cap to fix the low end response.
-3dB at 10Hz would be a good compromise. Trust me, I did the math 30 years ago.
Dimitri, you are WAY out of line. Let us continue.
Dimitri, you are WAY out of line. Let us continue.
Dimitri, the passive pole lowers overload margin. That is the reason i presented the Self design that takes only care of the 4th pole without lowering overload margin by much but it was considered as too complex.
Another thing i do not like with the ( well known ) Baxandall topology is that transients from the cart. are only filtered out after the opamp.
- Status
- Not open for further replies.