I’m pretty unimpressed with it. The gold standard for elegance, simplicity and performance is Leach’s MC head amp that with ZTX851’s will get you to 232pV/rt Hz (measured).
My personal view for what it’s worth.
Anyway, let’s get back to the Tran Bal. if everyone is happy with the schematic in post #122 and the EQ bit I’ll put up in a day or two, I’ll start laying the board.
My personal view for what it’s worth.
Anyway, let’s get back to the Tran Bal. if everyone is happy with the schematic in post #122 and the EQ bit I’ll put up in a day or two, I’ll start laying the board.
I'm too lazy to calculate of fire the simulator today, what is the expected bias collector current? I'm not sure how much dynamic range this setup may have. Also, the 15Hz pole in the servo (100k/0.1u) is too high, IMO.
But the values may be adjusted later.
Last edited:
I had a look (page 483 Fig 8.12) and it seems you can go quite some way beyond 10mA before the noise reduces no further and becomes asymptotic at somewhere between 150pV/rt Hz and 200 pV/rt Hz.
I'm going to go with the already suggested 10mA on this for now - easy to change it lower though by adjusting R7//R8. The 10mA per collector is about the upper limit from a LM4562.
In the plot below, the lowest noise density is with Ic = 15mA, middle 10mA and highest noise at 5.5mA
I'm going to go with the already suggested 10mA on this for now - easy to change it lower though by adjusting R7//R8. The 10mA per collector is about the upper limit from a LM4562.
In the plot below, the lowest noise density is with Ic = 15mA, middle 10mA and highest noise at 5.5mA
Attachments
Last edited:
No, but then why would anybody use bipolars instead of a LSK389 (or any pair of low noise JFETs) for the same noise. No input current noise to speak of, flexible input impedance, gain independent on the cartridge impedance.
A cartridge with 0.5mV @1KHz will provide 5mV @20KHz. At Ic=5mA, the max input stage differential gain (for 300ohm load) is 60, or 300mV at the output, or about 1Vpp. Meaning the overload margin is only 6dB (to 2Vpp across collectors). You think this is enough?
I don't get your numbers. I've set the colllctor load resistors at 300 Ohms each, set the input to 10mV (so 26 dB overload) and stepped Rcart from 3-12 Ohms, Ic at 5.5mA and this is what I get:-
Attachments
This amp is meant as a head amp, true?
In that case 20 to 30dB gain should be enough.
OLM at 20Khz would then be 50mVrms or 20dB for:
2R Cart with 100R collector resistors,
10R Cart with 200R coll. resistors and
15R Cart with 300R coll. resistors.
In these three examples giving 32dB gain.
Hans
In that case 20 to 30dB gain should be enough.
OLM at 20Khz would then be 50mVrms or 20dB for:
2R Cart with 100R collector resistors,
10R Cart with 200R coll. resistors and
15R Cart with 300R coll. resistors.
In these three examples giving 32dB gain.
Hans
Why? The need for balanced path ends once the cartridge signal in processed in the input stage. Some may prefer to have a balanced signal path all along from cartridge to the power amp, to me that's not worth the complexity. We are talking home audio, not pro studio equipment.
I was thinking about a situation where an MM cart is used as an alternative and you connect that behind the MC-head-amp to the following stage.
That does not mean, that the head amp needs to feed a symmetrical signal into next stage but if the next stage would be able to also use a symmetrical input with improved CMMR, wouldn't that be nice?
So this isn't so much about the specific MC head-amp design but rather thinking about how it would be part of a bigger phono stage that offers both MC and MM inputs, like the X-Altra.
Optimal would be to set the voltage on the collector resistances halfway between 4.5Volt and 1.5Volt to get the maximum achievable voltage swing.
With the chosen collector current, this is when both collector resistances are 220R.
In the image below the OLM in dB versus Rcart is visible in blue, now with 220R collector resistances.
Also shown in orange are the needed resistance values on the diff to se amp (the ones that are now 2K) to get an overall gain of 30dB, just to get a proper idea.
For a higher or lower gain this value can simply be adjusted.
So IMO it's better to make those resistances changeable and leave the collector resistances at 220R, that will in all cases give the highest possible OLM.
Hans
.
With the chosen collector current, this is when both collector resistances are 220R.
In the image below the OLM in dB versus Rcart is visible in blue, now with 220R collector resistances.
Also shown in orange are the needed resistance values on the diff to se amp (the ones that are now 2K) to get an overall gain of 30dB, just to get a proper idea.
For a higher or lower gain this value can simply be adjusted.
So IMO it's better to make those resistances changeable and leave the collector resistances at 220R, that will in all cases give the highest possible OLM.
Hans
.
Attachments
But why not just switch the collector load resistor values in parallel? The issue now is there is a string with a total of 350 ohms and I am tapping off the string. The collector always sees the same high load resistor. If I rearrange the switch, then it’s easy enough to change the load R from say 25 ohms (high output low R cart) to say 220 or 250 for a low output current cart.
Also, very easy to raise the 5 V rail by 1 or two volts for more headroom.
Also, very easy to raise the 5 V rail by 1 or two volts for more headroom.