The BC547 are only placeholders. I use the ROhm 2SD786. I find them superior even to the ZTX851 because they have a much bigher Hfe, so direct coupling is less problematic.
Remember that 12 years ago i started that thing with super low noise here.
I know my noise calculations and my candidates.
Yes, i designed the Virgo 2. It got a 16 page test with home story in Stereophile in the late 90th.
That put AudioPhysic on the map and we had many years of prosperity.
I am from time to time in the Netherlands and could visit you.
Remember that 12 years ago i started that thing with super low noise here.
I know my noise calculations and my candidates.
Yes, i designed the Virgo 2. It got a 16 page test with home story in Stereophile in the late 90th.
That put AudioPhysic on the map and we had many years of prosperity.
I am from time to time in the Netherlands and could visit you.
Well, you’re not short of transistors.
The ZTX has an Rbb of 1.5R versus 4 to 6R for the ones you have.
With two transistors that’s 7R less or 0.34nV/rtHz.
And the ZTX has a sister with much higher hfe with an Rbb of 3R.
But differential input current for MC can be easily compensated as you have shown yourself.
Hans
The ZTX has an Rbb of 1.5R versus 4 to 6R for the ones you have.
With two transistors that’s 7R less or 0.34nV/rtHz.
And the ZTX has a sister with much higher hfe with an Rbb of 3R.
But differential input current for MC can be easily compensated as you have shown yourself.
Hans
Last edited:
Transistor is the ZTX1053A with an Hfe between 300 and 500.
But Rbb is with 6R slightly worse as your 2SD786 so if you want a high Hfe, the 2SD786 is a perfect choice.
In your pre-pre 2*1/2Gm comes in series with 2*Rbb, with the 3R between the emitters plus the source resistor of your Cart.
Assuming a 20R cart and 6mA collector current, this calculates into:
For the ZRX851:
2*2.08R + 2*1.5R + 3R +20R = 30.2R giving 0.71nV/rtHz
For the 2SD786:
2*2.08 + 2*5R + 3R +20R = 35.1R giving 0.76nV/rtHz
So in effect the noise difference between both transistors, with a 20R Cart connected, will only be a negligible 0.6dB
Hans
P.S. What a fantastic promotional review in Stereophile.
But Rbb is with 6R slightly worse as your 2SD786 so if you want a high Hfe, the 2SD786 is a perfect choice.
In your pre-pre 2*1/2Gm comes in series with 2*Rbb, with the 3R between the emitters plus the source resistor of your Cart.
Assuming a 20R cart and 6mA collector current, this calculates into:
For the ZRX851:
2*2.08R + 2*1.5R + 3R +20R = 30.2R giving 0.71nV/rtHz
For the 2SD786:
2*2.08 + 2*5R + 3R +20R = 35.1R giving 0.76nV/rtHz
So in effect the noise difference between both transistors, with a 20R Cart connected, will only be a negligible 0.6dB
Hans
P.S. What a fantastic promotional review in Stereophile.
Thank you for the flowers.
I have used ZTX1051A.
As far as I can tell they have around 7 Ohm Rbb‘ and high Hfe.
They also have a somewhat lower 1/F corner then the ZTX851.
Yes, the resistance of the coils of an MC cartridge produces thermal noise and often is the limit how low you can go with noise. Therefore I am always trying to find an MC cartridge with relatively low coil resistance and healthy output. A good example is the Goldring Ethos.
I have used ZTX1051A.
As far as I can tell they have around 7 Ohm Rbb‘ and high Hfe.
They also have a somewhat lower 1/F corner then the ZTX851.
Yes, the resistance of the coils of an MC cartridge produces thermal noise and often is the limit how low you can go with noise. Therefore I am always trying to find an MC cartridge with relatively low coil resistance and healthy output. A good example is the Goldring Ethos.
I had originally similar figures as you just showed, but after extensive searching this showed to be caused by the Nichicon supply caps.
By replacing them by Panasonic 105C FM caps the problem was completely solved.
See noise image below with those initial Nichicon caps.
Syn08, a well respected member on this forum changed these caps in his low noise phono amp with similar positive results.
Hans
.
By replacing them by Panasonic 105C FM caps the problem was completely solved.
See noise image below with those initial Nichicon caps.
Syn08, a well respected member on this forum changed these caps in his low noise phono amp with similar positive results.
Hans
.
Attachments
When noise is becoming that low, every component should be of superior quality and leaking electolytic caps that normally do what they are supposed to, can obviously cause severe problems in very low noise designs.
Hans
P.s. with the Panasonic caps installed, I tested noise with tens of ZTX851/951 and they all had their 1/f under 10Hz.
Hans
P.s. with the Panasonic caps installed, I tested noise with tens of ZTX851/951 and they all had their 1/f under 10Hz.
Last edited:
Here is the same spectrum with the Nichicons but now also on a log freq scale.
Here you see a very unusual 30dB/oct slope, a slope that was also visible in Syn08’s completely different design.
Because that being quite strange for 1/f made me search for a possible cause, eventualy landing at the Nichicons.
Hans
.
Here you see a very unusual 30dB/oct slope, a slope that was also visible in Syn08’s completely different design.
Because that being quite strange for 1/f made me search for a possible cause, eventualy landing at the Nichicons.
Hans
.
