I used the 389 in the X-Altra MC/MM phononstage for the MM part and got stirling distortion and RIAA conformance results. Since the drain was configured inside the composite amplifier loop, the voltage changes at the drain are very small somewhat minimising the effect of the input capacitance. The RIAA all-active feedback also helps here since it’s sloping at ~-20 dB per decade from LF. This of course would not be the case in a passive pre where you typically have common source gain stage driving the EQ network.
The 20 kHz audio BW is small in the big scheme of things and not that demanding in the small signal regime. I guess if you were trying to build a very wide BW amp (100’s of kHz and up) the 389 would not cut it.
The 20 kHz audio BW is small in the big scheme of things and not that demanding in the small signal regime. I guess if you were trying to build a very wide BW amp (100’s of kHz and up) the 389 would not cut it.
Quick update on the X-Altra Mini II.
The MC/MM phono amp is tested and working very well with tight RIAA EQ as measured on the QA401. Its also very quiet - the MC only stage is under 400 pV/rt Hz (sim).
In general, I still have some work to do on the overall noise. I have mains harmonic peaks on the line inputs fitted with -15 dB attenuators at around -115 dBV giving a measured S/N of -103 to -106 dB (20 kHz BW). On the non attenuated inputs, the mains harmonics are at -130 dBV giving better than -110 dB . I’m shooting for all mains harmonics peaks to be better than -125 dBV at any volume control setting and on any line input.
On line level inputs, the thermal spot noise is -140 dBV with the volume control set to 0 and the same at max with the selected input shorted.
I‘m away until the end of the month, but will pick things up again in May, when hopefully I can crack the problem.
The MC/MM phono amp is tested and working very well with tight RIAA EQ as measured on the QA401. Its also very quiet - the MC only stage is under 400 pV/rt Hz (sim).
In general, I still have some work to do on the overall noise. I have mains harmonic peaks on the line inputs fitted with -15 dB attenuators at around -115 dBV giving a measured S/N of -103 to -106 dB (20 kHz BW). On the non attenuated inputs, the mains harmonics are at -130 dBV giving better than -110 dB . I’m shooting for all mains harmonics peaks to be better than -125 dBV at any volume control setting and on any line input.
On line level inputs, the thermal spot noise is -140 dBV with the volume control set to 0 and the same at max with the selected input shorted.
I‘m away until the end of the month, but will pick things up again in May, when hopefully I can crack the problem.
Last edited:
Hi Andrew,
A like your design. It’s very well thought out with design that works with readily available parts. I agree on your choice of pot value and placement of the gain stage as early in the signal path as possible.
Attached is a simple simulation of the gain structure.
The top purple line is the source input which we can refer to as 0 db.
The blue curve is the output of the fixed input pad on the cd input. The attenuation is - 14.3 db.
The green curve is the output level of the Log 20k pot at 50% rotation and attenuates the input signal overall by - 29.8 db. The resistance is about 9k at 50% rotation.
The blue curve is the discrete line output level -14.7 db at 50% rotation of the volume pot.
ln the case of the un attenuated inputs at a nominal 300mv the discrete line output is of course 15db above the source input level. The level is about 1.7 volts ref an input of 300mv. Perfect for driving a power amp with an input sensitivity of 1.7 volts .
Excellent.
A like your design. It’s very well thought out with design that works with readily available parts. I agree on your choice of pot value and placement of the gain stage as early in the signal path as possible.
Attached is a simple simulation of the gain structure.
The top purple line is the source input which we can refer to as 0 db.
The blue curve is the output of the fixed input pad on the cd input. The attenuation is - 14.3 db.
The green curve is the output level of the Log 20k pot at 50% rotation and attenuates the input signal overall by - 29.8 db. The resistance is about 9k at 50% rotation.
The blue curve is the discrete line output level -14.7 db at 50% rotation of the volume pot.
ln the case of the un attenuated inputs at a nominal 300mv the discrete line output is of course 15db above the source input level. The level is about 1.7 volts ref an input of 300mv. Perfect for driving a power amp with an input sensitivity of 1.7 volts .
Excellent.
Attachments
Last edited:
That is excellent, then an abbreviated version can be built and later expanded as required. Looking forward to start with the discrete line amp.
The KX2 housing is just built from recycled materials, of importance is sq and it sounds lovley. Thanks for excellent design and clear instructions. I trust the preamp will be of the same high class.
The KX2 housing is just built from recycled materials, of importance is sq and it sounds lovley. Thanks for excellent design and clear instructions. I trust the preamp will be of the same high class.