Hei igjen alle mann og godt nytt år.
Har hat sørgelig lite tid til denne hyggelige hobbyen i det siste, men man kan ikke i de uendelige forsømme sine forpliktelser.
Jeg fikk en del mer info fra Wayne før jul som jeg klipper inn her. De teknisk kunnige kan tygge litt på den.
Føler meg snart moden for en bestilling og kommer primært å fremme innkjøp av ferdig loddede kort av ny type.
here goes
I received a small run of the new MC preamp boards on Friday and have spent the last few days running tests on it.
In addition to the improvements I mentioned earlier I, along with the help of Hans Polak, have discovered a way to significantly reduce IM distortion specifically the 1 kHz component of a 19/20 kHz 1:1 intermodulation test. The 19/20 kHz test and the resulting IM folded down to 1 kHz is a very difficult if not brutal test. As far as I know, other than the late Graeme Cohen, not much has been written about component matching certain resistors and compensation capacitors in the front end of this popular preamp topology. Cohen explained that he matched resistors and many infer that he did so for DC offset performance but he never actually came out and said why he did it.
On the MC preamp there are 4 pairs of resistors and two capacitors that, when matched to extreme tolerance, vastly reduce IM distortion. In the prototype the resistors were ratio-matched to <<0.1% and the capacitors to about 0.3% or less. I intend to hand-match all new assembled boards at no additional charge.
I've embedded below pictures of the boards. The original is on the top for comparison.
I've also supplied a distortion comparison of the zoomed-in 1 kHz IM component. The output level was +10 dBu which is about 2 dB over the internal (and often external) overload point of most A/Ds. The first frame is the original without hand-matched components, the second the new version with hand-matched parts. While there would be some improvement in IM for the original with matched parts the servo pulling technique of the new board also lowers IM.
The IMD figures shown in red are NOT correct as they were not properly set for 19/20 IM. The actual distortion is about 34 dB lower in the new board compared to the original. The correct IM percentages are 0.0075% for the original board and 0.00015% on the new one.
I now have enough boards in stock to build all or part of the group buy with new boards or can provide the quantities I have of the original.
With regard to gain switching off-board it can be done but the circuit node which controls it is super-low impedance, sensitive to capacitance and subjects the preamp to the potential for noise ingress. I don't recommend it. Also, the trim pot is suitable for high output MC carts but not low-output ones. At the high gains for low-output carts the adjustment is touchy. I stuck with jumpers to keep the performance up. The input termination however can be taken off-board if the leads are kept short and a six position Molex female plugged onto J1-J3.
As before you're welcome to share this information with the group. I'll be glad to answer any questions.
Best;
Wayne
Original board IM +10 dBu output. The actual IM percentage is 0.0075%. (A/D attenuated by 10 dB; 0 dBu displayed = +10 dBu)
(The spur to the right of 1 kHz is not distortion, is environmental and quite likely USB polling.)
New board IM +10 dBu output. The actual IM percentage is 0.00015%. (A/D attenuated by 10 dB; 0 dBu displayed = +10 dBu)
og noen bilder som fulgte med