Din Sync RE-909 Дополнительные инструкции - Страница 3

Просмотреть онлайн или скачать pdf Дополнительные инструкции для Оборудование для звукозаписи Din Sync RE-909. Din Sync RE-909 8 страниц.

1- OVERALL GROUND ISSUE FIX - GROUND & CROSSTALK FIX

Hi, just wanted to share some good news about noise and crosstalk issues on RE-909. Pictured below is a Roland

TR-909 factory mod with a wire that seems to reinforce grounding between C151 styrofoam cap in the hihat low pass

ﬁlter and the common analog ground track where R236 is also attached. By doing the same mod in RE-909, pictured

below also in the bottom of the board, cymbals crosstalk and digital noise leak into the hihats channel is eliminated,

which ﬁxes the last remaining issue in my build. I used a thick grounding wire to connect the lower pin of C151 to the

double hole in the common ground track next to R236. This ﬁx works perfectly on traﬀo PSU but I can't guarantee it

works on safe PSU (and preliminary tests show it doesn't) because of the diﬀerence in ground routing. Otherwise, enjoy

your digital noise-free, crosstalk-free RE-909!

2 - SD GAIN SCALING

Heads up for those who commented about their snare snappy being way too loud compared to the original. Some of us

worked together to look for a cause and we ﬁnally did ﬁnd something. It took a bit to locate an original TR-909 to

analyse under the scope but once that was available it was easy to see.

We followed the path of the noise signal in the snappy section of the snare in R286 right up to the Q47 and Q48 VCA

transistors and noticed it was exactly the same as the TR up until the voltage divider resistors R298 and R279 where

the signal was roughly doubled. We repeated this measurement across 3 RE-909 (ﬁrst and second run) builds and one

(second run) which didn't have this issue at all. We believe ground impedance could aﬀect some builds, which in turn

could interfere in the voltage divider resistors. We measured 4Vpp noise (3V avg) in the snappy noise measurement

point (between C78 and R294) in RE-909 builds vs 1,70Vpp (1,2V avg) in a TR-909. Carter Goertzen used a

potentiometer to match the noise gain exactly to the TR and a solution was found by replacing R294 with a 1% 200K.

This value might be more or less universal if your snappy noise measurements match ours. We compared the output to

other 3 TR-909 samples and the result behaves like those and ﬁnally feels much better balanced.

Another possible ﬁx would be to play with R298 and R279 values and 1% resistors to counter impedance before the

VCAs and get the gain right but it doesn't seem as practical IMO. 