Offline
Poland, Gdansk

Hey guys,

I designed a board that helps with ground loop problems using nanoloop 2 - https://github.com/kiwanowski/Nanoloop- … p-Isolator

I had some problems with noise while trying to sync nanoloop 2 on a AGB-001 with an Elektron Analog Rytm.

I hope that maybe someone will find this design helpfull - it basically uses a 6N138 on the nanoloop sync signal.

Last edited by scannerboy (May 21, 2016 9:14 am)

Offline
Sweeeeeeden

A few comments on the project:

You probably don't need the USB for power. You could probably use the enable signal on the DIN jack for that as long as you don't draw too much current! The DIN sync standard specifies that an instrument should wait 20 ms from turning enable high until it starts sending clocks. In theory all instruments should obey by this and if the microcontroller is ready to accept clocks by then you should be good to go.

For this to work, you need to change the SUT fuse bits 00 or 01 from the standard value 10, otherwise the microcontroller will take too long (~65 ms) to boot.

Also, if doing this, you should add something like a 220 ohm resistor on the sync input to prevent shorting sync and enable if the instrument ever sends a clock pulse while enable is low. Some instruments do this because certain DIN sync slaves require this to "warm up". The shorting mentioned would happen through the AVR's internal protection diodes, which are reverse diodes to gnd and vcc normally meant to absorb voltage spikes above Vcc or below Gnd.

I'll also question the use of a clock frequency as high as 20 MHz for something like this. Not only it this above the specified range for Atmega8, but you shouldn't need such a high frequency for what you're doing. Or the use of a clock crystal at all. Since there are no precise timings involved, you should be able to just use the internal RC oscillator at for example 8 MHz.

Then there's the optoisolator. You're using the SD pin as a voltage source. While this will mostly work on GBA link cables, this signal is typically unconnected on DMG/GBC link cables and the device may not work when using those. I have ideas about how to solve this, but I would need to test those.

Finally, the ground plane extends close to the secondary side of the optocoupler. This isn't really a problem in this case since you're not using the optocoupler to isolate between high voltages, but I would still have modified the ground plane to have some distance to those traces. (Or put a restrict polygon over it to achieve the same effect.)

Offline
france

Could you describe the issue ?
Today I'm using nanoloop 2.7.x with analog sync in with an electribe 2, it is rock solid. The cable come from one guy on this forum, great cable smile
I don't have try it with an analog rythm, because I don't own one, i'd like to have one, but I like detail, and what is this ground issue ?

Last edited by yoyz2k (Jun 29, 2016 4:35 pm)

Offline
IL, US
yoyz2k wrote:

Could you describe the issue ?
Today I'm using nanoloop 2.7.x with analog sync in with an electribe 2, it is rock solid. The cable come from one guy on this forum, great cable smile
I don't have try it with an analog rythm, because I don't own one, i'd like to have one, but I like detail, and what is this ground issue ?

ground loops cause a humming sound

Offline
france

Yes, in this case I have it too, not too loud, but there is some noise from the gba. And I hear the pulse sync.

Need to see more about this project.