Hi TMM,
You should be able to clean up all that grass above the harmonics. The loop back might be creating this. Do you have an external oscillator you can try? The EMUs are very sensitive to grounding.
This is the unmodded channel set to 0.1Vrms FS
Cheers,
You should be able to clean up all that grass above the harmonics. The loop back might be creating this. Do you have an external oscillator you can try? The EMUs are very sensitive to grounding.
This is the unmodded channel set to 0.1Vrms FS
Cheers,
Attachments
Last edited:
Unfortunately I don't have an external oscillator that performs better than the EMU's own DAC. The 'grass' is almost all harmonic content - i have a feeling it might be caused by ARTA only having a 20bit dither or just a hardware limitation of the DAC but i'm not an expert on the topic 
One weird (to me) thing that i noticed is that the 1/4" jack is a stereo one. The 'right' and 'left' terminals from the socket go to the opamps and the 'ground' terminal goes to AGND. The 'right' terminal has 10k between it and AGND. I guess this is so it still works with differential/balanced outputs? Any idea what affect this would have on using a stereo vs mono cable? I'm using a 30cm mono cable from the 1/4" 'right' output.
edit: the original circuit is perfectly well behaved when driven by a balanced input. It only becomes a problem when one input is tied to AGND. Both of our mods cause it to become asymmetric when driven by a balanced input. I guess there was method in their madness.
One weird (to me) thing that i noticed is that the 1/4" jack is a stereo one. The 'right' and 'left' terminals from the socket go to the opamps and the 'ground' terminal goes to AGND. The 'right' terminal has 10k between it and AGND. I guess this is so it still works with differential/balanced outputs? Any idea what affect this would have on using a stereo vs mono cable? I'm using a 30cm mono cable from the 1/4" 'right' output.
edit: the original circuit is perfectly well behaved when driven by a balanced input. It only becomes a problem when one input is tied to AGND. Both of our mods cause it to become asymmetric when driven by a balanced input. I guess there was method in their madness.
Last edited:
Unfortunately I don't have an external oscillator that performs better than the EMU's own DAC. The 'grass' is almost all harmonic content - i have a feeling it might be caused by ARTA only having a 20bit dither or just a hardware limitation of the DAC but i'm not an expert on the topic
One weird (to me) thing that i noticed is that the 1/4" jack is a stereo one. The 'right' and 'left' terminals from the socket go to the opamps and the 'ground' terminal goes to AGND. The 'right' terminal has 10k between it and AGND. Any idea what affect this would have on using a stereo vs mono cable? I'm using a 30cm mono cable from the 1/4" 'right' output.
If you could put a schematic up it would be easier to answer that question.
Cheers,
ok, some schematics.
Standard circuit, driven single ended. The tip of the connector is the + of the sine voltage source
, the ring is the - of the sine voltage source, and the sleeve is connected to AGND.
Shorting the negative side of my input source directly to GND makes no difference in the simulation compared to just letting it ground itself through the 10k resistor (I haven't traced out where this resistor is on the board yet).
Standard Circuit driven with balanced input - tip and ring are a balanced pair, sleeve is GND.
As you can see, it doesn't suffer the same problem.
If someone wanted to, they could remove that 10k resistor wherever it is and have a HighZ balanced input. This would probably then cause problems if it was driven single ended and the sleeve was left floating.
Standard circuit, driven single ended. The tip of the connector is the + of the sine voltage source
, the ring is the - of the sine voltage source, and the sleeve is connected to AGND.
Shorting the negative side of my input source directly to GND makes no difference in the simulation compared to just letting it ground itself through the 10k resistor (I haven't traced out where this resistor is on the board yet).
Standard Circuit driven with balanced input - tip and ring are a balanced pair, sleeve is GND.
As you can see, it doesn't suffer the same problem.
If someone wanted to, they could remove that 10k resistor wherever it is and have a HighZ balanced input. This would probably then cause problems if it was driven single ended and the sleeve was left floating.
Last edited:
ok, some schematics.
Standard circuit, driven single ended. The tip of the connector is the + of the sine voltage source
, the ring is the - of the sine voltage source, and the sleeve is connected to AGND.
Shorting the negative side of my input source directly to GND makes no difference in the simulation compared to just letting it ground itself through the 10k resistor (I haven't traced out where this resistor is on the board yet).
Standard Circuit driven with balanced input - tip and ring are a balanced pair, sleeve is GND.
As you can see, it doesn't suffer the same problem.
If someone wanted to, they could remove that 10k resistor wherever it is and have a HighZ balanced input. This would probably then cause problems if it was driven single ended and the sleeve was left floating.
Well the ring should short to sleeve which is ground in single ended use. If you plug in a mono plug this is what happens unless Emu used some funky jack that doesn't.
Last edited:
I'm going to get a good used HP 332,3,4A just for its variable notch filter and see if it can be used for tuning out selected harmonics or for use with QA400. They are cheap and the variable feature gives a lot of flexibility for many applications. Maybe some upgrades and mods to it for specific apps.
Thx-RNMarsh
Thx-RNMarsh
I Had a big difference configuring for to use ASIO driver in measures. The programs that allow ASIO driver is Arta and Rightmark.
Sound Card EMU 0404 PCI with Chipset ATI (using PCI card may have problems with new chipsets Intel) SO- Windons XP 32.
I believe it is related to latency WDM in Rightmark the value was 142ms, ASIO driver the value was 50ms.
Sound Card EMU 0404 PCI with Chipset ATI (using PCI card may have problems with new chipsets Intel) SO- Windons XP 32.
I believe it is related to latency WDM in Rightmark the value was 142ms, ASIO driver the value was 50ms.
Attachments
Last edited:
Has anybody here used vector signal analyzers like the 89410a or the 89441A instead
of the usual FFT analyzers? In comparison to the newer Stanfords or HP/Agilent FFT
analyzers they seem not to be costly, and the BW is 10 MHz vs. 50 KHz on many HPs
when used dual channel.
A source is also there, cross spectrum/waterfall/spectrogramm, everything available,
and the RF unit is not needed for AF people.
regards, Gerhard
of the usual FFT analyzers? In comparison to the newer Stanfords or HP/Agilent FFT
analyzers they seem not to be costly, and the BW is 10 MHz vs. 50 KHz on many HPs
when used dual channel.
A source is also there, cross spectrum/waterfall/spectrogramm, everything available,
and the RF unit is not needed for AF people.
regards, Gerhard
Interesting option but they are $2K+ on eBay and will require some learning to use. The Omichron is an alternative VNA with similar coverage .
I can get the VNA functionality along with good post processing from Praxis. Once you understand the VNA well you can do really interesting things with one. I use an 8753 often so I'm no stranger to the concept, but its way beyond most audio hobbyists.
Sent from my Nexus 7 using Tapatalk
I can get the VNA functionality along with good post processing from Praxis. Once you understand the VNA well you can do really interesting things with one. I use an 8753 often so I'm no stranger to the concept, but its way beyond most audio hobbyists.
Sent from my Nexus 7 using Tapatalk
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.