I've experienced this same issue with a bunch of Arduino type boards. The issue comes and goes with Windows updates. Are all of your motherboard drivers up to date?
I was talking about your motherboard in your computer. USB controller drivers can cause issues.
I've had the same issue with Win10, but sometimes the unit will just stop working between tests. Not often but enough so you really remember it. Unplugging the USB is a PITA in my case as I have to go around the basement to reach the rear of the bench to do it. Everything is in a rack. I think it is a windows problem because cycling the power on the RTX doesn't solve it. I am running v1.19 and have been since it came out. Same issues.
Powering down the RTX won't shut down the USB interface and the XMOS since it gets power from the host. My guess is that the XMOS sends something when powered off that tells the PC that its disconnected however it never does the initial handshake when powered back up because the USB VBUSS never went away. You could make a USB cable where VBUSS comes from the RTX. It may provide VBUSS internally as well so simply cutting the Vbuss wire would fix the problem. Looking at the enumerated devices on the PC when its "locked up" under devices may give some clues.
Windows 10 has had some recent updates to USB drivers that also indirectly impact UAC2 driver. I noticed these when the Windows UAC2 started behaving differently to the async feedback coming from my USBI2S board. The worrying thing is that Microsoft UAC2 experts were apparently not even aware of these updates and their impact to UAC2.
Can you share what changed specifically? Thanks a lot.
I wonder if MS actually has any UAC2 experts, since their UAC2 driver was developed by Thesycon https://docs.microsoft.com/en-us/wi...udio/usb-2-0-audio-drivers#driver-development .
I don't know what exactly was the update to the USB drivers but after that UAC2 responded to async feedback more aggressively which caused buffer over/underruns on my board as I was using quite small buffers. Slight change to the feedback scheme and increase to buffer sizes mitigated the issue. Update involved several usbxxxx.sys files in \windows\system32\drivers but usbaudio2 drivers were not updated. My hunch is that the UAC2 driver uses some lower level USB functionality for e.g. async feedback handling as that is not only for UAC2.
IIUC the Windows UAC2 driver was developed by Thesycon but is now maintained by Microsoft. But as said Microsoft's UAC2 experts were not even aware that there had been updates to USB.
IIUC the Windows UAC2 driver was developed by Thesycon but is now maintained by Microsoft. But as said Microsoft's UAC2 experts were not even aware that there had been updates to USB.
Last edited:
I need to do some measurements with more than 20 dBV (24-26 dBV) out from the generator.
Can I possibly series connect the two outputs to achieve this without risking to damage the instrument?
Thanks
Can I possibly series connect the two outputs to achieve this without risking to damage the instrument?
Thanks
Nic,
This is not durable as two AC voltage generators in parallel does not double the voltage. And if the AC phases are not same it becomes even messier. Use an external stereo measurement (line) amplifier for this purpose instead.
This is not durable as two AC voltage generators in parallel does not double the voltage. And if the AC phases are not same it becomes even messier. Use an external stereo measurement (line) amplifier for this purpose instead.
Yes Nic,
Standard lab practice is to use a fixed gain amplifier to provide higher signal levels if needed. However you can use any amplifier in this case. Never, ever connect outputs together on anything!
Standard lab practice is to use a fixed gain amplifier to provide higher signal levels if needed. However you can use any amplifier in this case. Never, ever connect outputs together on anything!
I guess an 1:2 audio output transformer could also do the trick if you only need up to 6dB boost and you don't care too much about extended bandwidth.
Don't use audio transformers to combine outputs in any way. Same issue. Voltage generators are ideally zero ohm outputs that never can be combined.
Use a line amp of some sort
Kurt
Use a line amp of some sort
Kurt
Sure, however the primary impedance must be high enough to avoid too high currents from the output.
A 600 to 1200 or 300 to 600 may work. Consult the load data of the Analyzer
A 600 to 1200 or 300 to 600 may work. Consult the load data of the Analyzer
Op amps have current limiting and the signal levels are not that high considering power. You will get distortion long before damage. So it comes down to the load impedance that is reflected into the generator really.
Most amplifiers have a gain between 29 to 35 dB. It might be just as easy to use a consumer amplifier for this job. I do have +10 and +20 dB RF amplifier modules (connectorized) for RF work.
Most amplifiers have a gain between 29 to 35 dB. It might be just as easy to use a consumer amplifier for this job. I do have +10 and +20 dB RF amplifier modules (connectorized) for RF work.
Really we just need turns ration unless we need to calculate load presented to the generator.
I don't think you will damage the RTX6001, but you won't achieve your goal either. The two channels are connected, and share a common ground, so you would basically just short the two outputs with opposite phase. Since the output impedance of he two outputs are equal, it would be like shorting the connection point to ground.
The single ended output impedance is 50 ohm by the way. Balanced output impedance is 100 ohm.