I have never had any proper training in electronics, engineering, or workshop manners. That's why I just "do stuff" when it comes to building and testing stuff in real life, and it also means I may do things differently than a trained electronics person would. Looking at signal spectra at different positions within an amp can be very useful to understand where and how signal distortion happens (yes, the feedback loops need to be disconnected first).
I think this is getting a bit off topic. Maybe start a new thread and let us know about it if you want to discuss this more.
Happy new year to everyone!
Thanks, that's a good idea! I have one of these isolators floating around my workshop...
@mbrennwa
With the RTX6001 Audio Analyzer this does not change the situation. The analyzer is already isolated from the PC. It is supposed to be connected to earth though, through the supply cable.
How do you measure the phase?
If you make a loop back connection, you will not have a phase of 0 between the (SW) generator in the PC and the (SW) analyzer in the PC. There will be a number of delays in-between, like buffers, USB interface, DAC and ADC.
With the RTX6001 Audio Analyzer this does not change the situation. The analyzer is already isolated from the PC. It is supposed to be connected to earth though, through the supply cable.
How do you measure the phase?
If you make a loop back connection, you will not have a phase of 0 between the (SW) generator in the PC and the (SW) analyzer in the PC. There will be a number of delays in-between, like buffers, USB interface, DAC and ADC.
The test signal used by REW is not a simple sine sweep. It is designed to measure not only frequency response, but also phase, impulse response, etc. Of course there will be some time delays, but they are removed by REW to make sure the results are not affected.
Paul
If you only want to check if the polarity is swapped somewhere use a test signal that looks different (modulo shift) if you swap polarity, e.g. a sawtooth.
Still waiting for my RTX6001 - is the current drawn at the RTX6001 USB port known ?
Ulli
FYI:
I did try Intona with a PicoScope 4227 - didn't work.
4227 draws about 450 mA from the USB port.
Even with a AQVOX USB Low-Noise 5V Isolated Linear Power Supply connected to the 4227 USB port it is not working properly 😀
Using a notebook running on battery is my current solution.
cf. How to make floating or isolated measurements and break the ground loop with a USB oscilloscope?
Still on my todo list:
WaveIO has a 5V input at the output side (lower right corner).
Might be possible to modify Intona accordingly.
4227 draws about 450 mA from the USB port.
Even with a AQVOX USB Low-Noise 5V Isolated Linear Power Supply connected to the 4227 USB port it is not working properly 😀
Using a notebook running on battery is my current solution.
cf. How to make floating or isolated measurements and break the ground loop with a USB oscilloscope?
Still on my todo list:
WaveIO has a 5V input at the output side (lower right corner).
Might be possible to modify Intona accordingly.
You are right, the phase is indeed wrong on the DAC 😱
It will be corrected in the next FW release (for the XMOS).
To be honest, mine didn't work in the first try, until I changed the USB cable for one that has decent supply cabling. USB cables vary a lot in wire size which is no issue for the data connection but sometimes makes a difference on the power side. Some of these wires are minuscule, especially in the thick cables ;-)
Jan
And there will be a software utility that allows us to perform the firmware upgrade I suppose?
@ JensH: On one of the pics of nyt (attached cropped) there are white lines, on the pcb, forming a cuboid and some ground (?) clips along these. Are these lines show the traces of a provisional of some kind of metal shield over the inputs circuitry which are not included in GB units?
Happy New Year to All!
Happy New Year to All!
Attachments
Last edited:
They are for input stage shielding. I've tried to install two SPTE shielding boxes, but the crosstalk is not reduced, which confirms what Jens has said. In addition, the crosstalk level does not change by switching to different input gains/attenuations. Therefore I think that it might be the crosstalk inside the ADC. Since these shields make no difference, I've removed them from my unit.
However, the measurements are usually done with single channel. Therefore I guess crosstalks should not be an issue for most cases.
However, the measurements are usually done with single channel. Therefore I guess crosstalks should not be an issue for most cases.
Attachments
Last edited:
@Ghianni
Yes, the clips were included to have the option of mounting shields over the input stages.
As described earlier, it was decided not to mount any shields.
@PaulBC
Very nice shields you have made!
As someone pointed out earlier, there is some crosstalk, primarily from the right generator output to the left input, due to the proximity. The shields will reduce this crosstalk. The ADC will cause some crosstalk between the input channels, which the shields will not change.
Some components in the input stages become relatively hot, so closed shields like the ones you have made are not recommended and will void the warranty.
Yes, the clips were included to have the option of mounting shields over the input stages.
As described earlier, it was decided not to mount any shields.
@PaulBC
Very nice shields you have made!
As someone pointed out earlier, there is some crosstalk, primarily from the right generator output to the left input, due to the proximity. The shields will reduce this crosstalk. The ADC will cause some crosstalk between the input channels, which the shields will not change.
Some components in the input stages become relatively hot, so closed shields like the ones you have made are not recommended and will void the warranty.
Hi Jen
one question on Single ended input - is it worthwhile building an electronic converter for RTX or cable solution ( earlier reference on this thread for suitable converter cable) is more than adequate ? I Was wondering whether electronic converter ( similar to AP sys series analyzers) as that can be using long length balanced cable from DUT will be better than cable? If so can you please suggest suitable schematics. Most of my use will be single ended connections from DUT.
one question on Single ended input - is it worthwhile building an electronic converter for RTX or cable solution ( earlier reference on this thread for suitable converter cable) is more than adequate ? I Was wondering whether electronic converter ( similar to AP sys series analyzers) as that can be using long length balanced cable from DUT will be better than cable? If so can you please suggest suitable schematics. Most of my use will be single ended connections from DUT.
Last edited: