Products of this sophistication are rarely flawless for all use cases when initially deployed. I'm confident that important issues which arise will be addressed equitably and effectively.
There is no "magic bullet". Additional time, effort and expense are inevitable.
There is no "magic bullet". Additional time, effort and expense are inevitable.
Good question! Apparently it wasn't tested sufficiently in this respect. 😱
I put a lot of focus on the low level performance, making sure that the noise will be at a very low level and that distortion will be very low etc. The attenuator for the higher levels was not tested as much.
The problem that has been discovered now is a crosstalk issue between the overvoltage protection circuit (mainly some comparators) and the nearby attenuator. I am now working on a solution (a small add-on circuit), which will only enable the overvoltage detection when it is needed, that is, when selecting 10 dBV or lower input level settings.
ARTA can do multitone tests. I also did it with audioTester once, but I cannot recommend that part of audioTester.
Will you hold the units that have not yet been delivered and add this fix before sending them out?
Can this fix be performed ourselves (assuming reasonable DIY and SMD soldering capability)?
I'm expecting my unit next Monday, and would very much like to avoid sending it back to Denmark from Asia.
How do you track it?
I can't track it with the code provided in the email sent by RTX.
@mbrennwa
That would be my preference. I have designed a small PCB by now, but I need to test it, hopefully tomorrow, and check the logistics in this in the next couple of days.
@cwtim01
I think that should be possible.
Per channel you need to remove two resistors (0603), attach a small PCB on top of an IC (SO14) and solder 5 wires from the small PCB to the Main Board, e.g. using enamelled copper wire.
B&W_arthur
Once your unit has shipped you should receive a track & trace number. Use that at a DHL track & trace page.
That would be my preference. I have designed a small PCB by now, but I need to test it, hopefully tomorrow, and check the logistics in this in the next couple of days.
@cwtim01
I think that should be possible.
Per channel you need to remove two resistors (0603), attach a small PCB on top of an IC (SO14) and solder 5 wires from the small PCB to the Main Board, e.g. using enamelled copper wire.
B&W_arthur
Once your unit has shipped you should receive a track & trace number. Use that at a DHL track & trace page.
I received a mail with a DHL tracking ID "AWB+10_digit_number" that doesn't work, but tracking works if I remove the AWB prefix.
Hi Jens,
Further on, do you plan to redesign the motherboard in order to fix these issues?
No problems for those (such as Demian) who beta-tested your analyzer?
Further on, do you plan to redesign the motherboard in order to fix these issues?
No problems for those (such as Demian) who beta-tested your analyzer?
Hi Jens,
That sounds like a decent plan. Of course, we will have to open the case to install the board. The small package can be posted as the analyser will work, just not at higher voltages. Inconvenient for me, but certainly not the end of the world.
What you want to avoid is having all these come back and be shipped again if possible.
-Chris
That sounds like a decent plan. Of course, we will have to open the case to install the board. The small package can be posted as the analyser will work, just not at higher voltages. Inconvenient for me, but certainly not the end of the world.
What you want to avoid is having all these come back and be shipped again if possible.
-Chris
It will not affect the performance at low levels. But it will remove the internally generated distortion (crosstalk) at high levels.
The motherboard will of course be redesigned for future production.
Demian hasn't complained about it, but I don't know if he tested at high levels. As long as you work at line levels there is no issue.
Mine was delayed because I went on a trip just as it was about to be shipped so mine can be reworked at the factory if that is preferable. I'm capable of doing the retrofit if required, but it makes sense to wait a bit I'd say.
Even at high levels it is not a big deal for those of use who work with tube amplifiers as the residuals are at least 20 - 30dB below what is possible for most good tube amps at even modest power levels.
I designed semiconductor ATE for a living (including a fairly high performance AC test instrument) so if you want a second set of eyes to take look at the root cause/proposed fix feel free to PM me. (Or not as you prefer)
Thanks for answer. I can feel your pain. I mean after designing such masterpiece a bug like this is discovered.
When a redesigned PCB is ready I'm interested. Of course the 'patch' is ok meantime.
Not speaking for Jens here, but I wouldn't expect a replacement motherboard for an issue like this typically regardless of the situation, and in this case particularly in view of the astronomical discount we were offered in this group buy. (We're paying < 40% of MSRP)
Early production of a device as complex as this is likely to have a minor bug or two, if you weren't willing to take that risk the GB probably wasn't the best idea for you IMO.
Early production of a device as complex as this is likely to have a minor bug or two, if you weren't willing to take that risk the GB probably wasn't the best idea for you IMO.