08-Feb-2019, 12:58
OK, I accept the 2 posts above but I still think the following things need to be considered:
1- Soniclife: you said the voltage when the bug occurred was 237 V. Was that immediately before the shutdown or when you checked the voltage immediately after the shutdown. If it was after the shutdown, that doesn't tell us what the voltage was when the shutdown was triggered. There could have been a brief spike on the line which triggered it and the voltage dropped back to within acceptable limits immediately after the spike and before you got the chance to check the voltage. The problem with knowing the voltage at the moment the shutdown is triggered is that we usually are not watching the voltage when we're listening to music, we may check it every now and then for some reason. I used to check my voltage every now and then after getting my P12 just to see what was happening on the line, what the distortion on the incoming line was and what the distortion was on the P12's output just to get a feel for what my incoming power was like and how the P12 was working but after a week or so I stopped and now check only very rarely. Most of the time now I have the display turned off because it's distracting.
2- Pim: I don't think a dropout in one channel of a dual mono setup necessarily proves the cause is a bug. No 2 amps are ever going to be perfectly identical. The protection may be designed to trigger at a given voltage but do both amps measure identical voltages? Take 2 different measuring instruments and measure the same thing simultaneously and you will usually see a slightly different reading, simply because of individual variations in the instruments. The tighter the tolerances for the instruments are, the smaller the variation but they still may not measure perfectly identically. One amp may measure x V and the other may measure x+1 V. The extra volt that one measures may be the difference between it triggering and the other not. I used to work in health and safety where some things needed to get measured and the reliability of measurements could be a problem since any measurement which could be made always had a specified margin for error. Totally accurate and identical measurement by 2 different instruments is extremely difficult to obtain.
3- All of this discussion is based on an assumption that there are only 2 possibilities, that the fault is being triggered by the firmware bug or that it's being triggered by an over voltage condition. There is always a 3rd possibility, that it is being triggered by some other problem unrelated to the bug and other than an over voltage condition. We know that the bug was causing the problem for a lot of people at one point in time and we know that the shutdown with this error message occurs much less frequently now. That tends to indicate that the bug was fixed but it may not have been fixed perfectly so it still occurs though much less frequently, or another problem might cause the shutdown every now and then in the absence of an over voltage condition.
As I said, it's hard for us to know that when the shutdown occurs, there isn't an over voltage condition because we would need to know what the voltage is at the moment of shutdown to be able to say that it was or was not an over voltage condition causing it. Checking the voltage a few seconds later doesn't tell us what the voltage was at the moment the shutdown was triggered. If the voltage is high a few seconds after the shutdown, then it's probably over voltage but if it isn't high, then it may have been a brief spike in voltage, now over, that triggered it. Knowing for certain that it wasn't over voltage is difficult. I think we have good reason to believe that something was done in the new firmware for the CI board to solve the bug that Devialet acknowledged because people report the problem far less frequently now. That doesn't mean that the bug was perfectly fixed or that there isn't some other fault that occasionally triggers the voltage protection function incorrectly. Pim has to be right when he says that "It's something inside the amp" but the whole voltage protection circuit and firmware is inside the amp, His 2 amps may not act absolutely identically because of slight differences, even though both are within specification because specifications are always stated as being within a certain +/- range in which case the shutdown may trigger in one amp but not the other because of over voltage on the line so while the cause in his case can genuinely be said to be inside the amp, that doesn't necessarily mean that there's something wrong with the amp involved. Then again, the original bug may not have been fixed perfectly and still can cause problems o occasionally, or there could be another and yet unidentified problem which occasionally causes problems and in both those cases the cause is inside the amp.
I'm not saying that both of you are wrong and that it isn't the bug that was acknowledged. I'm simply saying that none of us can prove that the cause was the bug because there are other possibilities and we simply don't have the kind of evidence we need to say with certainty that it's one thing or another. I think we can say, given the much lower frequency of reports in recent months, that Devialet did do something to fix the bug they acknowledged, whether or not they managed to eliminate it completely.
Finally, Pim, the table in your link really shows up one aspect of the problem for manufacturers. Look at all of the differences across states and the Australian standard. There is no uniformity in what is permitted within Australia. I don't know what level Devialet set for the trigger voltage for the protection but it does not have to be the over voltage limit shown for any state in that chart. They're going to set it at a level they think it should be set at to protect their product and that could be 250 V or 260 V or anything they like above the nominal voltage. I think it has to be something above 250V but I don't know how much higher Devialet think is acceptable. We've got a national standard in Australia and every state has their own standard which differs from the Australian standard to some degree. Every country has its own standard and you can bet they all don't agree. The Australian standard was intended to provide uniformity with our trading partners but none of our systems implement it correctly and you can bet there are variations in standard across those trading partners our Australian standard was designed to be uniform with. You tell me what a manufacturer should set for an over voltage limit to protect their products for countries with a nominal 220/230/240 V power grid.
1- Soniclife: you said the voltage when the bug occurred was 237 V. Was that immediately before the shutdown or when you checked the voltage immediately after the shutdown. If it was after the shutdown, that doesn't tell us what the voltage was when the shutdown was triggered. There could have been a brief spike on the line which triggered it and the voltage dropped back to within acceptable limits immediately after the spike and before you got the chance to check the voltage. The problem with knowing the voltage at the moment the shutdown is triggered is that we usually are not watching the voltage when we're listening to music, we may check it every now and then for some reason. I used to check my voltage every now and then after getting my P12 just to see what was happening on the line, what the distortion on the incoming line was and what the distortion was on the P12's output just to get a feel for what my incoming power was like and how the P12 was working but after a week or so I stopped and now check only very rarely. Most of the time now I have the display turned off because it's distracting.
2- Pim: I don't think a dropout in one channel of a dual mono setup necessarily proves the cause is a bug. No 2 amps are ever going to be perfectly identical. The protection may be designed to trigger at a given voltage but do both amps measure identical voltages? Take 2 different measuring instruments and measure the same thing simultaneously and you will usually see a slightly different reading, simply because of individual variations in the instruments. The tighter the tolerances for the instruments are, the smaller the variation but they still may not measure perfectly identically. One amp may measure x V and the other may measure x+1 V. The extra volt that one measures may be the difference between it triggering and the other not. I used to work in health and safety where some things needed to get measured and the reliability of measurements could be a problem since any measurement which could be made always had a specified margin for error. Totally accurate and identical measurement by 2 different instruments is extremely difficult to obtain.
3- All of this discussion is based on an assumption that there are only 2 possibilities, that the fault is being triggered by the firmware bug or that it's being triggered by an over voltage condition. There is always a 3rd possibility, that it is being triggered by some other problem unrelated to the bug and other than an over voltage condition. We know that the bug was causing the problem for a lot of people at one point in time and we know that the shutdown with this error message occurs much less frequently now. That tends to indicate that the bug was fixed but it may not have been fixed perfectly so it still occurs though much less frequently, or another problem might cause the shutdown every now and then in the absence of an over voltage condition.
As I said, it's hard for us to know that when the shutdown occurs, there isn't an over voltage condition because we would need to know what the voltage is at the moment of shutdown to be able to say that it was or was not an over voltage condition causing it. Checking the voltage a few seconds later doesn't tell us what the voltage was at the moment the shutdown was triggered. If the voltage is high a few seconds after the shutdown, then it's probably over voltage but if it isn't high, then it may have been a brief spike in voltage, now over, that triggered it. Knowing for certain that it wasn't over voltage is difficult. I think we have good reason to believe that something was done in the new firmware for the CI board to solve the bug that Devialet acknowledged because people report the problem far less frequently now. That doesn't mean that the bug was perfectly fixed or that there isn't some other fault that occasionally triggers the voltage protection function incorrectly. Pim has to be right when he says that "It's something inside the amp" but the whole voltage protection circuit and firmware is inside the amp, His 2 amps may not act absolutely identically because of slight differences, even though both are within specification because specifications are always stated as being within a certain +/- range in which case the shutdown may trigger in one amp but not the other because of over voltage on the line so while the cause in his case can genuinely be said to be inside the amp, that doesn't necessarily mean that there's something wrong with the amp involved. Then again, the original bug may not have been fixed perfectly and still can cause problems o occasionally, or there could be another and yet unidentified problem which occasionally causes problems and in both those cases the cause is inside the amp.
I'm not saying that both of you are wrong and that it isn't the bug that was acknowledged. I'm simply saying that none of us can prove that the cause was the bug because there are other possibilities and we simply don't have the kind of evidence we need to say with certainty that it's one thing or another. I think we can say, given the much lower frequency of reports in recent months, that Devialet did do something to fix the bug they acknowledged, whether or not they managed to eliminate it completely.
Finally, Pim, the table in your link really shows up one aspect of the problem for manufacturers. Look at all of the differences across states and the Australian standard. There is no uniformity in what is permitted within Australia. I don't know what level Devialet set for the trigger voltage for the protection but it does not have to be the over voltage limit shown for any state in that chart. They're going to set it at a level they think it should be set at to protect their product and that could be 250 V or 260 V or anything they like above the nominal voltage. I think it has to be something above 250V but I don't know how much higher Devialet think is acceptable. We've got a national standard in Australia and every state has their own standard which differs from the Australian standard to some degree. Every country has its own standard and you can bet they all don't agree. The Australian standard was intended to provide uniformity with our trading partners but none of our systems implement it correctly and you can bet there are variations in standard across those trading partners our Australian standard was designed to be uniform with. You tell me what a manufacturer should set for an over voltage limit to protect their products for countries with a nominal 220/230/240 V power grid.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Shunyata Alpha V2 speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.
Brisbane, Qld, Australia
