Grestorn

I wouldn't have believed it, but you're right. I just reverted my system to stock clockings. And the value is now reasonable (about 16.5 W in idle, 130W with Prime95). Very curious. Unfortunately I cannot try HWMonitor since that doesn't display the CPU Package power value in either the free version or the trial of the Pro version. And I don't want to buy it (Aida64 provides all I need for system monitoring ). Anyway, I can live with the issue, especially since I'm pretty sure that you can't do anything about it, as it's most likely a problem Intel would have to solve. About the sensor issue, I'll repeat here what I wrote you in the PM as well, for others to read: I understand your problem. I wouldn’t mind that at all, if the displays would actually appear as soon as the device starts to report something else than n/a or 0. But as it is right now, if the fans are stopped when AIDA64 is starting (which is the usual case) I will never see these values. I always have to restart AIDA once the fans start to work. I guess you can see whether aquaero service is present in the system. My suggestion would be, just to enable all the sensors once you see the data source, no matter whether the actual sensors provide useful data or not (interpreting n/a just as 0). Maybe that saves you some work and hassle? People who have an aquaero probably don’t mind if there’s a huge number of useless values displayed for them…

1) See the screen shots. The first shot is with the fans inactive. Note that the gauges are missing. Also the text display for their Voltage and Current is missing. The second shot shows the same panel with the fans working. I had to reset the display after I the fans started to work to get the display to work correctly. The corresponding panel definition is also attached. 2) Makes sense. 3) Also makes sense. 4) It's an Intel 5960X on an Asus Rampage V. You can see the CPU Package value at the most bottom right on the screenshots. Do you need the sensor debug output as well? I used Prime95 to stress the CPU, so a value of 3.0W times 100 actually makes sense on the heavily OCed 8 core CPU. 2014-12-06.zip

Hi all, here's another short list of problems I currently experience with trying to create an elaborate panel for my iPhone (using Arx Control): All sensors that currently don't provide any data don't render on Arx Control. It doesn't matter what kind of control used, all are affected. To fix that you have to reset the Arx Control display (by changing the layout for example) once the sensor delivers any data. Some items are displayed differently on the preview paage and the actual display. For example, if the total width of a sensor item is to small, it will be rendered in two lines on the Arx display while on the preview page it's still rendered in one line just using the width it needs. The default font "Tahoma" doesn't render correctly on the Arx Display. It looks like a font with serifs (probably some Times Roman). Other fonts, like Verdana are working though The value for "Power Values / CPU Package" seems to be off by a factor of 100 - or even more. If it's supposed to be the total power used by the CPU Package. I know it's hard to guess what Intel actually reflects in these values, but I get around 0.6 Watt in idle and 1.5 Watt under load. 10x that value sound about right (if still a bit low, though, since I use an overclocked 5960X). Especially the first issue is something that bothers me a lot right now. Item 2 and 3 can be worked around and the 4th item is just a matter of interpreting the data. Thanks & regards, Martin

I've got problems matching the RAM SPD information displayed by AIDA64 with the one shown by CPU-Z or the BIOS. Here's what AIDA shows: And this is CPU-Z output, the BIOS shows the same values: Especially the timings for XMP 2800 and 3000 seem to be different in AIDA... Am I reading the values wrong somehow?

1) Nice 2) Well, it's just an alternate way to display. I know that it's always a compromise between implementation effort and gain. But some might like to have something like a gauge with hands, like on a classic car dash board. It's not really important to me, to be honest, but now that you got the basic routine, I could think of many different ways to display the gauge 3) ... oh my. You'll learn something new every day. Thanks

I'd like to suggest a few small improvements for the LCD Panels: I like the new Arc Gauges available for the HTML Server and Arx Control panels. However, an easy to implement improvement would be to add a setting for the start angle. I.e. if you want the arc to start in the lower left area of the circle, you'd input something like 220 degrees (assuming 0 degree is the top spot and you count clockwise). Also, alternatively to the fully drawn arc, you could consider an option to render a small triangle, something like this: It would be great if you could change the orientation of the sensor bars to vertical. I realize that you can simulate that using a graph using a histogram display. But you cannot use seperate colors depending on the value that way and the sensor bar looks better, too. Thanks a lot and keep up the great work!

Stopped fans are not rendered, at least not as Arc Gauges. They don't render in the LCD setup screen either. Even when the fans start to spin, they're not rendered unless I make the Arx Control to reset completely (by updating the layout or restarting AIDA64). However, if they're rendered in the first place, they will remain visible, even if the fan stops later. Sorry about the pump frequencies, I now remember that you already stated that they're only shown as RPM...

Updated, so far it works. If I encounter any problems I will let you know. By the way, I wanted to add Water Quality and Pump Frequency and realized, that these values are not available for LCD panels. Is that intended? And another thing: As discussed in another thread, fan controls are not rendered if the value is zero. That problem only exists for fans, not for flow values (even though most of them are also 0). Is there no way to avoid that?

Yes, it's 3244. And Auto Update is also enabled, so I'm usually using the latest version.

It's an iPad 3. Funny thing is that the display was fine and when I looked at the display after a while, it was corrupted. I don't think that I changed the layout in between. But who knows...

I'm in the progress to create a complex sensor panel for the Arx Control display. One recurring problem is that after a while the sensor inputs are mixed up. Meaning, all of the controls show wrong values which are obviously meant for other controls. See the following screenshot and the accompanied export file. As you can see, the value for the utilization of the GPUs are completely off (showing some fraction). All of the other values are off as well, but it's most obvious with them. It looks like the assignment for the sensor inputs are somehow moved around or randomized. 2014-11-30.zip

Well, you see, when my systems idles, my water temperature floats between 36.0 and 36.7°. In that range, the fans in the aquaduct are turned off. When I start to put some stress on it, it rises above 37°, at that temperature the fans are starting or work. To find the right spot when they should start is vital, because you don't want the fans to be on while the PC is idle. The aquasuite displays that you can put on your desktop show that the temperature measured is very stable and dependent. I don't think that it's actually really accurate, but that's not important. I don't need to know whether it's actually exactly 37.6°, but I want to to know when the temperature is starting to rise or fall, even if it's just by 0.5° or less. And the changes in the temperature are reflected well enough by all of their sensors, even the ones I've put in the water cycle myself. I understand your points and I agree that for most use cases, an accuracy of more than 1 degree is over the top. But unfortunately, water cooling system are an exception to that rule, I'm afraid. Again, I don't want to impose any additional work on your part, so I would be happy if the temperature would be shown in 10th of a degree, i.e. 375 instead of 37.5. I guess that would be a very minor change in your data collection routine only, wouldn't it?

Hi Fiery, I've been playing around with the new Arx Control feature to make me a all new panel for my "old" iPad. While doing that I realized, that the temperature sensor for the aquaduct using integer numbers is just not adequate. You have to understand, that a temperature in a water cooling cycle shows much less deviation than a CPU, GPU or even a motherboard temperature. My "Water inlet" temperature usually varies between 35 and 39°C. That's why AquaComputer's has a resolution of 0.1°C. I understand that it might be difficult for you to add fraction support for temperatures all through your framework, but I think it would be worthwhile. If that's too much to ask, I'd rather have the aquacomputer temperature values reflect the 10x value, i.E. showing 37.5° as 375.

If I didn't miss anything, it doesn't seem to be possible to control two different devices over Arx Control, is it? Like a tablet and a phone?

Yeah! I was so looking forward to this! Now I have to decide whether to use ArxControl or or your own HTML Server solution. With what are you more comfortable with?

Thanks for the info!

Looking nice. Why doesn't everybody share the source file though?

I've been trying to see the effect of Intel's Turbo Boost technology. On my 5960X, I can see in AIDA64, that the CPU clocks to a factor of 35 whenever under load. It never goes below that (since I'm using a watercooling, temperature is never an issue). When using OC, it goes to a factor of 43. If I understand Intel Turbo Boost correctly, it means that it can get an even higher multiplicator for a short amount of time. And when using the Intel Extreme Tuning Utilitty, this is exactly what I'm seeing: When using the Benchmark mode of Prime95, the "Processor Frequency" as displayed by Intel's utility boosts up to 36 (44 with OC) for short amount of times. When I'm increasing the "Turbo Boost Power Time Window" this time can be many seconds long. Still, in ADIA64 I'm not able to see that effect at all. Neither the frequency display nor the CPU multiplier ever changes to reflect the Turbo Boost. Is that a technical limitation? Am I doing something wrong? And, while asking this, the CPU frequency displayed in Win 8.1 seems to be off. It's supposed to be 100 MHz (FSB), but shows 98.1 MHz. Under load it even drops below 96 MHz. The overall CPU frequency also is shown as too low as a result. In Windows 7 the display seems to be stable though. Is this an effect of the changed way the RTC is used in Win 8, which has been noted to cause problems with certain benchmarks? Is there no way around this effect?

The value for the case fans can become erratic when the system is under heavy load. Maybe that's related to how the value is calculated. Is there something that can be done about it? Right now this makes the value quite useless, since I usually need to see it under heavy load.

In fact, I do. It's the official Aquacomputer UV active coolant. But the quality measurement is quite unstable, sometimes it's higher, sometimes lower.

Thanks a lot Fiery! All your changes seem to work fine. And nice new icons for the flow and level sensors Here's the sensor dump: Pump is working at 75.0 Hz, the pump's current is floating between 450 and 475 mA, it runs at 4500 rpm, The water quality is listed as 30%. The Aquaero CPU's temperature is listed as 44,0°C (I think that's missing from your list).

The Aquaduct only shows one level sensor, I guess the other one is for a different kind of hardware. So I can't tell you the actual value of the second level sensor. I also can only use one flow meter and up to 6 temperature sensors. The water flow is measured in liters per hour, yes, and you're showing the correct amount for the one flow sensor I have. The water level is listed in percent, 100% meaning it's filled up to the brim. BTW, now the level sensor shows "80%" on the aquaduct while you list "7954%" which would confirm the factor 100. And, yes, it's pretty normal that the level meter has a huge variation, I didn't put any additional water into the device. And there's one sensor missing from your listings: "Water Quality". At least I didn't see it.

You can use the product indefinitely, but you won't be able to get any updates.

Just tested the sensors on my Aquaduct Mk IV. Everything seems to work fine with the exception of the LiquidLevels. Both are completely off. Liquid Level 1 is reported as 7096%, Level 2 as 215% The actual value as reported by the Aquaduct is 71%. So my guess is that you're reporting 100x the actual value. Here's the sensor debug output:

You're the best. I mean it! Thanks a lot!