Hello,
I'm about to buy either 3 Picoscopes or 3 PeakTech PC based oscilloscopes. They will be used to instrument experimental inverters within an enclosed area kilometers away from any human. The data link between the test site and humans will be done using Ethernet over optical fibers. Once in the test site, no human will be able to access them for 2 years.
I'm hesistating between the two options :
- 3 x Picoscope 5000 + Linux PC within the enclosure and then access them using remote desktop
- 3 x PeakTech and a remote windows PC
They will be placed in a harsh environment, so I'm concerned about CEM and filtering the power supply of these scopes.
They will be used with high voltage probes to measure the output transients of high voltage frequency converters.
I wanted to have you guys opinion about these two products, their reliability and your opinion about the software.
Any advices ?
Regards
David
This sounds big budget, so why choose? Get both setups for redundancy.
I have the fully specced picoscope 3000. Although they are a bit pricey, they are definitely good units. I have never used the windows software, but it is supposed to have more features than the linux version. So make sure that the linux version supports what you want to do. I don't know about the 5000, but the 3000 is very rugged physically. It looks like the unit Dave is playing with in this video: is well shielded.
Their SDK is pretty good too; I have written a python script which interfaces to their C library which streamed at 125MS/s for >10 minutes straight with no issues.
For me, this wouldn't be a question at all.
PicoTech is a well known reputable British company that stands behind their products.
PeakTech on the other hand just rebadge all kind of cheap gear from various sources and sell it at a premium. Certainly not what professionals are dreaming of...
Thanks for all your replies !
I'll always lean towards reliability and I like PicoScope as well. It's just that if I could avoid USB that would be nice. Can't wait for Pico Tech to build Ethernet devices.
I am dubious about the reliability of operating instruments and PCs for long durations due to increasingly poor programming. (1) I would include a provision to reset the oscilloscopes and PC remotely by temporarily removing AC power. It might be better to use oscilloscopes which are directly controllable over Ethernet to avoid having to include a remote PC.
Beware of heat dissipation issues when the instruments and PC are fully enclosed. I have done this several times and always had to add air circulation through the enclosure with a filter of course to keep dust out. If humidity is a potential problem, then controlling the air inlet temperature can lower the relative humidity inside preventing condensation.
(1) I would trust my ancient DSOs to operate for weeks to months at a time without problems because they all use static memory allocation in their programming. I would not trust a modern DSO to do so because it likely uses a heap or even worse, garbage collection. With care Linux might be acceptable and I know BSD is but forget Windows.
Thanks David Hess
Thanks for your suggestions regarding remote power shutdown/reset. I think that you are right when saying we should avoid PCs. I have a hard time finding the tradeoff between the reliabilty of a Linux fanless industrial computer and a PeakTek scope (because I know the picoscope is good hardware). The PC may be subject to strong EMF this is also why we should maybe avoid that.
The hardware will be in a large air tight enclosure filled with nitrogen to avoid condensation (many cubic meters). It will be cooled using air/water heat exchangers and the average temperature will be arround 18°C. I'm not worried about heat in there. I'm worried about EMF and I need fanless operation.
Like you I trust my DSO and would be happy to use it using LXI, this is why I'm so hesitant because it's something I've never been doing before and I have only one shot at it.
Regards,
Thanks for your suggestions regarding remote power shutdown/reset.
They make controllers to do this but I was always in a position to cut the power manually and have the PC restart when power is applied.
I think that you are right when saying we should avoid PCs. I have a hard time finding the tradeoff between the reliability of a Linux fanless industrial computer and a PeakTek scope (because I know the picoscope is good hardware). The PC may be subject to strong EMF this is also why we should maybe avoid that.
I am dubious that EMF could be strong enough to cause problems unless it got into the input power or the grounding scheme was poor. At worst some minor shielding should handle it. I would probably include a dedicated power filter before the power supplies to reduce conducted EMI. Use a solid single point ground.
A fanless industrial computer should be pretty reliable. Even consumer gear can be if derated and modified by for instance replacing fans with long life ones. Unfortunately it is difficult to rely on long term reliability without historic data which argues for making things either redundant or as simple as possible.
Make sure strain relief is used near all non-captive connectors like USB and SATA to prevent them from vibrating or working out. (1) Tie everything lose down solidly.
The hardware will be in a large air tight enclosure filled with nitrogen to avoid condensation (many cubic meters). It will be cooled using air/water heat exchangers and the average temperature will be arround 18°C. I'm not worried about heat in there. I'm worried about EMF and I need fanless operation.
So you have considered the heat dissipation issue and have it handled. It sounds like an intrinsically safe application in an explosive atmosphere where positive pressurization is used.
Honestly fans do not bother me if I get to pick the fans.
I would think nothing of selecting them to operate for years without maintenance.
Like you I trust my DSO and would be happy to use it using LXI, this is why I'm so hesitant because it's something I've never been doing before and I have only one shot at it.
(1) Non-captive USB and SATA are terrible in long term installations or where vibration is present. USB was literally intended for desktop applications and SATA for backplane applications.
Thanks again David Hess for your advices.
I found out that PeakTech is indeed rebranding hardware made by other manufacturers. For example the PeakTech 3440 is in fact a CEM DT-988 also branded as SouthWireTools 15190T.
For the PeakTech 1330, it looks like that it's a rebranded Owon VDS3104.
It looks like however that it is a descent scope :
https://www.youtube.com/watch?v=vO712DvGAKA, however Dave seems to have hard feelings about Owon hardware...
Thanks for the tip about USB plugs and vibration, I've seen that for higher end PC based scopes, they provide à screwed usb plug like the KeySight U2701A/U2702A.
Thanks for the tip about USB plugs and vibration, I've seen that for higher end PC based scopes, they provide à screwed usb plug like the KeySight U2701A/U2702A.
Oh really? I have not seen that. I tie the cable down near the USB (or SATA) plug so there is no force on it. This should also be done with IEC power cables which will otherwise work lose. Just temperature cycling over months is enough. You would think that they make an IEC power socket which a wire clasp to hold the plug in but I have not seen one.
Some SATA connectors now have a clasp which holds them captive but not all SATA sockets support them.