|
|
Post by dmitry871 on Oct 16, 2023 12:15:56 GMT -5
Dear LakeShore team! Why do you supply Cernox sensors in CU (CU-HT) package with WQL wire while you are reccommend to use twisted wire in manuals for 335 and 336 temperature controllers (see attachment)? Meanwhile diode sensors in CU package are supplied with WQT wire. Why? Best regards, Dmitry  |
|
|
|
Post by Lake Shore Ryan on Oct 18, 2023 8:11:28 GMT -5
Hi Dmitry, Good catch. I can see how this could be interpreted this way. As with most things, there is nuance to recommendations. Here is my explanation for why we made these product choices. In general, we find that noise inside the cryogenic environment is minimized due to it being inside a grounded metallic enclosure. We found that poor thermal anchoring is the greatest source of measurement error in user systems, so we chose to use ribbon cable to maximize thermal contact when wrapping the wire around a thermal anchor. The ribbon cable we use on temperature sensors with maximum temperatures of 420 K or less uses a bonding agent to create the ribbon cable. This bonding agent would melt before reaching the maximum rated temperature of the diode (500 K), so we instead use twisted pairs for this CU-HT package. This wire is rated for use to 500K. Also, diodes aren’t expected to go as cold as Cernox sensors, so the best possible thermal anchoring isn’t as critical as if it was a Cernox sensor attached to something reading below 1 K. We felt this was a reasonable trade-off. Outside of the cryostat, we absolutely recommend using shielded twisted pairs for all sensor types regardless of the wire inside the cryostat. Hopefully this helps explain our seemingly inconsistent product offering. If your application would benefit from twisted pair wiring on the Cernox CU-HT package, there are several options: - You can cut the wire right back to within a few centimeters of the sensor and attach your own twisted pair wiring. This won’t change the sensor calibration curve as it is wired in a 4-lead configuration.
- We can customize the sensor for you for an additional cost. This usually becomes economical when dealing with many sensors as the cost of customization is amortized over many sensors. You can request more information about this by contacting your local sales representative.
Thanks for your post Dmitry.
|
|
|
|
Post by dmitry871 on Oct 26, 2023 4:30:38 GMT -5
Hi Ryan,
Thanks a lot for detailed explanation! The reason why you use WQL wires become more clear to me. However it is still interesting to compare performance of WQL vs WQT wires in the same conditions, i.e. the same vacuum chamber, the same environment conditions, the same sensor measuring the same temperature, let say 4K. Do you expect to observe any noticeable difference in measurements (or measurement noise) using 336 temperature controller when using WQL or WQT wires inside vacuum chamber?
Another question I'd like to ask you is about shielding of measurement wires. 336 manual says: "One shield should be connected to the cryostat’s ground as long as it is near earth ground." What does it mean "near earth ground"? Should I connect vacuum chamber to the same earth ground as temperature controller (i.e. ground wire in the power cord)?
Vacuum pump, vacuum gauge and GM-cryocooler is usually connected to a vacuum chamber. Well it is possible to electrically isolate vacuum pump and vacuum gauge but it is not possible to isolate GM-cryocooler which can introduce some noise in the shielding of measurement wires if the shield is connected to vacuum chamber. What do you recommend in this case? To which point is is better to connect shield in case of using 336 temperature controller?
I also looked through 372 manual and it looks like there is more complicated shielding scheme due to using external scanners. Would you recommend something else in case using 372 resistance bridge?
Best regards,
Dmitry
|
|
|
|
Post by Lake Shore Ryan on Oct 26, 2023 15:35:47 GMT -5
Hi Dmitry,
Honestly, it's hard to say with the wire. If you have a lot of RF noise sources in your lab (like large motors) beyond what is present in a traditional cryogenic system (things like turbopumps aren't that noisy), then you will be fine with ribbon wire. The added thermal connection this wire provides to heat sink bobbins will be worth it.
The grounding issue is a deep topic. By suggesting to connect shields to the cryostat only, we're trying to avoid ground loops, which result in 50/60Hz signals flowing along the shields and potentially coupling into the sensor wires. Suggestions for both the 336 and 372 although they may use slightly different approaches, are all trying to avoid this issue. The instrument will be connected to an electrical ground, which may be at a different potential to the cryostat, which will likely be connected to a different earth ground. For your shield to be effective, it must be connected to one of these grounds. The cryostat is what we suggest. The main takeaway though should be to never connect both ends of the shield on the cable. If you connect one end to the cryostat, and the other end to the instrument ground, you could end up with a worse situation than no shield at all. Current will flow between the two potentials, likely with a 50/60 Hz component from the power lines which can couple into the signal wires.
Hope this helps.
|
|
