Several times per month we are asked about the feasibility of using Marki Microwave products at cryogenic temperatures for various research applications. Many customers have purchased various products (bias tees, diplexers, filters, couplers, power dividers, etc) for cryogenic applications, and so far none have complained about the products failing. Now for the first time we have actual data to show that our products function at cryogenic temperatures. As we acquire more data about the performance of our products we will post it here. If you are interested in a product that you don’t see listed here and you are willing to perform the test and share the results for this tech note, please contact [email protected]. We will send you a part to keep.
Cryogenic testing of our first product (the FLP-0750) comes courtesy of Patrick Harrington and Prof. Kater Murch of Washington University in St. Louis, who were the first customers to take us up on the offer. See their complete report here. Read some fascinating work on quantum mechanics here.
This measurement shows the insertion loss of the FLP-0750 in a room temperature and cryogenic (50 mK) environment. The loss in the cryogenic chamber is de-embedded from fixturing with significant loss and ripple, so it is unknown how much of the ripple comes from the degradation of the filter and how much comes from the measurement system. It seems likely that the filter still successfully eliminates rejection band signals.
Bias Tees and Diplexers
Marki bias tees and diplexers have a relatively similar construction. These products have been used by numerous customers successfully at cryogenic temperatures. Most recently, Nathan Holman from the University of Wisconsin-Madison Department of Physics, has taken scattering parameter data of the BT-0024SMG as well as time domain data of the settling time of the DC input at 1.6K. It was shown that Marki Microwave bias tees (and by similarity, diplexers) will operate close to the datasheet specifications at cryogenic temperatures.
The plots above show that at 1.6K, the insertion loss was less than 3dB for frequencies above 8MHz and the DC port isolation is ~40dB which is within the datasheet specifications. Settling time was <100 μs for the DC input and it was shown that this bias tee was tolerant to repeated fast thermal cycling. For a more detailed look at the cryogenic performance of the BT-0024SMG, you can read the full report here.
Marki triplate stripline couplers are expected to work, based on the construction and performance over smaller temperature swings, as well as the feedback we have received from customers.
Our high directivity bridge couplers (CBR series) do not operate at cryogenic temperatures. The directivity degrades due to construction shifts across temperature, and some of the components may become superconducting. We have received measured feedback from Simon Zihlmann from CEA Grenoble, France on the CBR16-0003 which was characterized at both room temperature as well as at 8mK. The data shared with us characterized both the direct line insertion loss as well as the coupling of the coupler and showed significant degradation in performance when tested at 8mK. The measured data is provided below with the room temperature performance in blue and the 8mK performance in orange.
Marki DC blocks have been fully characterized at cryogenic temperatures by Simon Zihlmann from CEA Grenoble, France. In this series of tests, an insertion loss sweep was performed at room temperature and then the DC block was placed in a dilution refrigerator with a base temperature of 8mK and re-tested. Once the performance at low temperatures was characterized, the DC block was brought back up to room temperature and tested once more to investigate whether there were any irreversible effects of cooling the DC block to milli Kelvin. The results of these tests are shown below.
Marki Microwave’s MQH/MQS series of 90-degree hybrids have been used successfully in cryogenic applications with minimal degradation in performance. We received measured feedback from Philip Dindo from The National Radio Astronomy Observatory who was able to perform cryogenic measurements on the MQS-0418UA connectorized quadrature splitter module at 4K. In this series of tests, insertion loss, amplitude balance, phase balance, isolations, and return losses were measured at room temperature and then the MQS-0418UA was installed in a cryostat and cooled to 4K multiple times. The data collected by NRAO proves the operation of our MQH/MQS MMIC quad hybrids down to 4K. The measured cryogenic test data is shown plotted below:
The data above shows agreement with all maximum/minimum guaranteed specs in the datasheet for this quad splitter. The full report including details on the measurement setup, calibration, and measurement methods can be found here and the measured s-parameters at 4K can be downloaded here.
The PD, PD3, and PD4 series of Wilkinson power dividers are expected to work, though possibly with degraded isolation, down to very low temperatures. These products do use surface mount resistors that could become superconducting at very low temperatures, however. We have not received meaningful customer feedback about these products.
The PBR series of high isolation bridge combiners is NOT expected to function at cryogenic temperatures for the same reasons as the CBR bridge couplers, listed above.
Our higher frequency, banded, capacitively coupled baluns (including the BAL-0106, BAL-0212, BAL-0520, BAL-0208SMG, BAL-0416SMG, BAL-0620SMG, and all MBAL products) are expected to operate well at cryogenic frequencies, as they consist exclusively of coupled transmission lines. The rest of the balun product line uses magnetic cores. We recently received feedback from Fabio Sebastiano from TU Delft that these baluns will still operate at cryogenic temperatures, despite their magnetic cores as shown below.
In these tests, the BAL-0003SMG was immersed in liquid Helium and data was recorded at 300K and at 4K. The CMMR measurements extend beyond the 3GHz upper frequency limit that we spec this balun to, however within the 500kHz to 3GHz band we see that this balun behaves very close to the datasheet specifications for insertion loss, amplitude balance, and common mode rejection. These measurements confirm the operation of our baluns at cryogenic temperatures. This data also suggests that products with a similar construction (BAL-0006SMG,BAL-0009SMG, BALH-0003SMG, BALH-0006SMG, BALH-0009SMG, BAL-0012SSG, PBR-0003SMG, PBR-0006SMG, and PBR-0012SMG) should also operate well at temperatures down to 4K.
Mixers, Amplifiers, Multipliers, and other Nonlinear Products
We have not received any feedback from customers about these products, but they are not expected to operate at cryogenic temperatures due to the degradation of the semiconductor devices.