THE ARO 0.4mm ( GHz) SIS MIXER RECEIVER. Revision 1.0

Transcription

1 THE ARO 0.4mm ( GHz) SIS MIXER RECEIVER Revision 1.0 April, 2008

2 Table of Contents 1 System Overview Mixer Operation Setting the Mixer Voltage and Current Setting Vj: Setting Ij: Setting the Magnetic Field: Mixer Defluxing:... 7 Limitations to Cartridge De-Flux Heater Use... 7 Mixer Defluxing Procedure... 7

3 1 System Overview Those who may be familiar with the Atacama Large Millimeter Array (ALMA) Project, may recognize the band of operation for this receiver. This is because this receiver uses the same mixer used for ALMA band 9 ( GHz). These mixers have been provided to us by the Space Research Organization of the Netherlands who is responsible for the front end cartridge for band 9. This same mixer is used for various other receiver systems at other facilities such as the CHAMP + receiver for APEX and at the JCMT. 2 Mixer Operation The biasing of the mixers is relatively straightforward. There are three main parameters in the operation of the mixer: The mixer bias voltage and current (Vj and Ij, respectively), and the magnetic field. Vj and Ij are dependent on LO frequency (but doesn t vary that much), and the magnetic field strength is not. This means that the magnetic field strength does not have to be adjusted every time the receiver is tuned to another frequency. Vj is set directly from the mixer bias supply which can be controlled be either a manual control box or by computer. The mixer current is controlled by the amount of LO power used to pump the mixer. This control is done by monitoring the mixer current while the power level to the multiplier is adjusted. In operation at the telescope, these parameters are controlled by the main computer on the receiver tuning screen. The magnetic field is the third parameter and takes a small bit of technique to optimize. The other thing the operator has to watch out for is for trapped flux which can occur for the long periods when the receiver is in operation. Section 2.X will describe how do remove this trapped flux and to de-magnetize the electromagnet. 2.1 Setting the Mixer Voltage and Current Setting Vj: The first parameter to set is the mixer bias voltage Vj. This is done by simply typing the desired value in the proper field of the receiver tuning screen on the control computer or by turning the Vj adjustment potentiometer on the manual control box. In either case, Vj will be displayed on the computer screen or control box. This value does not varies from between 2 and 2.5 mv. The optimal value for Vj as a function of LO frequency is shown in the upper half of Figure Setting Ij: The mixer current Ij is set by adjusting the LO drive to the mixer. Again, this is done by typing in the desired mixer current in the proper field of the receiver tuning screen of the control computer. In manual operation, this level is adjusted by turning the knob of the level set attenuator on the output of the LO power amplifier. Turn the knob until the desired value for the mixer current is displayed on the manual mixer control box.

4 Figure 1. Optimal mixer bias values for Vj and Ij, respectively as a function of LO frequency.

5 2.1.3 Setting the Magnetic Field: Setting the magnetic field takes a little bit of practice, but after a few tries, it becomes pretty straightforward. It is especially important to make sure that the electromagnet is turned off when warming up the receiver. Having current flow through the electromagnet at room temperature could potentially damage the electromagnet by overheating. The main purpose of setting the correct magnetic field is to suppress Josephson s noise which lowers the sensitivity of the receiver. However, an arbitrarily large value of the magnetic field will not yield the best results, and can damage the electromagnet solenoid. Figure 2 shows the concept of obtaining the proper electromagnetic field current. As can be seen in the graph, there are two minima in the critical current. The first null around 5 ma typically gives the highest sensitivity, but with such a sharp null, the long term stability of the receiver may not be optimal. The better null is around 11 ma. The sensitivity may not be as high (we re only talking about a few Kelvins), but the broader null will provide better overall stability of the receiver. This is the target that the operator should aim for Critical current pol. 1 I Critical (ma) sweet spot I (ma) Magnet Figure 2. Sweeps of SIS junction critical current as a function of mixer magnet current of the mixers. However, determining the best magnet current (I magnet ) will be done in a more qualitative manner by observing the total power trace on the oscilloscope. The operator will vary the bias voltage while observing the total power curve and adjust I magnet. Figures 3 and 4 illustrate what the screen on the oscilloscope looks like before and after Josephson s noise is minimized. The trace in Figure 4 shows the effect from Josephson s noise being minimized which would correspond to the second null in Figure 2. The desired function is to minimize the amplitude of the bumps, but they will never be totally eliminated. This effect is independent of LO frequency, so it only has to be set once. However, this effect is dependent on the environmental conditions of the mixer. For example, if the mixer is not at the same physical temperature or if there is trapped magnetic flux in the mixer, I magnet will need to be re-optimized. Removal of the trapped flux requires the heating of the mixer which will be described in the next section.

6 Bumps in the total power curve with no magnetic field showing the effect from Josephson s noise. Figure 3. Pumped IV curve without magnetic field showing the effects of Josephson's noise on the total power curve. In this case, the maxima of the bumps in the total power curve are off the screen. Figure 4. Pumped IV curve without magnetic field showing the effects of Josephson's noise being minimized on the total power curve. In this case, I magnet is found to be 13.7 ma.

7 2.1.3 Mixer Defluxing: Limitations to Cartridge De-Flux Heater Use In order to ensure rapid and efficient defluxing of the cartridge s mixers, the deflux heaters are capable of dissipating a significant amount of power. In order to avoid damage to either the heater or the mixer itself, operation of the deflux heater at room-temperature shall be limited to the application of short pulses (< 1 sec. in length) for the purpose of verifying that electrical contact is present. This shall only be done when necessary. In order to minimize the heat-load on the cryostat during defluxing, and to minimize the time required for the mixer temperature to stabilize after defluxing, the voltage pulse applied to the heater shall have a duration that is not significantly longer than needed. Maximum recommended pulse lengths for use when the 4 K stage is at 4.0 or 300 K are: Table 1 De-Flux Heater Pulse Lengths Deflux heater pulse length T 4K stage = 4.0 K Deflux heater pulse length T 4K stage = 300 K < 1 sec. (recommended value in tuning tables) < 1 sec. (max. allowed) Applying power to a deflux heater for a period of seconds while cold (at 4 K) should not damage either the heaters or the mixers, but will result in unnecessary heating of the entire optical assembly, including the other mixer. This will result in a need to wait for a period of minutes to allow the mixers to stabilize at their operating temperatures. (Stabilization of the mixer temperature may be observed by measuring the mixer bias current as a function of time at a bias voltage of 2.2 mv, preferably with magnetic field applied.) Mixer Defluxing Procedure The mixer defluxing procedure is: - Set the bias voltage and magnet current of the mixer to be defluxed to zero. Turn the LO power OFF. - Apply a demagnetization sequence to the magnet. This consists of a consecutive series of current settings to the magnet, starting at a high positive value, and oscillating between positive and negative values, while reducing their magnitude to zero. The recommended sequence is: +50 ma, 0, -50 ma, 0, +40 ma, 0, -40 ma, 0, +30 ma, 0, -30 ma, 0, +20 ma, 0, -20 ma, 0, +10 ma, 0, -10 ma, 0. - Apply a defluxing pulse to the mixer heater: 24 V (fixed-voltage) for ~0.2 sec. Done manually, this duration is not critical. The idea is to just apply the heating pulse to the mixer for a short period of time (< 1 sec.). - An example of these steps is illustrated in Figure 4-1. Again, the duration for the pulses is not critical, but should be kept relatively short to minimize the effect to the other mixer and the environmental conditions of the receiver. Note that if defluxing at lower temperatures gives problems, it may be necessary to increase the heater pulse length slightly. If this is needed, the SIS mixer may be biased at 4 mv while applying the heater pulse to allow the temperature of the junction to be probed during defluxing (as seen in the blue curve in Figure 4-1, the junction s normal-state resistance will increase suddenly when it has warmed up sufficiently).

8 I Magnet (ma) Demagnetization Current Time (sec.) 1 Defluxing heater pulse 25 V Bias (au) Time (sec.) V Heater (V) Figure 4-1. (top) Magnet demagnetization sequence to be applied prior to deflux heater operation. (bottom) Deflux heater pulse and associated change in SIS bias as a function of time (for 4 K operating temperature). 3 Warm Up Procedure 1. Turn down the LO drive to the mixers, either though the receiver tuning screen on the control computer or manually. 2. Turn down the bias voltage to the mixers. 3. Very important: Turn down the electromagnet current I magnet to 0 ma, then turn off the magnet current control. Leaving it on at room temperature can damage the electromagnet solenoid. 4. Turn off the power supply to the LO power amplifier and electromagnet. Note: It is very important that the LO power amplifier is on before the signal is applied to it. Applying signal power before applying DC power to the amplifier can damage it. Never apply signal to the amplifier without having its DC power applied first. 5. Turn off the bias to the HFET IF amplifiers. 6. Plug in the two pin heater cord to the connector of the receiver and plug the other into the isolated VARIAC transformer. 7. Plug the VARIAC into the wall, turn on the power and turn the knob up to about 85%. The warm up time is about 3.5 hrs. Pay close attention to the receiver temperature towards the end of the warm up cycle in case one of the thermal switches fails during the warm up cycle. No temperature inside the Dewar should ever exceed 314 K.