I bought a new photodiode for the West Bridge Labs. It will be housed in the QIL (WB265B) in the central cupboards. There is a QR code on it linking to this page.
I've modeled the theoretical "best" achievable configuration in Megastat, replacing the flexible strap with a solid copper element, and leaving all but one aperture open in the inner shield [Attachment 1]. In this configuration, the wafer can reach 123K in ~20 hours.
- The difference in time constant between the copper bar end and cold head (purple and blue curves) is due to the thermal resistance across the bar and 2 joints, plus realistic radiative losses from bar.
- The difference in time constant between the cold plate and copper bar end (brown/green and purple) is the thermal resistance across the additional solid copper finger and 2 joints.
- The stead-state temperature difference between the inner shield and the wafer (orange and green) is maintained by heat leaking from 1 aperture in the shield. (We've closed up 4, but best case there will need to be 1 open for the copper bar and leads.
Attachment 2 shows power budgeting of heating and cooling power delivered to various components: 1. copper bar, 2. cold plate, 3. inner shield, 4. Si wafer. The points of intersection of the heating and cooling power curves correspond to the steady-state temperature for each component, which can be verified by Attachment 1. The total heat load on the copper bar and cold plate is just over 70W.
The ideal model shows that there are two areas of potential gains:
Analyzing the most recent cooldown data will hopefully validate the efforts of closing up apertures and painting the entire wafer enclosure with aquadag.
The cryocooler was turned off 6/8 at 11am.
On Tuesday 6/7 we met to discuss next steps for getting PD testing up and running in CAML (cryo auxiliary mariner lab). Here is a rough 2-month plan:
Step 0 [1 week]: Get CAML workstation running and confirm connection to EPICS/CDS.
- Organize BNC cables coming in from QIL and connect to the PD testing table.
Step 1 [1 week]: Outline desired measurements and sketch diagram of electronics/components.
- Decide how we want these component organized; i.e. how many boxes, number of inputs/outputs per box.
Step 2 [2 weeks]: Implement desired setup.
Step 3 [2 weeks]: Replicate QE measurements on JPL PDs
- Tweak setup as necessary
Step 4 [2 weeks]: Test MCT detectors
I encountered several puzzles in the data from this cooldown (Attachment 1):
1. Due to residual challenges with getting the Megastat SLOW channels recorded to frames (after flood), we've been using a USB to extract data form the CTC100. However, only ~48 hours of cooldown were recorded on the drive - we lost data from the rest of the cooldown. I'm hoping this is due to storage limits on the USB, and if so I will decrease the sampling rate of the CTC100 and/or source a USB with larger storage. Of course this is a temporary solution, and efforts would be better spent on getting data recorded in frames. For now, since every component had reached steady state (except the test mass), most parameters can theoretically still be inferred from the data we obtained. (Note that data logging started a bit after the cryocooler was switched on.)
2. The inner shield appears to have reached a steady-state temperature lower than that of the cold head. This is bizzare, and non-physical - the cold head should be the component supplying cooling power to the rest of the system. Our current Megastat model would not predict such behavior under any choice of parameters. I replaced the inner shield RTD before this cooldown, so I am hoping the new RTD has a different calibration curve than what the CTC100 was expecting. If this is the case, I should be able to use the data to undo the existing calibration and obtain correctly-calibrated temperature for the inner shield. It is worth noting that the new RTD accurately measured room temperature, so if this theory is correct then the calibration error only exists at low temperatures.
I looked closer at the RTDs, and they are all PT100s with the same resistance-temperature curve [Attachment 1]. So we might need to re-calibrate them.
The chamber is currently warming up, and I plan to cool down again on Monday to a) obtain data for the entire cooldown and b) see if this inner shield behavior is repeated.
From the plot, it seems like the slop is ~50 Ohms / 120 K. So a difference of 5 K corresponds to ~2 Ohms.
It could be that your measurement of the resistance is off slightly due to the lead resistances, etc. Are you using a 3-wire or 4-wire probe?
You can get a higher accuracy relative calibration by dunking all the RTDs together in ice water and then in boiling water.
RTD Calibration Plan Developed on 16-06-2022:
Prepare an ice bath and liquid nitrogen bath to calibrate the RTD.
Ice bath: fill cup (?) with ice. Fill with water until 2 inches below the top of the ice. Let sit for 2 minutes before calibration.
We will use a DMM to measure the resistance across the leads. Right now, we are thinking that we will connect the leads to the resistor using alligator clips or solder them together.
Use linear fit to calibrate the RTD value as a function of resistance using these two reference values.
Repeat for all three RTDs
Q: What is the tolerance of the resistor?
100 ohm +/- .06%. Means that our calculation if done correctly will be between 99.4 → 100.6 ohms in the ice bath. Our RTD is class A, which has tolerance +/- .15 degrees. Pt-100 SHOULD be 100 ohms at 0C – the temp changes linearly.
Here I summarize the analysis of thermal conduction through 3 ceramic ball bearings supporting a test wafer in Megastat. I considered a baseplate or housing for the ceramic ball bearings at the temperature of the cold plate, the 3 ceramic ball bearings themselves, and the wafer seated on top. This pathway can be broken down into the following components:
1. Thermal resistance across joint from Al baseplate to ball bearing
2. Thermal resistance of bulk of ball bearing
3. Thermal resistance across joint from ball bearing to wafer
These resistances in series can be summed, and there exist 3 of these pathways in parallel (1 for each bearing) and the net resistance can then be found using the corresponding formula.
Attachment 1 is a plot of the thermal conductance of alumina ceramic (Al2O3), and it was taken from from this paper. The peak value of 99% Al2O3 agrees with what is listed in the ceramic ball bearing spec sheet: 28 W/mK. I've used these tabulated values to obtain component 2 above.
Attachment 2 is a plot of the thermal conductances across various joints, taken from Ekin. I could not find data for the conductance across Al--Al2O3 pressure joints or Al2O3--Si pressure joints, but I have used the value corresponding to a stainless steel--stainless steel pressure joint as a conservative upper bound. (*When I was calculating non-negligible conduction last Friday, I had forgotten to update the joint conductance value from an Al--Al pressure joint (orange box in Attachment 2). This greatly overestimated the joint conduction, and the ceramic joints should be a few orders of magnitude below this.)
The results of this modeling show that the maximum cooling power delivered via conductive contact is below 7e-5 W. Compared to the 6e-3 W of radiative cooling power (~1%), this is negligible. Attachment 3 shows the nominal difference in wafer cooldown - the solid and dashed orange curves are barely distinguishable (T difference of ~0.2K at steady state - RTD uncertainty is +- 0.15K). We should proceed with this design plan for emissivity testing in Megastat.
I took the Moku and iPad from QIL over to 40m for PLL Measurement at the 40m. (SURF)
Here I analyze potential gains from using LN2 to kick off a Megastat cooldown, and transition to a cryocooler to cool under 77K.
Here I analyze potential gains from using LN2 to kick off a Megastat cooldown, and transition to a cryocooler to cool under 77K. I've removed the heat loads in the system by modeling an empty chamber (no wafer) and no exposed apertures in the inner shield. I've also modeled the flexible strap is a solid copper element, for an ideal cooldown scenario.
Attachment 1 is a reminder of the cooldown of the cold head of the cryocooler. The cooling power of the cryocooler is a function of the temperature at the cold head, and this is tabulated in its manual. The cold head reaches 77K at ~3.8 hours from past data. This provides an upper limit on gains from using LN2 - if the cold head is cooled to 77K instantly, its cooldown curve will be shifted to the left by 3.8 hours, and thus the system will reach steady state 3.8 hours faster. In reality, there would be a time constant for the cold head to reach 77K due to the LN2-copper interaction, and the mass/heat capacity of the cold head.
Attachment 2 is a plot from Ekin which shows the heat transfer rate (Qdot) per m^2, for the boundary between LN2 and a solid. The different curves correspond for different between the LN2 and cold head, and for a majority of the cooling from RT, we will use the curves for ~ 100K. These have lower heat transfer because film boiling is occurring, and a layer of film "insulates" the surface of the solid from the 77K bath. These curves vary in the diameter of the contact area, and we use the bottom curve corresponding to D >= 1.0 cm. Once reaches 10K, the heat transfer rate jumps up to the nucleate boiling curves. We use the one corresponding to 1atm of pressure.
[Radhika, Hiya, and Clare]
We debonded the RTDs from the chamber using isopropanol and acetone, then soaked the RTDs in isopropanol for about 15 minutes to remove residue. We then took resistance measurements in ice water detailed below:
Ice Water Calibration Measurements (completed 28-06-2022):
1) Prepared ice water bath by filling beaker with crushed ice and then water to one half inch below the ice surface. Let the ice bath sit for about 1 minute.
2) All six RTDs were measured (labeled A-F). For each RTD, the resistance in the ice water bath was measured by swirling the RTD held by the DMM leads in the water until the DMM readout stabilized.
3) These measurements were repeated in the reverse direction (Started with RTD F).
*Note: in the process of ice bath calibration, RTD D broke.
Liquid Nitrogen Calibration Measurements (completed 30-06-2022):
Procedure: We clipped the DMM leads to the RTDs and taped the two clips together. This is very secure - good method for the future. We dipped the taped clips with the RTD into the LN2, and swirled under the liquid surface until the DMM readout stabilized. We only took one set of measurements because we are very confident of the boiling point of N2.
*Note: RTD A broke. This is because we spread the RTD wires out to put into the aligator clips, making them very prone to snapping when we move them. This means that we do not have an RTD on the cold head.
On Thursday 6/30 we worked towards a cooldown with a glass wafer rested on ceramic ball bearing supports. We used a glass wafer for this cooldown because it has a higher emissivity than a Si wafer, and therefore could be more easily validate our cryostat setup.
The procedure went as follows:
1. RTD preparation [Attachments 1, 2]
- Re-soldered the leads of the 3 RTDs we previously calibrated (2791)
- Bonded RTDs to the inner shield, outer shield, and glass wafer with varnish
*Note that we did not have any RTDs left to clamp to the coldhead
2. Wafer support structure [Attachments 3, 4]
- Applied grease to the bottom of the baseplate and bolted it down to the cold plate using 2 screws and a dog clamp. The hole spacings on the cold plate and base plate did not align align along the longer axis of the baseplate, so we needed the dog clamp to hold down one end. To get the dog clamp elevated to the right height, we folded a piece of aluminum foil until it was tall enough to rest the dog clamp flush on top of the base plate.
- Rested 3 ceramic ball bearings and wafer on top (*The ball bearings created enough clearance between the dog clamp and the wafer)
We closed up the chamber and started the vacuum pump [Attachments 5, 6], but soon after I realized the wafer RTD was glitching/shorting. I turned off the pump and re-opened the chamber, and upon inspection noticed that a few of the RTD lead pins were fraying. I replaced these pins and left for the day.
On Friday I verified that all RTDs were operating properly and closed up the chamber. The vacuum pump was started at 3:10pm and the cryocooler was switched on at 4:10pm on 7/1.
The cryocooler was switched off on Tue 7/5 at 11:45am.
[Stephen, Clare, Hiya, Radhika]
When we opened up the chamber, we noticed some of the aquadag-coated foil was touching the bottom face of the glass wafer [bottom edge of wafer in Attachment 1]. This is a potential conductive short, and analyzing the cooldown data will reveal how significant the conduction was. We also noticed several aquadag flakes on the ceramic ball bearings/wafer [Attachment 1]. Stephen pointed out ways to improve the clamping of one end of the base plate to the cold plate, given that the hole spacings of the two did not line up (explained in 2792). We mapped out a new bolting/clamping approach for the baseplate using metal spacers instead of folded Al foil [Attachment 2]. We applied grease to the bottom surface of the baseplate and then bolted/clamped it down to the cold plate [Attachment 3].
We removed the glass wafer and debonded the varnish from the RTD. We then re-varnished this RTD to the undoped Si wafer [Attachment 4] and placed it on top of the ceramic ball bearings on the bolted/clamped baseplate [Attachment 5].
Lastly, we trimmed off excess aquadag-coated Al foil to ensure there would be no future shorting to the wafer. We then verified all RTDs were responsive and closed up the chamber.
The vacuum pump was turned on at 5:35pm, and the cryocooler was turned on at 6pm.
We started a cooldown on Thursday with an undoped Si wafer in Megastat . The chamber does not have the Maglite flashlight installed, due to struggles with soldering the ground wire to the aluminum surface of the flashlight. We decided to pivot to drilling a hole in the aluminum, wrapping the ground wire through and around, and then applying solder to the joint. This will be completed by next week.
These tasks will ideally lead up to a Si wafer cooldown next week with optimal heat excitation and with an RTD re-installed on the cold head, and potentially with the finalized baseplate and aquadag-coated Al sheet on the cold plate. It is with this cooldown that we can start to estimate the sample’s emissivity, first with least-squares fitting and eventually with MCMC analysis.
We comfirmed that the DC ouput of one of the 1811s is bad. We set out to measure the AC response of the PDs.
For this, we decided to use the current modulation on the Diabolo laser which is rated to have 0.1 A/V and a bandwidth of 5kHz. We calibrated the current to optical power by swiping the current and measuring the power at the homodyne PDs using a power meter. The laser power before the 1064nm PZT mirror was measured to be 5mW.
Attachment shows the measurement and a linear fit with slope=0.97 mW/A.
We drove the current modulator using a sine wave from a function generator with 1kHz 0.5Vpp. When we looked on the PD AC signal port in the Oscilloscope we saw 2 Vpp 12Mhz signal. We passed the signal with a low pass filter but again we saw mostly noise.
We took the PDs to the 40m PD test stand but we accidently fried Jenne's laser.
Next, we should just use the Moku network analyzer instead of the scope to measure the response in the QIL using the Diabolo current modulator.
Update to QIL/2795:
- RTD calibration with Clare scheduled for Wed 8/3 morning.
- A heater will be bolted down to the cold plate for the next cooldown (to aid in chamber warmup).
- For the maglite, I took the wire intended for ground and wrapped it through the drilled hole and around [Attachment 1]. When I applied solder, it bonded to the wire but made no contact with the Maglite aluminum surface. The near-instant oxidation of the aluminum still prevents the solder from bonding to it. Previously when we tried scratching the surface and immediately using mineral oil (vaseline) to prevent oxidation, the solder joint still did not form due to the layer of oil. So I am not convinced that applying vaseline will help this joint with the drilled hole - I am in favor of considering an alternative connection. In addition, the wrapped wire does not allow the maglite shaft to slide completely into the "head", and thus the bulb does not make it through the opening of the reflector.
We performed a calibration procedure on the newly ordered RTDs for Megastat, in LN2 and ice water. We noted that these are 1kohm RTDs at 0C, as opposed to the 100ohm RTDs previously calibrated that are inside the chamber. We chose to calibrate 6 RTDs and labeled them alphabetically starting with G.
The first calibration point was taken at the boiling point of LN2. We dispensed the LN2 into a dewar and dipped each RTD just below the surface of the liquid. Since the temperature right below the surface is 77K with relatively high accuracy, we only took 1 measurement of each RTD resistance. The results are tabulated below:
The second calibration point was the melting point of ice. We created an ice bath by filling a beaker with crushed ice and adding deionized water until 1/2" below the surface of the ice. We let this equilibriate for a few minutes and then proceeded to dip each RTD and record its resistance. When we started this process, we realized we had not let the system equilibriate for long enough, since the temperature of the water was dropping with every measurement. The water surrounding the ice had not yet been cooled sufficiently. We scrapped the first set of resistances, but the 2 rounds of valid measurements are tabulated below. The multimeter was overloaded when set to display the resistance in ohms, so we had to display kohms and thus lost a significant figure.
RTD G was dropped and lost after calibration, but we calibrated more than we needed in anticipation of an accident like this.
I recall at one point we had one of these NF1811 with a broken power suply pin. It was from a limited production run with the smaller micro 3-pin power connectors. Maybe check yours is not that one.
Long story short it still responded with only the positive rail but DC will gave a bad photovoltaic mode response and the AC had a large unstable oscillation that was only viewable on a high speed scope (if I recall right higher than the 125 MHz nominal bandwidth). I would check the power-in pins aren't bent/broken and also check the AC out on a higher speed scope (i.e. >=500 MHz).
I acquired internal tooth lock washers to bolt down the ground wire to the aluminum maglite body, but I soon recognized that the curvature of the aluminum does not allow for flush contact of a washer or nut. Attachment 1 shows a 4-40 screw and nut relative to the maglite body.
I am continuing to search for the best way to get electrical contact, given these constraints and the issues with soldering in the last ELOG.
I torqued down the toothed washer and bolt holding the ground wire to the Maglite body and locally flattened the aluminum, securing the electrical connection. With the Maglite wired up, I connected a Tenma power supply to a mini breadboard and attached the Maglite leads [Attachment 1]. At 3V, 0.38A flowed through the incandescent bulb [Attachment 2].
- I will wire a current-limiting circuit to protect the bulb.
- Normally the Maglite batteries press up against an internal component to complete the flashlight circuit. Since we’ve cut off the body and of course have no batteries, I plan to use vacuum/cryo-safe epoxy to hold it in place (or use some spring mechanism to apply pressure to that component).
- The head of bolt used to connect the ground wire to the aluminum body sticks out and interferes with screwing on the flashlight's reflector. I will need to cut away some of the metal to allow the piece to screw in all the way.
Once these steps are complete, we need to address the feedthrough for the maglite leads into Megastat. Then it will be ready to be mounted inside the chamber and used for wafer heat actuation.