Differences

This shows you the differences between two versions of the page.

--- ice:ice_quickstart_guide [2021/11/23 22:51] – external edit 127.0.0.1
+++ ice:ice_quickstart_guide [2022/08/09 00:14] (current) – external edit 127.0.0.1
@@ Line 24: / Line 24: @@
-<imgcaption ice_4_laser|The ICE Box controlling an atom-locked 4 laser system in maser/slave configuration>{{ :ice:ice_4-laser_system.png?400|}}</imgcaption>
+<imgcaption ice_4_laser|The ICE Box controlling an atom-locked 4 laser system in master/slave configuration>{{ :ice:ice_4-laser_system.png?400|}}</imgcaption>
@@ Line 75: / Line 75: @@
 ====Installing and Using the ICE Control GUI====
+<WRAP center round important 100%>This guide demonstrates interfacing with the ICE Box using the Vescent ICE GUI purely because it is a convenient visual aid. However, the official ICE GUI has several known issues and as such, we recommend that users primarily interface with ICE through the [[ice:commands:overview|Serial API command structure]]. The GUI should be used only at set up for basic confirmation that a system is working before transitioning to the use of Serial API commands.
+</WRAP>
 The ICE Control GUI can be found on our Github page [[https://github.com/Vescent/ICE-GUI/releases/tag/V1.2|here]], and installed by scrolling to the bottom of the page and choosing between the three file formats. If you are using a Windows 10 machine, it will be easiest to download "ICE_Control_v1.2_32bit.zip", which includes an executable file. Once you have downloaded the .zip file, extract it to whichever directory will be most convenient to run an executable from. The executable must be left in the extracted folder, as it will have trouble finding some of its dependencies if you move it. It is important to note that you must both save the file (the Windows Default is to "Open" files), and also extract the file before the ICE Control GUI can be used. You will not be able to open the ICE Control GUI if you do not perform both of these steps.
@@ Line 89: / Line 93: @@
 The list of board slot numbers on the left hand side of the ICE Control GUI will illuminate, indicating which boards are installed. Boards can be selected from this list by clicking on the numbers to navigate to the control options for the corresponding ICE board.
 =====Locking a Laser to Spectroscopy=====
-After making all the necessary connections you will need to find spectroscopy. This can be done by enabling the laser through the ICE CS1 menu and setting the **Laser Current** to a value which gives the desired output (see <imgref step_1>). From there, modify the setpoint temperature of the QT1 card to the temperature specified in your D2-100's documentation. To do this, first ensure that the **T Min** and **T Max** temperatures are configured to safe values for your system, then click on the **TSet(C)** field to enter a new value with the keyboard. For the D2-100 it is recommended to set **Stage 1** (Top) to 25°C, and to keep **Stage 2** (Bottom) between 15°C and 30°C.
+After making all the necessary connections you will need to find spectroscopy. This can be done by modifying the setpoint temperature of the QT1 card to the temperature specified in your lasers documentation. To do this, first ensure that the **T Min** and **T Max** temperatures are configured to safe values for your system, then click on the **TSet(C)** field to enter a new value with the keyboard. For the D2-100 it is recommended to set **Stage 1** (Top) to 25°C, and to keep **Stage 2** (Bottom) between 15°C and 30°C.
 <imgcaption temp_1|QT1 GUI showing TSet(C) and T Min/T Max locations>{{ :ice:ice_temp_1.png?400 |}}</imgcaption>
 If using a D2-100 DBR laser purchased from Vescent, refer to the documentation which came with the laser to find the approximate **Stage 2** setpoint. Once the desired value is entered, click the **Servo: Off/On** button to engage the temperature servo. The **Temp(C)** field displays the measured temperature of the thermal plant, and **TError(mK)** displays the difference between **TSet(C)** and **Temp(C)** while the values are within the range of the monitor. **TError(mK)** is plotted in the graph below.
 <imgcaption temp_2|QT1 GUI showing engaged servo and live error graph>{{ :ice:ice_temp_2.png?400 |}}</imgcaption>
-If your D2-100 documentation is unavailable, start by setting **Stage 2** to 15°C, engage the servo, then increase it 0.5°C at a time until spectroscopy is found or 30°C is reached. Assuming that **Ramp** is enabled, and the value of **Range** is non-zero, this should be a reliable method to find spectroscopy.
+It is now safe to enable the laser on the CS1 card. Check that the current limit, and current setpoints are both within the specified range found in the laser documentation, then click the "Laser On/Off" button next to the current dial.
 <imgcaption step_1|Rb Spectroscopy (top) and Peak Lock Differential signal (bottom)>{{ :ice:ice_step_1.png?400 |}}</imgcaption>
+If your D2-100 documentation is unavailable, start by enabling the **Ramp**, then  setting **Stage 2** to 15°C, engage the servo, then increase it 0.5°C at a time until spectroscopy is found or 30°C is reached. Assuming that the value of **Range** is non-zero, this should be a reliable method to find spectroscopy.
 Once spectroscopy has been found, align the feature you want to lock to with the center line on the GUI plots. You can do this either by changing the **Laser Current** supplied to the diode, or by shifting the **Center** dial in the **Ramp** box. Turning **Range** up and down will make your ramp amplitude larger or smaller and show you more or less of your spectroscopy. Whatever is lined up with the center line on the plots is the feature the ICE Box will try to lock to. If you want to lock to a peak, align the top graph with the peak you want, and note that the bottom graph's peak lock signal should be near the middle of a slope.
@@ Line 106: / Line 113: @@
-<WRAP center round important 100%>The intuitive thing to do is to center right on the feature you want, but you actually want to be a little bit offset from that. The ICE Box gives a slight kick when the servo is engaged, so if you're perfectly centered on a feature it can sometimes bump you over to the next peak. It doesn't matter which side you offset from (a little right or left will be fine) as long as you're still on the correct slope. The servo will always kick towards the center of the selected slope. It's recommend to aim for a roughly 5% offset from center when aligning the feature you wish to lock to.
+<WRAP center round important 100%>While it may be intuitive to center right on the feature you wish to lock to, it is better to center the feature with a small offset either to the right or the left. The ICE Box servo gives a slight kick when the it is engaged, so if your feature is perfectly centered it can sometimes bump over to the next peak. The servo will always kick towards the center of the selected slope. A roughly 5% offset from center when aligning the feature is ideal.
 </WRAP>
@@ Line 116: / Line 123: @@
-<imgcaption step_2|TEXT>{{ :ice:ice_step_2.png?400 |}}</imgcaption>
+<imgcaption step_2|CS1 GUI showing the Ramp Box, and the Center Line to which the Servo Locks>{{ :ice:ice_step_2.png?400 |}}</imgcaption>
 Engage the **Servo** button and watch the signals. The top graph should go to a value that corresponds to the feature you were trying to lock to, and the bottom graph should go to zero.
-<imgcaption step_3|TEXT>{{ :ice:ice_step_3.png?400 |}}</imgcaption>
+<imgcaption step_3|When the ICE Box locks to spectroscopy, the top graph will flatten out at a DC offset corresponding to the locked feature (Yellow Line)>{{ :ice:ice_step_3.png?400 |}}</imgcaption>
-If these do not happen, it might mean that your **Op Offset** and **Gain** settings are not optimal like in <imgref wrong_gain>. To optimize these, turn your gain down as low as it will go without losing lock, and then adjust the **Op Offset** so that the signal on the bottom graph goes to zero. You may not be able to get it exact due to the digitization of that setting, but there will be a setting for which it is closest and changing up or down will flip what side of the line you are on. Once you've done this, turn **Gain** back up until the signal looks nice. Getting **Gain** to the right setting is a little difficult to do with just the ICE Box, but a good way to approximate it is to increase **Gain** until you lose lock, and then back off by ~40-50%. The optimal **Gain** setting ultimately comes down to your measurements of noise on externals such as a beat note between two lasers.
+If these do not happen, it might mean that your **Op Offset** and **Gain** settings are not optimal such as in <imgref wrong_gain>. To optimize these, turn your gain down as low as it will go without losing lock, and then adjust the **Op Offset** so that the signal on the bottom graph goes to zero. You may not be able to get it exact due to the digitization of that setting, but there will be a setting for which it is closest and changing up or down will flip what side of the line you are on. Once you've done this, turn **Gain** back up until the signal looks nice. Getting **Gain** to the right setting is a little difficult to do with just the ICE Box, but a good way to approximate it is to increase **Gain** until you lose lock, and then back off by ~40-50%. The optimal **Gain** setting ultimately comes down to your measurements of noise on externals such as a beat note between two lasers.
+<imgcaption wrong_gain|Demonstration of how incorrect Offset and Gain can affect your lock>{{ :ice:went_wrong_change_gain_and_op_offset.png?400 |}}</imgcaption>
+Finally, if you are only seeing the spectroscopy on the top graph of the CP1 GUI, and not the peak lock signal on the bottom graph, or if your peak lock signal appears to be small or weak, it is likely that your Phase and Dither settings are incorrect. Generally, it is best to optimize the Phase setting by adjusting the corresponding dial on the GUI, and then minimizing your Dither such that you are still able to reliably lock to the desired feature. The 4MHz frequency dither which is used to generate the peak lock signal is written onto the D2-100 being controlled by the ICE box, so the smaller the dither amplitude can be the better.
+<WRAP center round important 100%>If using the mouse to drag the position of the GUI dials for dither and phase, note that the corresponding values are not updated until the mouse is released. For this reason, it's recommended that the + and - buttons on the face of the dial are used when attempting to optimize phase and dither.
+</WRAP>
-<imgcaption wrong_gain|TEXT>{{ :ice:went_wrong_change_gain_and_op_offset.png?400 |}}</imgcaption>