S Muusse, M. Todd
New QCL unit (0920) has been put in place of original unit (0923) which is malfunctioning (cause unknown). Initial profiling has taken place and L1 focal length is 10% larger than spec.
QCL unit failure summary:
All day on 2026-01-21 we were running the laser at 900mA doing beam profiles and alignment work. No malfunctioning was witnessed and the laser unit was operating as expected, similar to the laser unit we had run in December for a week. Before lasing we set the following limits on the LD and TEC, per the datasheet.
- LD current limit was set to 1.05A
- LD voltage limit was set to 12.9V
- TEC operating temperature was set to 20C
On 2026-01-22 we turned the laser on to do some more beam profiling with the same limit settings as yesterday, but setting the LD current to 1A (forward voltage was around 12.2V) as was done when we originally profiled this laser in September. We were in the middle of setting up for a new beam profiling measurement (alignment showed the beam was fine, as usual) and we blocked the beam with a high power beam dump at the laser head while installing the profiler. Upon removing the beam dump, we noticed no power was coming out of the laser, and then noticed the forward voltage had dropped to around 800mV. We noticed no sounds or smells or any other signs that something had stopped, only sudden lack of light coming from the unit. All other settings were fine, meaning the controller had not faulted and was still outputing 1A LD current and the TEC was maintaining 20C.
We ran the following checks to see if we could remedy the problem, without success.
- Checked the cable connection to the laser unit (pins on the laser unit looked fine)
- Checked the Laser Controller cable connections (pins in the DSUB connectors looked fine)
- We tried a completely new laser controller box and new cable. The settings were similarly set and the same behavior was seen with the forward voltage response to the LD current setpoint.
- We tried changing the TEC setpoint to be 25C, and the same behavior was seen with the forward voltage response to the LD current setpoint, and failing to lase as usual.
- We used the same cable and laser controller on the previous QCL unit, and it operated as expected --- lasing fine as usua
QCL 0920 profiling summary:
Replaced broken unit with 0920 and confirms it lases as spec'd. We built telescope as modelled but beam profile was 75% of expected beam width at L2. We subsequently profiled the laser output and confirmed QCL output q measured previously by Matt was still correct. We became suspicious of the true focal length of L1. Then we profiled multiple places after L1 and fit a q parameter using a non-linear fit. A plot of this fit is attached. Using these 2 q parameters the focal length of L1 was estimated to be 220mm instead of 200mm as spec. We think this is mostly because L1 focal length stated by manufactorer is for 588nm where the refractive index is almost 3% larger than at 4.6um. Focal length in the model was modified to reflect our estimated f1 which predicted a beam size closer that measured earlier.
Side note: the scanning slit beam profiler reflects significant amount of 4.6um light which is observable on a thermal beam card.
This morning we continued profiling to characterize the true L2 focal length and install the modeled telescope and characterize it.
We replaced L1 with L2 in the telescope and made several measurements to fit for a q-parameter after L2, because we think we know the q going into it quite well. Then with the q parameters, you can estimate the focal length using ABCD matrix for a thin lens (our estimates yield relatively low complex angles, less than 1 deg, making confident estimates).
| Parameter | x [m] | y [m] |
| Input q (coming from the laser) | 0.411 + 0.067i | 0.376 + 0.056i |
| Output q (fit from profiles) | 1.496 + 1.289i | 1.221 + 0.632j |
| Focal Length (-q1/(q1/q2 - 1) | 0.503 | 0.511 |
Average focal length of L2 = 0.507 m. Which is around 1.5% different from the spec'd value. To reiterate Sophie's log above, the focal length of L1 is estimated to be 0.217 m. Which is almost 9% different than the spec'd value. With these values in hand (as well as the updated value for CaF2 refractive index at 4.6um), we can make a more accurate model and see if measurements of the outgoing q-parameter from the telescope match that model.
We installed a telescope as set up in a model and measured several proviles to fit for the q-parameter after L2. The modeled waist size in both the horizontal and vertical are within the fit uncertainty; however the waist position of the modeled beam is roughly 20cm off in both directions. I think this is because of the Gouy phase regime that we are sampling gives better estimates of the waist size and since we did not sample near the waist we do not have a good idea of where it is.
This afternoon we will try and re-build the telescope as optimized in our models with these measurements to see if we can get a q-parameter that will be 53mm at the "ITM" (propagated 35m from L2).
