I recently purchased the low-cost Thorlabs LD1255R precision constant current driver in an attempt to replace the expensive LDC205C driver that I’m using to drive the laser on the [∞] OpenRAMAN project.
For those who haven’t followed me recently, I’m struggling a little bit with the laser and TEC drivers to make the spectrometer cheaper. At the moment, the system uses lab driver that are completely overkill for the application and cost therefore a lot (2000+€). I therefore planned since a long time to replace these drivers with much lower cost versions.
My very first attempt was to adapt the [»] 1A LED driver I already presented to drive the laser diode which requires up to 400 mA. This circuit was a catastrophic failure and fried my laser almost instantaneously despite all the protection I had put (transient suppression, slow start, voltage clamps etc.). I then decided to check for commercial alternatives as I clearly had under-estimated the sensitivity of laser diodes to damages.
Unfortunately, the alternatives I was able to find did not ship outside US and I was left with two circuits from Thorlabs: the LD1255R and the LD3000R drivers. The former delivers only 250 mA but requires much less custom circuitry to be made than the latter which goes to 3A but need a PCB to be done. I therefore decided to go step-by-step and start with the LD1255R first.
The driver can be bought from Thorlabs just a little under 150€ (ex-VAT) and Thorlabs also offers a power supply (LD1255-SUPPLY, 110€ ex-VAT) which makes your laser driver for a little bit above 260€. You will have to add some relatively standard components like a SUB-D9 connectors to plug into the LDM21 block and also a toggle switch to connect/disconnect the DISABLE pin to ground. The circuit with the extra wiring is shown in Figure 1.
Thorlabs strongly recommend to disable the LD when switching on/off the power supply to avoid damaging the LD with transient voltages. Note that using the DISABLE pin does not stop the current from flowing but rather redirects the current through a transistor that short-circuit the LD. This has consequence for what I will discuss later in this post.
Although Thorlabs website advertises a 250 mA max current, mine was limited to about 225 mA. I suspect that this limitation comes from the power supply which delivers a maximum of 250 mA according to its datasheet while Thorlabs itself recommends using a 300 mA power supply (sic!).
At maximum power, it required 8 sec exposure time to take the acetone spectrum shown in Figure 2. This is better than the [∞] Starter Edition performances (and with higher spectral resolution too!) but still well below the typical ~1 sec required when running the LD at higher currents using the LDC205C.
A quick look at the circuit itself shows that the driver uses very simple circuitry that is priced ridiculously high for what it is. For instance, the component that can be seen in the front of Figure 1 is most probably a simple linear voltage regulator used to smooth out the voltage by lowering it to something like 5-7V (minimum required input voltage is 9V so it makes sense). The big problem with linear regulators is that they dissipate a massive amount of heat that is proportional to both the current and the voltage drop which can be huge. This makes the circuit burning hot after only 1-2 minute of operation as can be seen in Figure 3.
Thorlabs LD1255R driver is clearly not suitable for driving high loads for extended period of time. Running the circuit for more than a few minutes poses major risks of dramatic failure like burnt components or components desoldering from the board. You don’t want any of that to happen because you have no idea what such kind of failure may do to your LD that is connected to this circuit.
In an attempt to mitigate this problem, I machined a 20×50×10 mm block of aluminium with two M3 threaded holes to act as a heat sink to protect the (supposed) linear regulator. I added just a bit of thermal paste on the component for good heat transfer between the two. Be careful not to spill thermal paste on the leads of the component that would then short-circuits. A better alternative would have been to use thermal pad but I did not have some available at the moment. Note that the screws are tighten just enough to maintain contact between the component and the aluminium block. A picture is given in Figure 4.
The custom heat sink helped lowering the temperature of the board to somewhat acceptable temperature but after ~15 minutes of operation the circuit was barely touchable and the aluminium block had probably reached ~60°C.
Further analysis revealed that there are still over-heating components on the back of the board this time as can be seen in Figure 5. I was able to measure temperature above 120°C there too but I was not able to take a proper picture to show it here (I miss an assistant to hold both the burning hot circuit and the FLIR camera).
The component that produce the heat observed in Figure 5 is a small 2-leads SMT component that I believe might be the current sense resistor of the circuit.
Even the front of the circuit with the op amps show strong heat dissipation. This is illustrated in Figure 6.
I decided to turn off the circuit after 30 minutes of operation to avoid damaging anything. This was already enough to draw conclusions anyway.
Considering the results obtained here, I would not recommend using Thorlabs LD1255R precision current driver to power up the laser of OpenRAMAN [∞] Performance Edition. The circuit does not operate safely, even below the maximum recommended load according to its datasheet. Despite our best efforts to patch the problem by adding a heat-sink to the circuit, we were not able to provide with a durable solution due to a poor initial design of the circuit as it seems that Thorlabs was way too optimistic when setting the maximum admissible load knowing how the circuit operates.
The only possible way to use the circuit would be to constantly turn it off and on to avoid over-heating but that would not be a good solution to reach a steady state with the laser for precision experiment like those offered in the Performance Edition. Note that it is not possible to use the DISABLE pin to switch off the system because it will simply short-circuit the LD but still deliver the same power through the driver, causing the over-heating to go on.
This is then unfortunately a second failed attempt… I hope Thorlabs LD3000R will perform better. See you soon to discover the answer!
I would like to give a big thanks to James, Daniel and Mikhail who have supported this post through [∞] Patreon. I also take the occasion to invite you to donate through Patreon, even as little as $1. I cannot stress it more, you can really help me to post more content and make more experiments![⇈] Top of Page
You may also like: