9V dummy battery

Never worry about depletion and replacement.

Reflector + collimator = reflective collimator

It's already more than 10 years ago. One day, I got curious... Collimator lens vs reflector, which one would be better? Regarding this question, I think I am not alone, because I could find some other threads related to the same question.

It was a very hard-to-answer question, so the solution I found was having them together. To this end, a thin layer of aluminum was vapor deposited on the back of collimators. In the next figure, the left hand side is an aluminum deposited collimator, and the right hand side is a normal collimator.

unfortunately, this project was not so successful, because I didn't see any noticeable improvement of brightness even after the aluminum deposition. Maybe, that is because the backside leakage of collimators is quite negligible.

UCO Clarus, my new backpacking gadget!

After getting back from a long hike to Rae Lake Loop, I thought I need a small light weight LED lantern. After some research, I decided to purchase a UCO Clarus lantern, and now I am so happy with that. It just weighs about 2 ounces and super bright. A little downside is that its burn time is less than 2 hours in its high-mode (150lm). However, this is not a downside at all in fact, because it can be easily overcome by using a high capacity Li-ion battery in place of original 3AAA batteries. A 22500 Li-ion battery (2300mAh) nicely fits UCO Clarus as if originally designed for it, and provides about 5 hours' burn in the high-mode - it's more than enough for two days' backpacking trip.

 

Left - 3AAA, right - 22500 Li-ion

Left - 18500 Li-ion with spacers (18-22 O-rings), right - 22500 Li-ion

Test equipment - impacting device

It's already 10 years ago. I built a test equipment at my friend's request. He was a material engineering PhD candidate and needed an equipment that could apply repeated and consistent impact on a sample surface. He also asked that the intensity and rate of the impact should be adjustable.

So I devised this device:

1. The electromagnet (EM) header first picks up a steel bearing at a time, and drops it onto the surface of the test sample. The impact intensity can be adjusted by changing the height of the EM header and the period is programmable at the host PC.
2. The steel bearings are reclaimed after the impact and raised to the upper repository to get ready for the next drop.
3. Experimental data is collected by the sensor module and analyzed at the host PC.

The EM header is an air core electromagnet, which has minimum effect of residual magnetism and housed within a PTFE case.

AC source noise filter for audio amps - activated by USB

I own a pair of AudioEngine A2 speakers and I really like them. To enhance my love for them, I planed an AC power source upgrade project.

I planed this project for two things:
1. Suppress the AC source noise - this is the primary goal.
2. Whenever I turn my PC on and off, I want the speakers to be turned on and off together.

The second goal, however, was a small concern, in fact. If you can say you are an audio guy, you must have heard some people say that turning amps on and off without turning down the volume may damage the delicate circuitry. However, at the same time, the others say that today's amps are advanced enough and they no longer need such circumspection. I believed the latter is true. But, just for peace of mind, I sent an email to the Audio Engine customer service and asked them to assure two things for me:

1. Unplugging the speakers directly from the AC source will not damage the speakers.
2. Even if it causes damage on the speakers, it will not void the warranty.

OK. The green light is on. The following are my AC filter design and implementation:

DC source noise rejection filter for DAC/headphone amps

When I first purchased my DAC 4 years ago, it came with a 15V DC linear power supply. As you know, linear power supplies are very heavy and bulky, especially when their power rating is high. On the contrary, switching power supplies are light and small. But there are always pros and cons. Despite all the excellence over the linear counterparts, switching power supplies usually have high output ripple and noise. The output noise is quite intrinsic due to its mode of operation and very hard to avoid. Therefore, at least for applications where the noise is a big concern, linear PS is a good choice.

If for some reason there is no other choice but using a switching power supply, a hybrid type power supply can be used. For example, let's say that we want to get 15V DC source. Then we start with a 18V switching PS. Because the output may have a lot of ripple, we can trim it out with a linear regulator of 15V output. There are bunch of linear regulators out there with superior ripple rejection characteristics, so we have very high degrees of freedom in the part selection.

The followings are my design:

Parts:
LM1084 x 1
capacitors
  10u x 2
  100u x 1
resisters
  150 x 1
  5k x 1 (variable)
panel mound DC jack x 1
5.5/2.1mm male DC connector x 1
heat sink x 1
volume knob x 1

The most challenging part? Well, it was so tricky to pack all the part within a tiny case.

Solid-state turn-signal flasher relay; it's completely noise-free.

Mechanical flasher relay has been replaced by a solid-state relay.

Left - solid-state relay
Right - OEM flasher relay

It's completely silent! So, I found it was sometimes dangerous because I kept forgetting to turn off the turn signal. So I decided to add artificial sound. A door chime unit is used to this end.

Rather than irritating mechanical ticks, it makes clear ringing when the turn signal is on. The sound is so refreshing.