Pages

Wednesday, 19 May 2021

9V 1Watt Led Driver for single Li-ion cells

 Recently I stumbled across the MT9284 led driver IC. The datasheet shows it can be used for strings of 7 white leds. There is a suggestion in the graphs that the maximum led current is perhaps 20mA. 

I also had some 1 watt white leds by Bridgelux, BXEN-65E-13H-9B-00-0-0. These are approximately 9V at 110ma. 

Can the MT9284 drive one or more of these? Yes. I have tested to date one Bridgelux device at currents up to 80mA. The efficiency has varied from 83% to 92%, depending on led current. 

Nice and bright! I'm going to update my collection of homebrew pocket torches. I find as I get older that the part markings are getting harder to read and a bright light source that can illuminate the part from different angles makes a big difference.

More soon.

73's

Monday, 10 May 2021

Noise in "Jelly Bean" Regulators - Part 3

I repeated the exercise with a 78L08 and noise fell from 480mVpp to 30mVpp. Two surprises - the 78L08 was less noisy than the LN6206. I had expected it would be noisier since higher output voltages tend to have more noise for the 78Lxx parts. And apart from the 78L08 being quieter than I expected my test apparatus is not perfect. 

I had expected that the capacitance multiplier would reduce noise by the same factor as it did with the LN6206 ( 30 fold, 1800mV / 60mV ). But in this case I was only getting a 16 fold reduction (480/30). I checked and the unterminated amplifier was producing ~1.5mVpp on the CRO. However, when I put a 75 ohm termination on the input BNC the noise increased to ~30mVpp. Which requires more investigation.

Still, the capacitance multiplier makes a substantial difference and will become a standard inclusion on future analogue projects.

Flush with some sort of success I put a TL431 onto the test board with no capacitors anywhere in circuit and with an approximate output of 8 volts. 

There was no change in the 30mV of noise on the CRO. So while I cannot quantify how quiet a TL431 is, it appears to be a substantial improvement on the 78L08 alone. I am also unable to determine if the TL431, with capacitors, is an improvement. Or indeed if a capacitance multiplier is worthwhile.

But I'm not beaten yet.

73's


Thursday, 6 May 2021

Noise in "Jelly Bean" Regulators - Part 2

On the second attempt my low noise amplifier for measuring regulator noise worked. It's based on a dual low noise op-amp with an overall gain of 300 (x10 followed by x30) and no signal noise of a few mVpp. The very thorough noise calculator I used suggested I should have a better noise floor so further investigation is needed. 

I quickly checked a LN6206 regulator to confirm the approach and saw 1.8Vpp of amplified noise on the cro. That works out at around 6mVpp for the regulator which I might have been able to estimate with the cro itself. But now I can experiment with different noise reduction techniques and relative comparisons between regulators without having to estimate differences at that low level.

I started by trying a capacitance multiplier.  I used a 10k resistor, a 1uF tantalum and a general purpose npn transistor. That reduced the noise from the LN6206 from 1.8pp to 60mVpp. That is a large reduction in regulator noise with really no effort. 

 73's