Just a quick update; I’ve got a couple of electronics projects on the go which should be finished after Christmas. One is a personal project which is a modified version of an animated clock using two 24×16 modules from sure electronics. That displays some pong animations amongst other stuff. It’s still a WIP but I hope to get it completed soon. The other is a modified version of the AA battery tester I built – one of my first projects. I have changed it so that it can test LiFepo4 batteries with a nominal voltage of 24V. It can also be used to test car batteries and other high capacity lead acid batteries. It works but just needs a few tweaks.
Secondly I have returned to the 8051 microcontroller and built a replacement logic board for a 20 year old LED sign which is obsolete. This was for work not personal use so I am unable to post anything on this site about it due to IP reasons but basically the old controller board had around 30 chips on it of circa 1980 which was replaced with a board containing a Silicon Labs C8051F342 microcontroller, a RS232 chip and a couple of logic gates. The board works perfectly replicating the original sign’s functions with the added benefit of a much lower power consumption and faster updating of the displayed data. Of course the biggest benefit is we no longer have to potentially scrap 300+ signs and can continue to support them.
I’ll be posting information on the LifePO4 battery tester and pong clock in the next month or two.
The gas detector I built died also. It failed to turn off the power which was found to be a stuck relay so we used it as a monitor only. The sensor also failed to work and upon inspection I believe it is actually a MQ135 not the MQ136 as it is labelled. This is due to sensitivity to various gasses – the sensor is sensitive to gasses it shouldn’t be and displays the typical characteristics of a MQ135 sensor. The sensor was bought off amazon from China and the MQ136 is considerably more expensive than the MQ135. So this is typical; it looks like I’ve been ripped off and sold a cheap sensor labelled up as something else and sold at a higher price. I posted an article once on Chinese fake electronic components and it is becoming more prevalent especially high selling but not necessarily high value parts, typically sold to hobbyists.
Lesson learned – don’t buy sensors from China for a product you need to depend on. We were asked to build it as cheap as possible but at the end of the day it doesn’t do the job it’s meant to do so it’s no good. It makes a good air quality monitor but not a good H2S detector which is what we wanted. But having said all that the MQ135 can detect hydrogen which it does well and this is given off by lead acid batteries during charging. Hydrogen is explosive and odorless so although it can’t detect the rotten egg smell of H2S it can detect a defective battery so it isn’t a total failure.
Yeah, counterfeit parts are always a problem. It’s not just the sensors, either – everything gets counterfeited, including common transistors and ICs.
For example, I’ve seen some 2N3055 counterfeits that had a much smaller die in them that they should, considering their rated power handling capability…
“the MQ135 can detect hydrogen which it does well and this is given off by lead acid batteries during charging.”
Yes – it’s also given off during *normal* charging however, so the presence of hydrogen is NOT a reliable indication of a failing battery; it’s just an indication of a venting battery, and all lead-acids vent to at least some degree during charging, even the AGM/gel types; this is most noticeable during the constant voltage stage of charging.
H2S gas is extremely deadly in “high” concentrations – especially so because when deadly concentrations of it are present, it can’t be smelled at all: at such concentrations, it disables the smell receptors in the nose immediately, so there’s no indication that anything is wrong… at least, not until it causes a rapid loss of conscsiousness a few seconds later.
And in this case, “high” concentration means in the high PPM range; far below any kind of flammability limit.
I will try and swap out that sensor for the genuine thing if our budget allows. Also I’ve updated the code with a response curve (I used the same formula as my other gas detector) and it does work somewhat better. However without any gas samples calibrating it even reasonably accurately isn’t really possible.