The previous 9V batteries were carbon-zinc batteries, and due to their insufficient output, they didn’t function properly. The batteries I purchased this time are alkaline batteries, which provide better output than carbon-zinc batteries. Thankfully, they work well. The output is sufficient to make the speakers strike the metal plates effectively, and the sound is loud enough. Additionally, the coding patch I previously set up (Ableton – Max – Arduino) works seamlessly, and the speakers respond perfectly to the patch. So, this time, I connected the Arduino and the breadboard holding 12 speakers to test the patch. Everything functioned as intended, with proper output, and even the part where the speakers vibrate six times per second worked without any buffering.
However, during the experiment, the batteries started heating up. This was a recurring issue from before. Previously, the heat was somewhat tolerable, but as I ran the sequence I created, the batteries went from cold to warm in less than a minute. I began to suspect overheating (which is a critical concern as battery overheating can pose a fire hazard). After urgently researching, I found out that the cause was excessive current being drawn due to the batteries providing less power than what the speakers require. For reference, I’ll detail my findings here.
The speakers I am using have a maximum output of 60W and require 12V or 24V to operate.
60W / 12V =5A
This means 5A of current is required. However, the 9V alkaline batteries I am using provide only about 0.05 to 0.2A of current. In other words, my batteries fall far short of the 5A needed, and even if I attempted to increase the current by connecting the batteries in parallel, the total capacity of a typical 9V battery is only 500–600mAh, which would barely last six minutes before the batteries are completely drained. Fortunately, I realized the overheating issue quickly, avoiding the risk of starting a fire at home.
After researching possible solutions, the only viable option was to use better batteries. While I could use a DC-DC converter to boost the voltage and minimize current loss, this would not resolve the overcurrent issue. To create a parallel connection, I would need more batteries and a more complex circuit to ensure stability. However, the best battery I knew of was lithium-ion, which was not ideal for several reasons: a 9V or 12V lithium-ion battery costs around £20 each—an expensive option—and lithium-ion batteries are prone to fire or explosion due to risks such as over-discharge, overcharge, overheating, and physical impact. Thus, I was hesitant to choose this route.
While continuing my research, I discovered nickel-metal hydride (NiMH) batteries. NiMH batteries are rechargeable, like lithium-ion, but much safer, offering similar performance. For example, a single 1.2V NiMH battery typically has a capacity ranging from 2000mAh to 3000mAh, which suited my needs perfectly. However, since I needed 9V, purchasing 9V NiMH batteries was costly. Instead, I decided to buy AA-sized NiMH batteries and connect them in series using a battery holder.
There is a simple yet critical difference between series and parallel connections:
- In a series connection, the positive terminal of one battery is connected to the negative terminal of the next, increasing the total voltage.
- In a parallel connection, the positive terminals are connected together, and the negative terminals are connected together, increasing the total current.
NiMH batteries have very low internal resistance, allowing for higher current output. Additionally, AA-sized batteries, due to their design, have sufficient capacity for extended use.
Even after deciding to use NiMH batteries, I encountered additional issues:
- Back EMF from the relay module:
Back EMF could potentially damage the circuit. While this hasn’t caused problems yet, it could occur while running my sequence. Back EMF happens when the relay module cuts off power, causing a very brief but extremely high voltage spike (100–10,000V). This spike can travel back through the circuit, potentially damaging the Arduino or even the power-supplying computer. To prevent this, I need to use a reverse-biased diode, which redirects the current spike safely to the ground. - Battery capacity and sequence duration:
Running my 32-minute sequence continuously could strain the battery capacity. While high-capacity NiMH batteries (3000mAh or more) could handle this, they are approximately twice as expensive as 2000mAh batteries. Due to budget constraints, I decided to purchase 2000mAh batteries and reduce the sequence length. Instead of the original 32 minutes, I plan to shorten it to about 16 minutes. Additionally, the 32nd sequence, which required the speakers to turn on and off 24 times every three seconds, posed a high risk of overheating the relay module. Hence, I decided to limit the sequence to 16 rhythms (composed of a maximum of three notes per rhythm) while maintaining the repetition counts. This adjustment would reduce the total duration to approximately 16 minutes. Despite the shorter sequence, I expect to still experience enough rhythmic diversity and will test this in Ableton.
The batteries, battery holders, and other components I ordered are expected to arrive on Monday. Tomorrow, I plan to refine the sequence further, organize the wires neatly with the insulation tape that arrived today to ensure noise-free operation, and calculate the required length of copper wire for each speaker based on the planned arrangement. Finally, I will connect the speakers and inspect for any power loss issues.