Still remember the Extreme Capacitor Battle article? Well, it’s not done yet, but we are very close on it. The break-in session has reached around 180++ hours. I can feel the characteristic from each capacitor.

To simplify (and easier) comparison session, I have prepared two ELMA switch (2 poles, 6 positions). You can see the photo below.

You might ask, why do I need two pieces of this ELMA? One should be sufficient, right? Keep reading. As you may also notice that I have a “bad habit” as a perfectionist, so this is still related with it 😉


This ELMA switch has gold pins to ensure “better” connectivity and also quite “precise” and “smooth” switching mechanism. Perhaps this is not the best one, but surely this is more than enough for our project now!


I check the continuity of each position from the switch. Seemed ok! I use my old Sanwa Analog YX360TRF Multitester. I know it’s quite strange to use such vintage meter for this purpose, but I love to see the smooth swing from the needle of this meter (compared with small buzzing sound from my digital Multitester).


Now, back to the technical analysis. You will ask, why I don’t use method as below. It’s cheaper since only use 1 switch only. Why do I prepare 2 switches for this purpose? Keep reading…


With method above, yes, I can save money by using 1 switch only. But another problem will rise from this configuration. Check the “Signal Flow Analysis” below. I draw a small diagram to easier myself in explaining what’s gonna happen with that configuration.


The output of all capacitors will be mixed together. As you can see from the diagram above. The real signal is marked with green shadow at first capacitor. But since all the capacitor will be connected together at the output side, I can guarantee, the signal will also enter the other capacitors (the 2nd and 3rd capacitors, marked with yellow shadow). The signal will travel to the 2nd and 3rd capacitors, then return back to the output (since they are not connected at the input side). The result is, not only a single capacitor is “singing”, but the rest of the capacitors also will contribute on the signal output (although the contribution is quite small, but it will “give coloration” on the original signal – the 1st capacitor – and makes more difficult for us to distinguish the different between capacitor on test). I don’t tolerate this, so I draw another configuration as below.


The perfect schematic for comparing capacitors! Each capacitor on test will have their own signal path. We need to switch the front and back switch on same position to make sure the connectivity of the signal path. If you want to test the 1st capacitor, then you will need to move the front switch and back switch to 1st position, etc. Otherwise, you will not be able to create a connectivity.

This is the perfect configuration and you have to remember when creating a comparison jig. This is not only for capacitor comparison, but also applicable for other, let’s say an interconnect cable comparison.

Story from my own:

I have ever experienced myself few years ago, between the “cheap” and “perfectionist” design above. That was the interconnect cable comparison. We were surprised that “average” and “more expensive” cables could sound so similar in “cheap 1 switch” configuration  above. It’s hard to distinguish the different. They were different, but subtle to hear. Before, we have tested the cable independently (without switch) and the different was very clear. But due to the comparison test that forced us to use a switch to easier the A-B comparison, then we have to use it. After we changed to the “perfectionist 2 switches” design, then we could really hear the different between those two cables, just like we used each cable independently.

Your miles may vary! Trust your own ears!