I was thinking for a while to create a booster that is as small as possible. Ever since I bought an old guitar that was sitting unused on someone’s attic for 15 years I was thinking what could I do with it.
I have a thing for Ibanez guitars … I don’t know, I just couldn’t let the guitar go on a skip. In the end, buying tremolo arm for it was more expensive than the guitar 😣. It is great thus for my own little experimentation.
All the boards in this article were designed by me, and I got prototypes for free from PCBWay – thanks for this fellas, great service as usual 🙂.
You can skip straight to the video if you wish, below are some extra details of the build.
Intro
I did a bunch of booster pedals before, so I had a sort of a template of what I was going for. What was different this time was that I wanted to make this as small as possible, battery operated and get it to use as little power as possible.
The reasons for the above are – I wanted to be able to put it into my guitar’s body – hence small … and battery operated. And since I didn’t want to change the battery for as long as possible, low power usage was essential.
Natural choice for the battery was 3V coin cell battery. I wasn’t really thinking beyond that. I might in future consider something else, but one of the design goals was using a 3V coin cell battery.
Since the size of the battery is the minimal size I can get away, I thought, that’s the size of the board I’ll aim for. Naturally, I’ll need to use SMT components, and I’m no stranger to SMT.
Finally, I thought that the simplest and the most efficient design would be with an op-amp. I was looking for something preferably rail-to-rail to give me maximum head room, and something that will be very low power when idle.
There’s a few op-amps I considered and I settled for giving a try to: OPA379 (OPA379AIDBVT) and TLV313 (TLV313IDBVT) both in tiny SOT23-5 package. For this build I used OPA379, I’ll try the other op-amp later.
Schematic
Here’s the schematic:
Pretty simple. It’s really just a simple non-inverting op-amp design. Nothing special to be honest. The biggest thing was figuring out which components to use.
If you find some of the values odd – I had leftover resistors and capacitors from earlier so I designed the circuit around them. A conscience decision was made to keep gain lower and fixed (about 10dB doh! that’s rather about 12dB and that’s higher than some commercially available boosters … might need to lower the gain for the next design).
I used a through hole version of coin cell battery holder, I’ll change that to surface mount for the next version of the design.
The op-amp I used is not really designed for audio so we’ll see what we get out of it. I’m pretty certain it will work, how good though, we’ll see. This is a proof of concept more than anything else, so another iteration is coming for sure.
Consumption
One of the goals was to use an op-amp that can last a while on a single battery. If we assume that 3V coin cell has capacity of about 500mAh, let’s see what can we get out of our battery. This is a very crude calculation and I probably got it wrong on the video.
Quiescent current per amplifier is maximum 5.5uA. Our chip has only one op-amp, so, in theory, if the op-amp is idle, we should squeeze about 90,000 hours???? Is this right? That’s over 10 years. (Typical consumption is half of this so … 20 years!???)
Reality is that I used 100Kohm resistors for biasing, so they alone use 10uA … doh! That means the battery should last around 3 years. Again, this is not entirely true, because the op-amp won’t be completely idle in reality.
One thing for certain, it beats my Vox amPlug 2 … that thing is eating batteries. Unfair comparison … I know, amPlug has a bit more going on, but still.
If anyone knows how to better calculate consumption, please chime in the comments. I wonder how far off I was with this super crude calculation.
Noise
Another thing to consider is noise. When I compare this OPA379 to TL072, it will be probably more noisy (just by looking at the parameters). All I can hope is that it will be good enough 😊.
| TL072 | OPA379 | |
| Input voltage noise density (nV/√Hz) | 20-37 | 80 |
| Input current noise density (fA/√Hz) | 80 | 1 |
| Gain Bandwidth Product (GBW) | 5.25 MHz | 90Khz |
Current noise density is negligible compared to voltage noise, but voltage noise might be a deal breaker in the end.
Also, it’s worth noting that GBW is way wider for TL072. For OPA379, I chose 10dB gain in part because of GBW. Closed loop bandwidth is about 30Khz. When I plug-in noise figure for the noise bandwidth and include noise contributed by resistors … I get 0.3mVpp … for the equivalent bandwidth TL072 would have 0.2mVpp … I have no idea what to take of these numbers though :))))
Final Result
Anyway, here’s the final result:
Looks pretty much the job!!! The finished board looks the same as the 3D rendering from KiCad. Well, my soldering is suspect, but the goal – size-wise, complete success!
Conclusion
In the end I wasn’t really happy with the result. While I did manage to make it work, the circuit is too noisy. I am not entirely sure if I messed up something, maybe it’s my soldering. TLV313 has comparable voltage noise to TL072, so maybe I’ll have more luck with it. I haven’t given up on OPA379 yet, so stay tuned for more messing around with it.