When it comes to DIY projects, nothing beats the thrill of crafting something that rivals expensive commercial products. In the microphone build video below, [Electronoobs] found himself inspired by DIY Perks earlier efforts. He took on the challenge of building a $20 high-quality microphone—a budget-friendly alternative to models priced at $500. The result: an engaging and educational journey that has it’s moments of triumph, it’s challenges, and of course, opportunities for improvement.
The core of the build lies in the JLI-2555 capsule, identical to those found in premium microphones. The process involves assembling a custom PCB for the amplifier, a selection of high-quality capacitors, and designing lightweight yet shielded wiring to minimize noise. [Electronoobs] also demonstrates the importance of a well-constructed metal mesh enclosure to eliminate interference, borrowing techniques like shaping mesh over a wooden template and insulating wires with ultra-thin enamel copper. While the final build does not quite reach the studio-quality level and looks of the referenced DIY Perks’ build, it is an impressive attempt to watch and learn from.
The project’s key challenge here would be achieving consistent audio quality. The microphone struggled with noise, low volume, and single-channel audio, until [Electronoobs] made smart modifications to the shielded wiring and amplification stages. Despite the hurdles, the build stands as an affordable alternative with significant potential for refinement in future iterations.
Neat looking project. Any test data to back up claims of quality to the $500 microphone? I could not find any.
I think the claim was based on ‘uses the same capsule’ and ignored the bit where engineers solve physics problems with build choices within the electronics. Using high quality components is great, but the open source board is still getting revved and there are some places where additional care would resolve issues minor enough to be background noise in the current process.
I’m pretty confident that there is a path to success available with an open source mic preamp, but I’m not confident this video or the diy perks one gets us closer to it.
Both rely on a prefabbed cheapo dac for usb conversion.
It seems staying analog a bit longer down the chain might help as well.
I would almost prefer to stay analog and pull a mic level out for that new sine taste,
then also have a preamp option to pull up to line level (tube?)
I have a hard time with their aliexpress dac selection as well. I suspect the 500 dollar mic didn’t steal a easycap dac. I think you could probably shape your digital artifact noise to an inaudible frequency and perhaps use a better dac than the easycap and maybe bump up to a higher sampling rate.
I don’t pretend to be good at designing audio equipment, but as far as I’m concerned analog signal work is up there with rf. It’s black magic to everyone except for the guys who have the intuition to read and adjust the instrumentation during the dev iterations. The rest of us are operating on rumors and traditions.
Now that I think of it, maybe I’m all wrong, and I should just go test my ideas on a workbench before I rain on the parade of these guys who have done it.
One should not confuse adc with dac when talking about quality devices… lol… can I blame it on autocorrect?
Probably uses a capsule that is just labeled like those used in high end microphones.
It’s from Aliexpress, where pretty much everything costly is swapped with lower quality items relabeled as needed,from instrumentation OpAmps which are LM358 labeled as AD-something or low noise JFETs that either are plain low grade ones or sometimes even relabeled BJTs. Modules bought there are usually safe, although not ready for anything critical, industrial etc, but they usually work, however I wouldn’t touch any semiconductor coming from there, unless it’s designed and produced there from the beginning.
Check out the DIY Perks build. He has audio samples of his mic and this one.
As he says, this is just copied from another project, which itself had numerous problems and room for improvement. In fact, he says he couldn’t even get that design working, so made changes, yet I couldn’t find any dox for them. Still a crap design. Good idea though. Boggles me how some expensive mics use cheapo capsules. One day I’ll create an actual good design…
Honestly I dont think that capsule is used in anything thats actually high end. In reality finding a large diaphragm clone made by behringer and upgrading the circuitry on that might get you the nicest affordable device, simply because they actively clone the diaphragms instead of using whatever is available on aliexpress.
I recommend people buy like a Nat King Bee and be done with it because peoples homes are not studios and the acoustic treatment for a condenser is more important than the mic itself
Yup. I spent about 100 on a condenser mic, which I now pair with a zoom H4. It’s more than good enough for anything I’ll ever need, noise is low, and frankly a bit too sensitive for most things – it considerably exceeds my ability to make a room quiet, and unless you’re recording John Cage, there’s diminishing returns in a mic so good it captures all the ambient noise clearly.
This is my experience. A decent but not hypersensitive mic, with the talent’s mouth close to it, can provide a much higher signal/noise ratio in the presence of background noise or bad echoes or whatever.
If your mic can hear a pin drop across the room, it can also hear your refrigerator running. You don’t want that.
Max desired signal in, either drowning out or nulling out the sounds that you don’t want, is the name of the game outside of studio conditions. Turning the gain down, and getting more volume out of the source works wonders.
Totally agree! The cheap desk lamp arm shown is great for LED lights, a smoke fan, or a mic. They position right where you need them. I listen to a lot of radio not TV. There they are tight on mic but since covid in radio journalism a lot of builtin computer mics on desks with the cubicle or room acoustics are making for crap sound on a national radio program. If you do a video or are on TV I expect to see a mic near your face especially everyone in conference mode. Hold phone up to your mouth on a call to me no speakerphone garble please.
And/or a decent denoiser to blank the last remaining noise in the background, especially if you’ve got to have some significant agc or compression because you expect significantly varying speaking volume. The hardware accelerated ones that the big two gpu companies offer are fine, or there’s rnnoise, and you can definitely get it in the form of a vst plugin that you can put inline with a virtual audio cable or suchlike if you are doing a live call.
Concept is good, but he had a lot of unresolved noise problems in the end. I saw a few ‘noob issues in the build: at 6:00 the case pin of the FET was cut off, would have tied to the signal shield. at 8:50 noise was reduce with grill attached, but there is still noise on the scope. DC supply and audio grounds need special attention to keep any DC convertor current out of the audio signal, and probably the main source of the noise.
Putting that noisy USB board in the shielded case with the amplifier is not a great idea either.
Absolutely!