Sonic Cold Brew... Coffee with Sound

I love coffee. Too much.
I like cold brew. But it takes forever to make.
But that’s why it’s so delicious… right?
Let’s find out.

I had all the gear already, so why not shake up our grounds for several hours by sticking a transducer directly on the container?

I tried it last week, and the results were… disappointing.
It was a fast and loose test; I used some grounds that I already had, just to use 'em, but they were much too fine, and the kicker was that the transducer I used was only rated for ~5w. I gave it too much power for hours, and accidentally brewed the coffee because it overheated so much. :stuck_out_tongue:

Yesterday, I obtained a new transducer that can handle more, so today, I’m giving it another go. Fresh, coarse grounds; two brews from the same batch of beans. One is a control pot in the same room, to rule out ambient temperature differences.

The math is simple.
To generate a signal, I found an oscillator on the play store and hooked up the amp to a tablet that I can leave for a couple hours.
Using a sine wave, I sweep the spectrum while touching the pot to approximate resonant frequencies.
Find a good resonant low freq, in this case ~240.
Find a higher one, in this case ~505.
Then, I change it to a square wave for all the added harmonic content, and repeatedly sweep it up over a half second.

The grounds vary in size, so I took a shotgun approach, hence the sweep.
Choosing a lower constant freq didn’t give me quite the vibration I wanted, especially in the harmonics.

I duct taped on the transducer to get as much energy transferred to the coffee as possible. Unfortunately, that was kind of the best of the worst ways to attach it, at least for now.

I also put it on sorbothane pads and put the assembly by itself on a wooden stool. The stool is mainly not to annoy the neighbor, by providing just enough isolation not to hear the whole thing through the ceiling (neighbor above).

So… what will come of it? I’m not sure. As I understand it, making cold brew has to do mostly with lowering the brew temperature to reduce bitterness, but sacrificing the time it takes to saturate. So, if that’s all it is, then I suppose it could work. It could also be the case that I’m still not pumping enough power into it, and it needs paint- shaker levels of power to be effective.

The hardware I’m using is basically a 12w amp, and a 20w transducer. Interference on the audio cable is of no concern, so it’s nothing fancy. Just some cheap stuff I got from amazon a while back for another purpose.

I’ll check it in a few hours, and we’ll find out.

7 Likes

Yes, that’s right.

When overexposed, due to being ground too finely (typically with a blade grinder, so it’s uneven), brewed too long, or brewed in water that is too hot, coffee is bitter. When underexposed, it’s sour.

That’s why you’ll often find bad cold brew to be sour but never bitter. Of course you can add sweetener to counter bitter but there is no real way to counter those sour notes. I always use dark roast beans for my cold brew as they have less sour flavor compounds to extract in the first place.

Your theory is that cold brewing takes longer in part due to lack of agitation, and I suppose there may be some truth to that, but I’m dubious whether shaking the entire container will produce a better result than a laboratory magnetic stir-bar. Makes for cooler pictures, though, for sure!

1 Like

I love cold brew but always forget to make it the night before… I’m off to go make some for tomorrow morning. Back in a few minuets…

One week. Good stuff OP.

1 Like

Stuff naturally breaks down and dissolves quicker in warm/hot water. So that makes sense that it is slower to do cold. Takes longer to penetrate and leach out the chemicals.

Yes, but I was not talking about temperature.

Updates

The last attempt turned out to be a bit of a misfire. The transducer ended up overheating again, though not as badly, and I stopped it sooner. I also used a coarse-ground light roast, which, in retrospect, should have been dark, following what @Ruffalo said.

I also cracked the glass container trying to work it out, go figure. Next attempt will be in the french press, which is double-walled stainless steel, so that should take care of the transducer heat, while still being able to carry the high freqs all the way to the water.

As for the results…

(Until I can upload a better pic from a better cam)

From left to right:

Sonic brew, 45-60 mins
Control cold brew, 11hrs
Control cold brew, 21hrs

Looking closer, the sonic brew, while cancelled prematurely, has what appears to be an equal amount of both oil on top, and sediment suspended throughout.

Taste-wise, the sonic brew is watery, as expected, although, the “taste-delta” between the sonic brew and the 11hr brew is about the same as the delta between the 11hr and the 21hr, which is the superior brew. The difference, though, is small. It’s something that may be at least a little tough to sort in a blind test (couldn’t get anyone to do it).

I expected less taste overall, but other than the strength, the taste is exactly the same. The thing I can’t effectively guage is caffeine content. Fortunately, none of the 3 were bitter, at all, so it wasn’t a complete failure.

I think I’ll try again this next week with a darker roast, maybe a finer grind, or less water (this one was 3:1, already pretty rich). For reference, I use a hand-cranked burr grinder. I also need to find a better way of mounting it. The transducer needs some pressure to be effective, but it’s attaching to a smooth, steel surface, and tape around the back won’t last an hour.

But man, imagine… get up at 6, start a brew, leave by 8 with a cold brew that’s 90% ready. Sounds nice.

New attempt in the french press.

Went without a lid for about the last 1:45 (added zipties after 15 mins, don’t know why it took me so long to realize this), then added the filter and lid, and remounted the transducer to that.
The filter is in the water, so as long as it stays on, it should be more effective in direct agitation than side mounting. In both cases heat transfer from the transducer is negligible, but the ambient temperature isn’t ideal. It’s hot here.

It’s a fine grind, this go round, and it’s the last of the light roast.

The control group is decidedly better than last time. Brew time is about 10 hours.

If it works out, I think I can add a hydrophone that detects the best freq for a given volume and auto adjusts based on amplitude during a sweep.

1 Like

Was it sour? Never had light roast cold-brew that wasn’t sour.

And I’ve got to say, 10 hours is not a great time-savings.

I suggest trying a simple magnetic stir bar, it won’t be as cool as using sound waves but should be much easier and offer a comparable result.

You could also experiment with heat. Common knowledge is coffee flavors develop best around 90C, but what if you were to agitate and brew at say, 55C for an hour? Would that lead to a product without sour or bitter notes?

This also has the benefit of being hot enough to kill bacteria that could result letting beans sit in room-temperature water overnight.

It has sour notes, but I wouldn’t call it sour. That was the last of the bag, so next attempt will be a dark roast. I was using what I had until I needed more.

10 hours was the control batch for comparison. The sonic brew was 2hrs.

It might. The room I leave it in sits around 27-30º C ambient (because it’s a poorly insulated room in an old house in Texas). Adding some heat might help since the sonic brew always feels at least a little watered down; however, the first attempt where the small transducer overheated basically brewed it with low heat for a few hours, and it tasted terrible, so my guess is no, but under the right conditions, maybe.

I might also try some distilled water + additives. I’ve been using tap water for every attempt thus far.

That’s definitely something I’m considering. The only reason I haven’t is that I don’t feel my silly experiment is worth the price of the stirrer. I might pick one up; still undecided.

As for the results

The 10hr brew was still better. The color was almost identical. There was a fraction more sediment in the control brew, and the same amount of oils visible on the surface.

Taste was very similar, but the control still has the complexity and depth that the sonic brew just can’t manage. They are closer than the last run, but I’d say the sonic brew has only about 85% of the strength that the control has. The brighter notes are exactly the same, but the deeper, darker flavors still taste watered-down. Cold brew is all about subtle complexities. I think I can improve it by finding a way to move the grounds more, rather than the container. The stir bar looks like an obvious solution, and it might come to that. We’ll see. What I really want from all this is for the grounds to oscillate in place. Shake the flavor out of 'em!

Next attempt will be focused on:

  • vibrating something (probably the lid + filter) suspended in the solution
  • more power to the transducer
  • static frequencies, rather than sweeps

Find the resonant frequency of you container with the coffee and water in it so it makes standing waves that agitates everything. Though I am not even partially versed in this at all, just makes sense.

Just had a quick chat with a local shop owner, he is a firm believer that it cannot be done quickly. I’m gonna try a few things anyway, because 3 hours is better than 12, but that doesn’t help me find the magic.

I may try a 12hr sonic brew and see how different it can be than a normal 12hr cold brew.

Logically, he should be wrong. Standard cold brew is a bunch of grounds sitting in water, over time they settle to the bottom of the vessel thus limiting their exposure to water. Agitation should ensure that every grain is exposed and thus should reduce the amount of time required. It should be a measurable difference. Question is whether it makes a substantive difference in either required duration or production quality.

1 Like

I wonder if ionization in a closed container would be a waste of time. It’d be similar to an oil diffuser, except with little to no volume loss. Silver lining, it’d be a supersonic brew.

Could also toy around with a 2:1 coffee to water, though that may be excessively wasteful since there would still be no temperature change to increase saturation.

It didn’t occur to me until recently that combining the egg method with this may work. Will give it a try if/when I get some time.

3 Likes

This thread makes me happy.

3 Likes

This has been an amazing read so far. Great idea and work!

The egg trick works very well.

2 Likes

Sounds like this, minus the egg.

1 Like