Here is something you can try if you want to produce a beer sample that almost certainly has diacetyl concentrations above the typical flavour thresholds. During fermentation, wort has a very low concentration of free diacetyl, which means you typically don't detect it in fermenting wort. However, the wort will at this point (especially early into fermentation) contain a large concentration of alpha-acetolactate (which will eventually decarboxylate into diacetyl). I've e.g. measured peak 'total diacetyl' concentrations of 300-1500 µg/L in recent years depending on yeast strain and fermentation conditions (while the threshold is in the 20-100 µg/L range). By heating the beer, we can increase the rate of the decarboxylation reaction, so that the alpha-acetolactate is converted into diacetyl in minutes instead of days. So:
- Take a small sample from your fermenting beer early during fermentation (2-3 days after pitching).
- Place the sample in a closeable container (e.g. beer bottle that you cap, soda bottle, etc.). You don't want the volatile diacetyl to escape once it is formed. Keep in mind that the sample will be heated, so be careful with anything that can be damaged by pressure.
- Heat the sample to 60 degrees C (140F) in a water bath (e.g. on the stove) for an hour.
- At this point your beer sample should contain a high concentration of diacetyl, and you will hopefully be able to detect it.
I haven't tried this myself at home, but similar protocols (e.g. in EBC-Analytica, ASBC MOA ..) are used during 'total diacetyl' analysis. I'm not sure how any suspended yeast will be affected by the heating though (probably won't have an effect), as in the lab we centrifuge and filter our samples prior to analysis.
Edit: Oh, and when deciding on when to take the sample, typically the higher your fermentation temperature, the earlier the 'total diacetyl' peak will occur.