90u or 120u: which micron for a CBD hash?
The number names the mesh of the sieve that held the fraction back, in microns, not the cannabinoid content. A 90u mesh is narrower than a 120u mesh, so it is more selective and less generous in yield. Mature trichome heads sit in a window of 70 to 120 microns: that is why the 73u, 90u and 120u meshes exist. The level itself is read on the batch certificate.
Contents
The question comes back in several shapes: 90u or 120u, 73u or 90u, should you move up or down the scale. This guide answers in three stages: what the number names, what each mesh holds back, and the four things the micron will never tell you. The size measurements come from a study published in 2023, the bag scale from a supplier to the trade, and the levels from the Phytogrammes catalogue as recorded on 16 September 2026.
> Key points > > - The micron is a sieve mesh, not a content figure and not a grade. > - The smaller the number, the more selective the fraction and the lower the yield. > - The 73-120u window concentrates whole trichome heads. > - No standard governs the notation: "120u" and "90u-120u" may describe the same jar. > - Two 90u references in the catalogue test at 15.1% CBD and 19.3% CBG: same mesh, opposite profiles.
1. What 90u and 120u name
"90u" reads as 90 microns, that is 90 micrometres: the size of the openings in the sieve or filter bag on which the fraction was collected. The 120u lexicon entry is the reference page for that definition and for the full scale of meshes; it points out that the smaller the micron, the more selective the fraction.
The number therefore describes a sorting operation, not a substance. It applies equally to ice-water separation, where stacked bags filter a suspension, and to dry sifting, where the powder passes through tiered sieves. The catalogue's Drysift 90u listing does indeed define dry sift as the powder obtained "by dry sifting of the hemp flower", the 90u there naming the mesh that held it. Detailed definitions of the two routes are on the dry sift and ice-o-lator pages.
Why this window rather than another? Because it matches the size of the objects being sought. A study published in 2023 in the Journal of Cannabis Research, which measured the glandular trichomes of cannabis plants at different ages, reports glandular heads between 40 and 110 microns across, and bulbous trichomes of 15 to 30 microns. The lexicon, for its part, gives a working window of 70 to 120 microns for a mature head. The two ranges overlap on the essentials: the useful meshes sit between 73 and 120 microns.
2. The scale of meshes, bag by bag
A wash does not produce a single fraction. The suspension passes through several stacked bags, from the widest to the narrowest, and each bag holds back a fraction that the extractor weighs and judges separately. The standard set runs to eight bags. The table below takes the role assigned to each by The Press Club, an equipment supplier to the trade.
| Mesh | Role in a set of eight bags | What it holds back |
|---|---|---|
| 220u | Main work bag | Lets the coarsest material through |
| 190u | Cleaning bag | Refining, nothing to collect |
| 160u | Contaminant removal | Low quality, rated 1 to 2 stars |
| 120u | Collection | Good quality, trichome heads |
| 90u | Collection | High quality, close to whole mature heads |
| 73u | Collection | The highest grade in the set |
| 45u | Collection | Secondary fraction, larger yield |
| 25u | Collection | Low quality |
Two lessons come out of it. First, the high meshes do not collect: 220u and 190u are there for the work and the cleaning. Second, finer is not mechanically better. Below 45 microns the fraction fills up with immature heads, trichome stalks and dust; above 160, plant debris goes through with the resin. That is exactly what the lexicon notes, and what the supplier's table reflects.
The method itself is less than thirty years old. Reinhard Delp presented his ice-water extractor at the Cannabis Cup in November 1997; Mila Jansen's Ice-O-Lator bags were launched at the 1998 Cannabis Cup, and Fresh Headies' Bubble Bags in March 1999. The micron scale as it is read on labels today dates from that period.
3. 90u or 120u: what really changes
In the catalogue, two references illustrate the comparison. They differ neither in their legal standing nor in their analytical method, but in their mesh, their variety and their aromatic profile.
| Criterion | 90u (Drysift 90u) | 120u (ice-o-lator) |
|---|---|---|
| Stated mesh | 90 microns | 120 microns |
| Dominant cannabinoid (HPLC) | CBD 15.1% | CBD 15.9% |
| Unrounded value on the listing | 15.11% | 15.86% |
| CBG | 1.0% | 1.3% |
| CBC | 0.3% | 0.5% |
| CBN | 0.3% | 0.4% |
| Δ⁹-THC measured | 0.2% | 0.2% |
| Aromatic profile on the listing | peppery and dry, α-pinene and β-caryophyllene | floral and herbaceous, linalool and α-pinene |
| Price from 2 g to 100 g | 7.95 to 1.90 €/g | 7.95 to 1.90 €/g |
| Batch number | PG-26-DBD7 | PG-26-A7E3 |
The gap in level between the two batches is eight tenths of a point, in favour of the wider mesh. The price per gram, on the other hand, is identical across all six tiers. The Drysift 90u listing and the 120u listing are the purchase pages for those two batches; that is where the analysis and the tiers in force appear. The lexicon entry remains the definition page for the mesh. The two roles are not interchangeable.
4. What the micron does not say
Four limits are worth knowing before choosing on that criterion alone.
- The cannabinoid level. The proof comes from a pair of references with the same mesh: the Drysift 90u tests at 15.1% CBD, while the Hash CBG 90µ, sifted to the same fineness, gives 19.3% CBG for 4.2% CBD (batch PG-26-83EF). Same number, different dominant cannabinoid.
- A steady ranking. In the catalogue, measured levels run from 14.8% for a 45 micron fraction to 15.1% at 90 microns and 15.9% at 120 microns. Fineness does not push the level up; on these batches it would tend to push it down.
- The melt grade. Full melt describes a hash that melts completely under heat, leaving no plant residue. The lexicon makes clear that this grade is a verdict from observation, awarded by the extractor or the reseller, and that no laboratory issues it.
- Compliance. Only the batch certificate establishes the Δ⁹-THC. The 0.30% limit is set by the arrêté du 30 décembre 2021 (French order of 30 December 2021), which applies it to extracts and to products containing them; on 29 December 2022 the Conseil d'État (France's supreme administrative court, decision no. 444887) annulled the ban on selling raw flowers and leaves, resins falling under the regime for extracts. These products are for adults only and are not foodstuffs.
The micron does keep its value as an indication, though. A producer who states it is committing to say how the resin was sorted, and the number can be set against the texture observed. What lacks seriousness is a micron on its own, with no analysis and no batch number beside it.
5. Choosing between two meshes, in practice
The question "90u or 120u" is usefully reframed as three checkable questions. What is the unrounded value of the dominant cannabinoid on the batch offered? Is the Δ⁹-THC measured below 0.30% by a laboratory accredited to ISO/IEC 17025? Does the batch number on the certificate match the one on the label? All three answers appear on the product page, never in the number of microns.
Then comes the aromatic register, which depends on the variety and not on the mesh: peppery and dry on one side, floral and herbaceous on the other for the two batches compared here. The our laboratory page explains how to read the four blocks of a certificate, and the CBD resin and hash page gathers the catalogue's references. For a comparison between families of process rather than between meshes, the frozen, piatella, 120u and static comparison and the comparison of the three ice-o-lator 45u, 120u and Caramelo use the same grid.
Frequently asked questions
90u or 120u: what is the difference?
The sieve mesh. 120 microns holds back whole trichome heads, 90 microns a finer and denser fraction, therefore more selective and less generous in yield. Neither number indicates the content: in the catalogue, the 90u batch tests at 15.1% CBD and the 120u one at 15.9%, measured by HPLC.
Why does the 73-120u window come up everywhere?
Because it matches the size of the trichome heads being sought. A 2023 study measures glandular heads of 40 to 110 microns across, and the lexicon keeps a window of 70 to 120 microns for a mature head. Below 45 microns the fraction fills with dust and immature heads; above 160, plant debris goes through.
What does 220u mean on a set of bags?
It is the work bag: it lets the coarsest material through and is not used for collection. In a set of eight bags, 220u and 190u handle the work and the cleaning, 160u removes contaminants, and the meshes from 120u down to 25u collect the fractions.
Is a hash sold as 120u really 120u?
No standard governs the notation. The lexicon notes that a jar sold as 120u may correspond to that mesh or to a blend of two neighbouring meshes, and that "120u" and "90u-120u" may describe the same contents. The micron remains an indication of transparency; what can be verified is the batch certificate.
Sources
- Phytogrammes, lexicon "120u (microns)"
- Phytogrammes, lexicon "Dry sift"
- Phytogrammes, lexicon "Full melt"
- Phytogrammes, Drysift 90u listing, batch PG-26-DBD7 (recorded on 16 September 2026)
- Phytogrammes, 120u listing, batch PG-26-A7E3 (recorded on 16 September 2026)
- Phytogrammes, Hash CBG 90µ listing, batch PG-26-83EF (recorded on 16 September 2026)
- Journal of Cannabis Research, 2023, morphology and maturation of glandular trichomes
- The Press Club, the role of wash bags by mesh
- Leaf Magazines, "Cannthropology: Bubble Hash Backstory" (1997, 1998, 1999)
- Légifrance, arrêté du 30 décembre 2021 (French order of 30 December 2021, article 1, I and III)
- Conseil d'État, decision of 29 December 2022, no. 444887
- ISO/IEC 17025, testing and calibration laboratories
Written by the Phytogrammes team
Every article in this journal draws on primary sources (ANSM, EFSA, EUR-Lex, peer-reviewed publications) and on the lab's own practice: batch-by-batch HPLC analyses, measured cannabinoid profiles. Our approach.
Article published on . CBD is not a medicine.