Everything I’ve learned so far about the EK43 and Espresso

I find myself in the slightly unexpected position of being repeatedly asked questions in coffee shops by baristas (and increasingly shop owners!) about the EK43 for the creation of espresso. (or EKspresso as it is known as).

I have spent the last half year making nothing but EKspresso and this is a dumping of what I have learned so far, if you are already making EKspresso shots happily then this blog post probably is not for you unless you want to teach me some more things in the comments.

So – there are rumours that this grinder is amazing for making espresso with as well as brewed coffee, and the increasing number of people with an EK43 available to them still mostly only use it for the brew. While they have probably tried a few shots through it out of curiosity it is hard to put in the time and research required to do know it well enough to achieve “properly” when you’re also running or working in a busy cafe/roastery/etc.

Being my only grinder I’ve had to actually commit to it and I’ve found the end result to surprise even me after spending many months perpetually chasing the kind of coffee I used to drink at places like Macintyre’s and then finding out that the new coffee burrs give an even better result than I could have ever dreamed of.

This is the result of many peaks and troughs of “aha” moments and “why can’t I consistently do this right, I should give up now EKspresso is clearly too hard for a mere mortal such as myself” momentsIf you can do me a favour and read most of the statements I make in this blog entry with an “in my opinion so far” in parenthesis beside them then I’ll do you a favour and strip that disclaimer out of the rest of this text.

This blog entry is long and split into several sections if you want to skip down to the appropriate one

The Goal of EKSpresso At Cafe Ashton

Consistent sweetness and clarity of flavour. No muddiness at all, acidity is welcome but it shouldn’t make my tongue roll up and hide in a corner – without at least some sweetness espresso is nothing. Higher EY isn’t necessarily everything but a more even extraction will often mean a higher EY and “more” flavour – it is definitely part of the equation. Consistency is everything – it’s no good loving one spro and then hating the next – I have to know my guests are going to be blown away by their spro.

If these goals aren’t your goals then carry on reading any way, I might change your mind.

Things we need for a good EKspresso shot

The beans

A well roasted coffee these days is one which we can push the extraction up with and not get horrible roasty notes or bitterness from; annoyingly this makes them harder to dial in as well as without practise we can’t easily identify this “over” extraction by taste, we just shoot past our optimal extraction and go right back to under-extracting again. A reasonable level of solubility is useful too as it means extracting more faster and less chance of muddy results.

For already-trained baristas this is probably easier than it was for me – probably, if they’re trained on good beans but most baristas haven’t ever even tasted a well extracted espresso on high quality beans in their entire life (I am convinced this is why so many people are so happy with the piss-poor dribbles from those miserable mazzers).

We do trust our taste-buds way too readily and enjoy what we already know and are familiar with – it’s not until we have had a few stand-out EKspresso shots that we realise that everything we have ever tasted has been bad and we have been “doing it wrong” our entire our lives ;-).

We do enjoy what we know though so don’t go down this journey if you’re not happy to accept that once you’re able to repeatably make good EKspresso you’ll likely start disliking coffee beans from roasters you used to really enjoy and find even well dialled-in Mythos shots start to fall short of your high expectations.  I have no doubts there are going to be many more improvements over the coming year or so and that some of the things I’ve written here will be entirely invalid – tea sales at coffee shops that don’t keep up are only going to increase then.

The burrs

Coffee burrs. New coffee burrs. If the EK43 you have access to is over a year old at time of writing it probably has the old coffee burrs.  I have had amazing shots from the old coffee burrs but it is not what I’m writing about here; they pour far more rapidly and messily and are far harder to be consistent with.

People will throw around remarks like “You need turkish burrs to get a decent espresso pour” and these people are incorrect. Before the new coffee burrs this was sort of true and there are still some benefits to those turkish burrs even now – but even well dialled-in spro on turkish burrs tend to tastes muddy to me now. Aiming for an espresso pour that looks just like the other espresso pours from other grinders is a weird goal when we think about it – focus on the end result only.

This is a whole new world of coffee, we should take nothing less than the new coffee burrs with we when join it.

The measurements

Get a refractometer, I wish I had done so sooner as it was completely transformative and saved me weeks of effort in the first two days of owning one. This wasn’t entirely an artifact of me being a completely ignorant amateur – several baristas I know with many, many years between them have also had major “oh man how did I do without this” moments once they started refracting.

We don’t need to buy a VST refractometer, an Atago one will do (miles cheaper) and we don’t need to use those expensive filters – we’re not doing science, just cool down and mix our liquid well and get it on the sensor pronto, repeat for a dozen shots across a spread of times and find the peak, now we understand our coffee better than we did before. We don’t even need the app, just the basic maths of (tds * brew weight) / dose.

Edit: Lots of questions about this – I’m not saying filters are useless by any means, but when comparing like-for-like and finding a ‘peak’ where extraction starts to fall again it’s good enough to do without. Want accurate results to  share with others online? Use a filter. Doing science? Use a filter. Casually dialling in some spro, don’t bother. Hope that clears up this sentiment.

When we cost up the time we spend (or can spend) dialling in coffee and look at how much time a refractometer can save us it simply doesn’t make sense that we can spend so many thousands of dollars on fancy machines to make coffee with and then not afford the few hundred quid a basic Atago would set us back.

I don’t tend to use the refractometer to dial in my daily coffee any more because it has given me so much insight already and I can usually taste the tipping point where my spro goes from under-extracted to unevenly extracted (or more rarely: over-extracted), but as a problem solving device when a new coffee comes in it is still worth its weight and more in gold. My house is littered with A4 sheets of paper with “dose, time, yield, tds, ey) written on them.

An open mind, but not too open

One of the biggest issues I encountered when learning how to do EKspresso was the lack of experience anybody around me has with it themselves. This meant when describing an espresso recipe of “22seconds for a yield of 60g” I got serious looks of distaste from the baristas I was talking to. This meant a real crisis of confidence regularly and it wasnt until I got the refractometer and was able to back up this sort of recipe with numbers that I was truly comfortable just letting my taste buds do the work for me there. (Try and do a 60g shot without getting an EY over 20%…).

Having an open mind and just accepting that the shots I would be pouring would be vastly different to what would usually be expected of espresso would have helped massively and letting the opinions of others who probably never even tasted a shot above 16% make me feel as if I was doing it wrong was foolish. It’s hard if you’re the only one tasting these shots – getting knowledgeable people visit Cafe Ashton to have a go and confirm that things did indeed taste great was essential.

Making that perfect EKspresso

Okay here we go – let’s do this – these are the steps I take when dialling in a new coffee bean that I know absolutely nothing about.

Pick a dose and stick with it

I pretty much do everything as 19g in an 18g VST basket, this is simply just because my espresso machine tends to suck the puck out of the basket onto the shower screen if I dose at 18g. Going with less with 18g tends to mean 20 second shots for desirable yield so it’s hard to control and still be consistent – so stick with 18g and above if you want great success without messing around with lower pressure profiles. I’d do 22g if I was in a commercial environment and serve everything as split shots but at home this is way too much espresso for one human to get through and still have room left for filter later. How much pre-infusion? I use three seconds then 9 bars all the way – no messing around at my house.

Aim high on the yield

The temptation here is to stick to a classic brew ratio of 1:2 or less because we think that mouth-feel is a really important part of espresso (it is, but more on that later). Trying to do this will either result in frustration over “not being able to grind fine enough” or constantly getting under-extracted shots. Throw away this style of thinking and aim for 1:3 or just under (to begin with – some coffees work better at 1:2.5 but for now let’s not worry about that).

Aim fast on the time 

The higher the extraction we can get in less time, the more clarity the spro will typically have – presumably because we’re grinding coarser, have an even grind size, less chance of channeling, etc. Certainly I’ve noticed with a naked portafilter that shot times under 25 seconds tend to come through super evenly with minimal effort on my part.

Nutation is pointless as while it might slow the pour down, it just adds another variable to be inconsistent about – a light tamp on a tapped flat surface is all I do these days, minimal interaction = minimal variation.

If it tastes super roasty and burned at the 1:3 all the way through 20 seconds and 30 seconds it’s probably a crap coffee, throw it away and give up if you can’t get something reasonable in a dozen shots. I wasted so many hours trying to dial in coffee from local roasters because I like the people and like the idea of buying local. Get some Five Elephant in and start there if you’re in doubt – a lot of coffee is roasted for crap gear and low extraction and isn’t for this style of coffee.

I find that with great coffee it’s far easier to start at the under-extracted side of espresso and move up to the point where it starts tasting “muddy”. A refractometer will tell us where this point is too, as extraction tends to start dropping or flattening out at that time. (This is counter to what I’d do with darker stuff where I’d start over-extracted and work backwards).

It’s very easy to have doubt when we do an espresso at 19g to 58g in 23 seconds and it tastes amazing, it’s even easier to have doubt if we’re at 25s for a similar recipe and have our taste buds tell us that we probably want to go through even faster! This is now a common experience and I just accept it but the refractometer telling me that I’m getting 22% EY at these times is a really useful signal that our taste buds aren’t wrong.

Give the coffee time to become more soluble

More so than we seem to be encouraging in most popular literature. I can absolutely take a coffee from a roaster and do a gusher as spro and have it taste good within 24hrs of roasting but it won’t be consistent and more often enough won’t be soluble enough to easily extract well for the grind-range we have available on the EK43. Most good coffee is best between two and four weeks – it’s not just de-gassing, solubility just increases with age for whatever reason and higher extraction simply comes easier.

Know your reasons for pulling shorter

Strength: While a long brew ratio for a lot of light roasted and fruity coffees works really well (acidity giving way to sweetness at a slightly lower strength), some washed coffees just don’t have the character for this and will taste “watery”. Dropping down by 5-10g and increasing time by 2-3s fixes this. More often than not this is actually because the bean isn’t that soluble and the strength really is lower – a 20% EY at a brew ratio of 1:3 will nearly always taste watery. If it’s not easy to push the bean over this extraction yield at this sort of ratio it either needs a bit more time to become more soluble or is just not a good fit for this style of coffee. Pull shorter or do it as a filter.

Age: As a coffee gets older I tend to find that I drift towards more classic recipes (I write down all my recipes and extraction yields over the course of a bean’s lifetime at Cafe Ashton). That acidic and sweet 22% EY starts to turn into a bitter or muddy mess (and jumps up beyond 23%) at the same recipe from two weeks ago. Pulling back down to 20% at a brew ratio of 1:2 is a last resort and is a good way of getting the last of that spro drinkable. An easier solution is to open a fresher bag and put those beans to one side for when we next want to season a grinder. Why would we want to drink anything less than perfection anyway?

Roast: Some beans are just so good that we’ll want to use them anyway, some slightly more developed roasts just won’t work at these extractions and we have to pull them back to keep them at or under 20% EY. Most of the time this isn’t worth it but sometimes there is that one coffee that we really want to use and doing a brew ratio of 1:2 to keep that EY down is the only realistic solution.

Omniroasts are our friend, as are filter roasts – most (not all) of these so-called espresso roasts are aiming to dull the acidity to compensate for the crap job most grinders and recipes do of espresso extraction.

In defence of the “lungo”

Okay okay – so I’ve saved the answers to common questions and the defence over this style until last – the common objections or remarks from people questioning the EKspresso are (I’ll add to this if I’m asked any more)

  • I can’t grind fine enough to get a good espresso pour
  • The long shots are too big to do as milk drinks
  • The long shots surely won’t have the mouthfeel or strength I want from spro
  • “what about crema? Do I still get crema??”
  • I can’t afford to pre-dose all those beans out

Grinding fine enough

So first off, if we’re aiming for a brew ratio that’s bumping up to about 1:3 and aiming for times less than 30s (which is fairly common even for the high 22-23% EY shots) then grind size is simply not an issue. I can even do most Pacamaras as spro using this brew ratio and target time. The problem here is nearly always in aiming for that tidy looking classic pour. EKspresso does not pour like classic espresso and the results are so much the better for it.

It is true that if we’re using a lower dose then we’ll probably need to lower pressure and flow-rate to get a manageable pour time but for my purposes the 50-60g shots and the 19g doses suffice – this is therefore something that I don’t do.

Aiming for this high yield means that the low time isn’t such a big factor in variability – a gram or two either side makes very little difference to the resulting flavour.

The long shots are too big to do as milk drinks

This is a fair enough point – while at home my solution is to just drink a bit of the espresso while I heat up the milk, that’s not really workable for the shop environment. I think there are several solutions for this:

Don’t use the EK43 for milk drinks (this also means we probably shouldn’t use the EK for espresso either because it’ll just be a pain to keep dialled in). In milk the amazing clarity and sweetness isn’t exactly a priority anyway, a well dialled in Mythos shot will probably be just as good.

Do split shots – either properly by actually splitting the shot after pouring or just by using a spouted portafilter.  In milk the taste difference between the two spout outputs isn’t going to be that big a deal or noticeable for most people. Milk hides most sins and if we’re doing well dialled in EKspresso those sins are going to be pretty small when compared to the sins that most coffee shops are committing anyway.

Charge the same amount, make more money – no problemo.

Use bigger cups, as an industry our supposedly standardised names for 4oz/6oz/8oz drinks is completely pointless anyway given how different most shops espresso and milk styles are. 7oz flat white? Sod it – why the hell not?

Encourage people to drink less milk and more EKspresso – Ho ho yes indeed. Sorry not sorry – but this whole patronising notion that “most people aren’t ready for modern light fruity espresso” comes largely from the fact that most shops still serve under-extracted and sour guff that is borderline undrinkable unless we’ve taught both our taste buds and stomach to tolerate it.
Sweet fruity well extracted espresso is a surprising joy to people and I’ve now lost count of the number of times I’ve now heard the phrase “I usually have milk but wow, I didn’t know espresso could taste like that..”.

The long shots won’t have the mouthfeel or strength

I love this one because it’s completely nonsense; I’ve lost count of how many times I’ve been into a coffee shop serving modern well roasted coffee such as Five Elephant from a crappy mazzer with a recipe of 18g -> (<36)g and it has been under-extracted to the point of simply being muddy acid water. Pull the refractometer out and we’re lucky if these shots are above 16% (Indeed most shops tend to hit 16% and stop because there is a slight sweet hump there – yay taste buds!)

I have actually carried my refrac around with me with travelling and this is pretty much what I have discovered across the UK

  • 16% EY at a brew ratio of less than 1:2 means a TDS of less than 8% and this is terrifyingly common across our industry
  • 19% EY at a brew ratio of 1:2 means a TDS of 8.5-9%, and this is where most Mythos shots seem to be sitting
  • 20-21% EY at a brew ratio just over 1:2 means a TDS of around 9%, and this where a well dialled in Mythos seems to sit
  • 23% EY at a brew ratio of just under 1:3 means a TDS of around 8-9%, and this is where a well dialled in EK43 sits

Obviously there are outliers with folk using super soluble darker roasted espresso and such but I don’t care about that style of coffee so this doesn’t factor up here.

Essentially most shops aren’t refracting and most shops are under extracting and therefore serving short watery dribbles of guff-juice, or they’re extracting well on a Mythos and serving long-ish shots of coffee at or around the 9% mark.

Mouthfeel is such a myth I don’t even know where to begin, everybody I know has started off thinking this (including myself) and then when shown the facts and the end results realised that the quality of coffee coming out of these higher brew ratios with these lighter coffees and far higher extractions is much higher for a fairly similar strength of coffee. The intensity of flavour when using a sensible light roast is as such that we probably don’t want it much stronger anyway.

The fact is by pulling these long shots and extracting well, we actually open ourselves up the possibility of serving even more single origin coffee as espresso beyond those dull Colombians and Guatemalans.


We get more of it with a longer shot, stop pontificating. With decent water we end up needing to really mix it in too because so much ends up being produced.

I can’t afford to pre-dose those beans

But we can afford to lose kilos of coffee because of mediocre grinders like the K30 (or duff grinders like any Mazzer made ever)?? Okay no fair enough – there is also a human element to this – weighing out beans for a busy coffee service is like stamping cups for a busy coffee takeaway service – nobody wants to do it and it is mind numbingly boring.

Thankfully there is a solution – the Five Elephant volumetric doser not only works but it works well, especially when you’re dosing at 22g (because the margin of error is less at high doses). Yes it’s not perfect but it’s the best we’ve got because…

Nothing better is coming in the near future

The Peak was supposedly an EK43 in a K30 form factor but from all reports it’s nothing of the sort, and the fact that Malkhonig has now decided to do a slightly more friendly “EK43Barista” seems to be a tacit admittance of this.

The Mythos seems to be the bare minimum for good coffee right now and I’ve still not had a shot from one that even comes close to the clarity and sweetness of a good shot from the EK43.

Even if we’re making the commercial decision to not use the EK43  in shops (and there are plenty of reasons commercially for making that decision), discarding the EK43 espresso genius as a gimmick when most of us haven’t yet had a proper EKspresso shot is missing out on a whole world of flavour that is available to us – and buying great coffee and sticking it through a Mazzer should be a hangable offence.

TLDR: Give EKspresso another go, it’s worth it.