Hi guys,

as mentioned before I'm having problems with the extractions on my Elektra, pretty much since day 1. I'm using small doses, 14-16g, usually 3s preinfusion with 1bar and 8.75 applied throughout the shot. Normally it starts well, but goes sideways near the end, like here:



goes blond pretty quickly and the flow isn't as calm as it should be. It's channeling pretty badly, especially the last 5s of the clip, but that's how it is with most coffee (although with most I only get the occasional bright spot here and there). The blend I'm using now is from a London roastery called Monmouth and is less than 2 weeks old. You can also see that the cone that forms is very shallow, far from the stuff you usually see on a naked portafilter.

Now tastewise it's not too bad, pretty balanced, quite sweet even, but I'm curious why pretty much every coffee goes blond so fast? Recently I made a test dividing the shot into 3 cups and the last one was very watery, almost transparent.

Anyone care to comment?

Regards,
dsc.

Hi guys,

I've been having extraction problems for a while now (a few months I would say), very annoying thing, especially when you can't figure out why that's happening. Yesterday I tried something very simple and easy and it seems to have solved the problem, at least with the coffee I'm using now.

My usual routine when preparing a shot is something like this:

- dose straight into the basket moving it around while the grinder is working (it's a doserless Mazzer-E type of config)

- tap/thump the PF on the fork to settle the grounds

- tamp medium heavy (around 15kg?) or light, it depends really I usually try both and do a twist to remove all coffee from the piston

- lock and go

I tried various tamp methods (twist, no twist, heavy/light tamps and so on) but nothing was working properly. I was either getting thin, fast blonding shots or tight ristrettos which were almost choking the machine. Especially the beginning of the shots were very strange, super fast during the first second or so, then slow for a few seconds and very unsteady and unstable for the duration of the shot. It looked as if the fine particle were being pushed down during the first few moments, but before settling there wasn't enough resistance and water was shooting straight through the puck resulting in channeling and sprites. Remember this is only during the first 1-2s after which the flow would slow down, sometimes dripping, forming a very thin stream. The stream would normally blond after around 15-18s with additional sprites and sometimes heavy 'dancing'. Taste-wise each shot was definitely not right, usually harsh and too acidic, lacking body and simply thin.

Because I already tried various tamping methods I decided to omit the second step which is the tap/thump to settle the grounds. I ground the coffee as usual, the same dose, straight into the basket with a small heap in the middle. Tamped medium heavy and locked it in the group. The result was something I wasn't expecting, totally different pour, slow (even though I haven't changed the grinder setting), dripping for a good couple of seconds, forming a nice thicker cone and stream, not blonding for at least 25s, steady, stable. Tried the same again and got the same result. Taste-wise definitely better, more round, less harsh flavours, more body.

I know that I've only tried this with two shots and single type of coffee (SO) but I'm curious why such a difference? I've ready posts where people said that tapping/thumping the PF to settle grounds is good, heck I even saw it and had the best espresso in my life prepared this way (although it was from a doser fitted Anfim). Is it possible that my modified Major produces a proper 'mixture' of coarser and fine grinds which somehow gets messed up when disturbed? Still tapping/thumping would normally move fines further down, not up and so would slow down the extraction, so I'm a bit confused here.

Any suggestions/ideas appreciated.

Regards,
dsc.

We should get the last chunk of this split out to techniques or some other forum (at least to its own "help me fix this" thread) where people will actually read and help diagnose your extraction.

I would have completely overlooked this older thread except that it was about the T1 in particular =)

How many grams? Perhaps grind finer and under dose?

Mark

Hi guys,

true about the last bit, but I didn't want to start another topic and decided to keep it all in here.

Mark I've tried both finer grind/low dose and coarse grind/higher dose, it was still far from perfect.To be honest I thought that the non-thumping would change something but the difference is marginal. The pour is simply very chaotic, the cone moves around a lot and almost never stays in the middle (and even when it does there's a lot of bad things going around anyway). I wouldn't worry if it was only the look of the extraction that's wrong, but it's actually bad taste-wise as well.

Regards,
dsc.

This has always been a perplexing problem, Tom. IMO a Major and T1 should perform better than you're experiencing, and I think if I were in your place I'd be suspecting a flaw in one of the two. I associate the way your pours look and your description of it with really dodgy or badly mounted burrs. I know you've said you replaced them, but are you sure that they're spinning true? have you checked the shaft? Is there a wobble?

If I were happy with that I'd put my doser back on and try dosing with that--just to make sure that the doserless mod isn't the culprit. A long shot, but it sounds like the kind of thing you're down to.

After that I'd be looking at my dispersion screen, how it's distributing water during a flush. If all that's good, I'd find a friend who could loan me a really nice grinder (hopefully a big conical, but in any case one that makes great pours on other machines) for a couple of days, and see if that improves things.

Beyond those diagnostic questions, what taste defect do you associate with these ugly pours?

Hi Nicholas,

agreed if someone told me a year ago that I would have a Major and an Elektra in my kitchen I would've never thought shite in the cup.

Things I'd like to try first before taking the rig down to the parking lot and running it over with my car:

- low dose vs. high dose - which is effectively checking how headspace affects the pours

- preinfusion vs. no preinfusion

- current dispersion screen vs. original dispersion screen (my current one has a bit of wire stuck between the mesh and the metal backplate)

- Macap vs. Mazzer - I still have my Macap and even though the burrs might not be superb I'm just curious what the result is going to be

Normally the answer to these kind of questions is 'go back to basics', so I will try to turn off the preinfusion and pack less coffee in the basket.

I've checked the burrs and they appear to be parallel, at least to some degree. I know that when they touch slightly it's not a constant noise like they're touching with the whole surface, but I've never heard a grinder that did this (I think they all touch partially). Perhaps it is a little bit off, but I don't think there's anything I can do about it. Can I somehow check their position, using home methods? I tried looking at it but the human eye is not good enough to check the position of fast spinning elements. I don't think there's wobble, but I will check again today.

I was thinking of removing the funnel and might do that if nothing else works. The dispersion screen is another suspect, I think there might be something wrong with how the water is flowing straight from the group. That or the preinfusion which is doing more bad than good, although you've tried it yourself and it wasn't too bad, was it? Of course a borrowed conical would be a handy tool to see what's wrong.

Taste-wise the shots just taste bitter and sometimes smell very strange. You know when you lean down to smell the shot that ended a few seconds ago it smells a bit like rubber/tire, synthetic. I was wondering whether it's because the temp is too high but I'm mostly brewing at 91*C-92*C and if I go below the shots are sour, acidic. I can't really get any real flavour from them, although a few days ago I had a shot which was awesome (at 92*C), lemon/sweet and not bitter at all.

The thing that worries me the most is how unstable the shots are. When I look at clips of various pours most of them seem very calm, dripping, forming a cone, slowly getting bigger. If I try to do something similar I choke the machine so that it drips all the time, or I get a 15s blond thin crap.

I will order 1kg of beans from HB (same blend) and do some debugging over the weekend.

Regards,
dsc.

Odd. The Elektra machines are one of the easier machines to use IMHO. Meltdown shots are usually caused by the beans, grinder or barista. Assuming you are using good beans that leaves the grinder and and the Barista. Are you weighing your shots? Elektra's are very picky about head space, across their entire product line. Because the shower screen is attached directly to the dispersion block, there is no head space above the screen unlike your average E61 box which has a good ¼ inch of air between the dispersion disk and screen. A half gram to much/little will make a difference in the shot. A grinder that is off just a bit will also cause a torrent of problems. I am also assuming that the ugly shots are bitter and lacking body, a symptom of channeled shots.

Hi Dave,

strange indeed. Yes the shots are thin, lack body and are bitter (although they do have a fair amount of crema).

I do weigh the shots when I'm dialing in my grinder, but once that's done I trust the timer. I will check how much coffee it's packing now and how consistent it is.

Regards,
dsc.

I cannot explain it, as I've never had that problem either. But I think Dave's certainly offered good advice . . .

The entire Elektra line is very touchy with dose. A half gram can make a big difference. I run 14-15 grams in the stock double basket, LM ridgeless I can go 16. Any more than that I go to a triple basket. When I get gushers it is almost always a dose problem or I am dialling in a new coffee. You could be having a completely different problem but that has been my experience with the A3 over the past couple years.

Hi guys,

it looks like I have figured out what the problem is (the major problem is actually something that was noticed by the good people on TMC):

- my dose was too high, I was packing 17g in a double basket which is too much, now I've lowered it to 14.5g and it made things better

- the main thing though was the water debit, when measured I was getting 180ml per 10s when cold and 170ml per 10s when hot. That's almost 3 times the suggested amount (75ml).

It turns out that the new pump I've installed is a Rotoflow 1504 model and the older one was a 104 model. They have different characteristics, but still nothing huge, perhaps around 10l/h difference or so. I've fitted a smaller gicleur I got from Nicholas and it lowered the water volume to 110ml per 10s, still too high, but better. The flow from the group is more calm, I don't get any spitting anymore (when waking up) and it appears to be working as it's suppose to. With the PF locked in the beading is slower (well duh) and the extraction is more stable. I'm getting 1kg of beans today which I'm planning to try out during the weekend, but things are definitely looking (and tasting) better. I've also decided to install a needle valve on the supply line that goes from the pump to the machine. This will allow me to tweak the flow to whatever values I want/need.

Will update when the beans get here.

Regards,
dsc.

Keihin motorcycle carb main jets fit the group jet riser on the A3. I have thread somewhere discussing it but dont have time to search for it this moment. I noticed it one day after re-jetting my bike. I went to the bike shop and got the smallest main jet they had and replaced the group jet on the machine with it to lower the flow rate a touch. As a general rule the flow from a rotery will be higher than a vibe but I have pulled very good shots on machines with no jet installed at all. But your grind and distribution have to be spot on.

Hi Dave,

yeah that's the jet I've got installed at the moment and I'm pulling better shots. Still it will be interesting to install the needle valve and play with the flow a bit, try a few different values and see how it affects extractions.

I'm curious if the machines with no jets have pumps with much lower flow/h values.

Regards,
dsc.

Which machines have no jets? all E61 machines have a gicleur (same thing) just forward of their preinfusion chamber; LMs and Synessos all have jets. These may be redundant on certain pumps (e.g. vibratory pumps), but in the case of the E61 home machines they are at least the result of pulling a group that was designed for multi-group espresso machines, where the pump must have a flow rate sufficient for all groups simultaneously, but must not overload a single group when it it operated alone.

A needle valve is probably a superior gicleur--it acts as controller for variable flow rate. Where are you installing it? you might want to make sure you install it after the 4-way junction so that flow isn't restricted to the boiler fill line. If you don't, you'll want to be careful not to overdraw from the boiler, as it will fill more slowly.

For reference, the discussion of the Keihin carburetor jets can be found in this thread. Tom has a #68 jet, and I have a #65 and #68. I took mine out after awhile, as I found that it made the cooling flush unbearably long without any real improvement, on my machine. I ended up installing a delay timer on my pump for preinfusion instead (which IIRC Tom is using as well).

Hi Nicholas,

I think some LMs are jet-less, but I wouldn't bet my money on it.

Yes a needle valve is like a gicleur, the only difference being the fact that it's adjustable. I will install it on the supply line to the machine, the which comes from the pump. I'm not really worried about filling the boiler slower then before as I don't use too much water from it anyway.

Pulled some shots yesterday with a super fresh espresso blend and finally it starts to look and taste as it should. I get almost no bitterness and other flavours which weren't there before. The cone actually gets bigger as the shot progresses and it blonds slower.

Looks like it's all on the right track.

Regards,
dsc.

dsc wrote:Yes a needle valve is like a gicleur, the only difference being the fact that it's adjustable. I will install it on the supply line to the machine, the which comes from the pump. I'm not really worried about filling the boiler slower then before as I don't use too much water from it anyway.

If the boiler decides to autofill while brewing, it will cause more disruption than normal in that setup, I think...

Hi gyro,

yeah true, but that happens very rarely on my machine and if it does the shot is usually bad anyway (or not good enough). I was thinking of disabling the autofill during the shot though...

Regards,
dsc.

shadowfax wrote:Which machines have no jets? all E61 machines have a gicleur (same thing) just forward of their preinfusion chamber; LMs and Synessos all have jets. These may be redundant on certain pumps (e.g. vibratory pumps), but in the case of the E61 home machines they are at least the result of pulling a group that was designed for multi-group espresso machines, where the pump must have a flow rate sufficient for all groups simultaneously, but must not overload a single group when it it operated alone.

A needle valve is probably a superior gicleur--it acts as controller for variable flow rate. Where are you installing it? you might want to make sure you install it after the 4-way junction so that flow isn't restricted to the boiler fill line. If you don't, you'll want to be careful not to overdraw from the boiler, as it will fill more slowly.

For reference, the discussion of the Keihin carburetor jets can be found in this thread. Tom has a #68 jet, and I have a #65 and #68. I took mine out after awhile, as I found that it made the cooling flush unbearably long without any real improvement, on my machine. I ended up installing a delay timer on my pump for preinfusion instead (which IIRC Tom is using as well).

You can remove them from any machine. Just unscrew it. The one I was using was a 4 group LM. Did not realize it had them removed until after a few hours of shot pulling. Then Berry asked me what I thought of the machine and told me it was running wide open and that is how they run it in the shop. As long as the dose/distribution/grind are correct, it will work with no issue but if you are off just a little, it will let you know. The jets just add a little padding for error and make the machine a little kinder on the Barista IMHO. I changed them on my A3 just to see what it would do. The flow rate dropped just a little but I don't think it made much impact on the overall performance of the machine. If I had another one, I would not bother changing it.

Dave,

I put the #68 back on my T1 yesterday. Moreso than last time, it's made a definite improvement in the forgiveness factor on pours. Still, I notice a definite and slightly irritating increase in flush time with the restricted flow rate. I agree there's not much reason to modify the gicleur on a stock machine--Elektra tuned the machines to work really well (at least at the intended doses). In my case it feels like a rather useless tradeoff. However, since Tom's added a higher flow rate pump, it probably does make sense for him to tune it down with a smaller than stock jet.

For what it's worth, Tom--If you want more jets to try, you can probably use newer Synesso Ruby jets, I think. They are little disks that insert under the gasket between the grouphead and the solenoid valve inlet, so they are compatible I think (looks like the same solenoid valve to me). You might be able to beg some spare jets off a high end shop with Synessos. It's a thought, though it does make the thing harder to get at (have to take of the solenoid valve instead of the group cap). I believe synesso has 0.6mm through 0.8mm, maybe more.