I don't see any problems in your video. Everything looks normal.

Mine does all of those things.

During the pull the piston is stabilized by the fact that the gasket is tight against the basket which is locked in place. Seal on the robot is between the gasket and the basket, unlike other machines where the top of the portafilter gets pressed again s a seal in the group head.

Maybe try pulling the shot with the Robot on a lower surface for better leverage?

nrxyn wrote:
Today, in my first cup I changed some variables:

1. No more double pour method. I tried overflow method by tilting a bit. This way I thought less disturbance to puck. Looks better.
2. Decreased the preinfusion time. Made preinf until first drop.
3. After preinfusion, slowly raised to 6 bar. Finished with 4 bars.
4. And of course made WDT and hard tamp and paper filter.
5. Placed the metal screen with no pressure.
6. Overfilled the pf nearly full. Is there any harm doing this?

Again, there was one main stream and one another dripping area. But the taste was sweet.This is an improvement.

You can try the following technique, which I recently adopted to deal with my channeling problems. A couple of months ago I switched to a finer grind (lower dose) because I noticed that it provides more aromatic and fruity taste, while coarser grind (higher dose) results in a more traditional espresso taste. So I wanted to grind as fine as I can to pull the shot but channeling became a real problem. Until I added the following technique during preinfusion (first 10-15 seconds) - instead of pressing at 2-3 bars for 10-15 seconds, I start with a fast push to about 5 bars, which usually results in a first drop appearing after 2-3 seconds, and then I start what I would call a pulse motion - gently dropping pressure to 1-2 bars and rising it agin to 4-5 bars within about 1 second intervals. Like breathing in and out, 1 second in, 1 second out. I repeat this pulse movement until the drops turn into a stream and then continue with 6-7 bars till the end. With this technique, I can easily pull through even the finest grinds, something that would choke my Robot before. I usually go with 16 grams, and only get channeling occasionally maybe 1 in 20 pulls. The taste is usually sweet, aromatic, clean and fruity (makes me crave for a second cup every time).

I think there are several benefits in this technique. It helps to saturate the puck with the hottest water faster, which helps with proper extraction. It allows to grind much finer by helping water to "find" its way through the puck rather then forcing it through. And it helps to avoid channeling for the same reason. Worth a try IMO. You just need to find the right rhythm of the pulse, each grind and beans require slightly different efforts.

From considering your ("VoidedTea") description, I would think this would havemore risk of puck fracturing and subsequent channeling. But (seriously) I am glad it is working for you. I would be curious to hear opinions of expert Home-barista forum members.

Back to the channeling question posed by "nrxyn" in post #3759, until proven otherwise, your described channeling is not from the Cafelat Robot. My novice guess of cause: your puck prep. What would be helpful is full video of your bean preparation (e.g. dose, distribution, tamp) and then extraction. You posted video of representative problem. Specifics help. If "nrxyn" have already done this then please accept my apology for missing it. This thread is kinda long:)

Here is my Cafelat Robot barista user experience do date [~ 52 espressos made from my Cafelat Robot barista (2 espressos per day) since arrived on 5/13/2021 from order placed 5/8/2021.....which still amazes me as order came from Hong Kong]
* zero channeling for all but first espresso made to date (I always skip using the spouts i.e. naked portafilter): first extraction had channeling, gushers, etc from naked professional basket as I was not used to such a deep PF basket.
* made lungos (e.g. 17 grams bean and 58 grams drink 3 fluid ounces), regular, ristrettos and tight ristrettos (e.g. 21 grams bean 21 grams drink, 1.5 fluid ounce)
* peak initial brew pressure range is 6 bars to 14 bars. The crazy-high 14 bar is for regular brew light-roasted bean that really are not meant for espresso. I brew these beans as a challenge. Rest are brewed as drip. Again, zero channeling and stream perfectly centered (e.g. naked portafilter porn) despite ridiculous high peak brew pressure.
* I always use Ross Droplet Technique (RDT) and Weiss Distribution Technique (WDT) and always weigh beans with 0.1gram scale
* above includes preheat and no preheat of Cafelat Robot barista.
* above includes use and no use of filter paper and tamp to 30 pounds

It should be noted my 1st real espresso brewer is my Quickmill Alexia with PID. I use nothing by VST portafilter baskets for this (15,17,20,21gram). These baskets, in my experience, are very unforgiving w.r.t. espresso prep resulting in naked-PF-porn extractions (e.g. ones with no channeling, perfectly centered stream, tiger striping). Also please realize espresso-porn looking naked PF extractions do not translate into "g-d shots."

mdmvrockford wrote:From considering your ("VoidedTea") description, I would think this would havemore risk of puck fracturing and subsequent channeling. But (seriously) I am glad it is working for you. I would be curious to hear opinions of expert Home-barista forum members.
"

Yes, I believe it does work. I wouldn't recommend it if it was leading to more channeling. But based on at lease 4-6 weeks of daily pulls with this technique, I am very happy with the results. The main advantage I see in this technique is that it allows for a faster pull (under 30 sec) with finer grinds. Which is very important due to unavoidable temperature drop in the Robot during the pull. The only viable alternative to the pulse technique for finer grind with the Robot is much longer preinfusion, which has a very negative drawback of temperature drop for the actual pull. Maybe that is why people try to compensate by preheating everything (I don't do that). And in addition, I found that even with longer preinfusion, it is quite difficult to press the water through the fine grind without ramping up the pressure to 9+ bars. I practically have to apply all my weight and strength in order to speed up the pressing time. Very uncomfortable. The pulse technique significantly reduces both preinfusion and extraction time and allows to use less force for the same grind. Meaning extraction happens at much higher temperatures and with much more comfort. You don't need to take my words, just try it for yourself, especially with a very fine grind, something that usually makes you struggle a lot. I am sure you will notice the difference.

I have on occasion "loosened up" a stuck shot by briefly letting up on the pressure, maybe going from 6-8 bar down to 1-3 bar, then back again. This has allowed for a more normal flow rate in what would have otherwise been a very slow shot. It was suggested to me that I was probably creating channels by doing this. Was I? Or was I doing a kind of "pre-infusion" a little later than usual, allowing the puck to expand a bit? I have no way of knowing for sure. I didn't measure the extraction. The shots didn't taste terrible.

FWTW.

I've done this with all my manual levers including the Robot, easing up on the lever to save a shot. I can't say if this was creating channels or not but I can say that without doing this it would be a sink shot.

I don't see why "easing up" on the pressure would be a problem, so long as you aren't lifting up the arms.