sashaman wrote:I was thinking about this topic too recently after making some improvements to my HX machine. It seems that the conventional wisdom is:

DB > HX > SBDU

However, I think that for most home users the following is now more accurate:

DB > PID brew boiler + thermoblock > HX > SBDU

Of course, as far as I know the CC1 and the Silvano are the only machines in the PID+thermoblock category, and the Silvano is the only one with build quality (at least from materials, appearances and brand) comparable to the prosumer HXes. It just seems like the dead-simple temperature control of the PID brew boiler is a big advantage over flushing on an HX, and if you can get good microfoam on the thermoblock then the main "brew+steam simultaneously" advantage of the HX is largely neutralized. Thus, if the tradeoff is "slightly better steaming power on an HX vs. better temp control on a Silvano", I would think that the Silvano would provide better results in the cup for most serious home users.

There are a lot of misconceptions in the threshold assumptions, reasoning and conclusions here. For instance, a PID can only go so far in terms of stabilizing brew temps; other things like boiler size, brewpath materials, brewpath mass, group design, etc., play extremely important roles. And while thermo-block designs can improve steam performance compared to an SBDU, they don't compete with better HX machines.

By way of a single example, you make a threshold assumptions that a PID provides "better temp control." A well designed and built PID controlled machine is more convenient for a user who doesn't care to adjust temps on the fly, but otherwise it's not meaningfully more accurate or consistent for a user with enough knowledge (it doesn't take much) to use an appropriate cooling flush routine. You also "misunderestimate" the relative importance of powerful vs adequate steam.

Based on the responses of early adopters, the CC1 and Silvano seem to be excellent performers at their price points. I think you can safely add that these machines make the Rocky and other old-fashioned SBDU designs obsolete, and that they introduce a fair level of consistency in the cup along with adequate steam performance -- but otherwise you can't carry it much farther than that.

Neither the CC1 nor Silvano put the same quality in the cup or steam at anywhere near the same level as a good, ~$2K "prosumer" HX, let alone a still more expensive Elektra T1, La Cimbali M21, or Nuova Simonelli Appia at anything.

BDL

Hmm, I don't think I agree with all of your conclusions. Granted, my experience is very limited, but I've found the following:

1. True, my Expobar Office Lever is considered one of the lower end HX prosumer machines, but the parts (E61 brew group, large copper boiler, etc.) are largely the same. Even after I've fixed the issue with the grouphead idle temperature, I do NOT find it easy to get the same temperature without the use of Eric's thermometer adapter. That is, if the machine has been idling for a while and I can depend on the grouphead being at the idle temp, it's easy enough to just time the flush. However, if I'm making a bunch of shots in succession, the flush time (if any) will change significantly based on the starting temp of the grouphead. Eric has posted some charts that show how the temperature at the grouphead can vary widely based on how long the machine has been idle and any recent flushes. While with enough experience one could learn to work around this, it seems like tons of folks (justifiably) poo-poo the temperature surfing routine used to master the Silvia - I don't see this as much different. All that said, I find it quite simple to get the right temps with the grouphead thermometer, but AFAIK that's an add-on to all HX machines.
2. WRT to the steaming, I don't think I underestimate the difference - the HXes are clearly more powerful. However, at least from hearing feedback people who have used the Silvano (EricBNC, and the video on Chris's Coffee), it is quite possible to make good microfoam on the thermoblock - it just takes ~60 secs vs. 30. Is there some other difference I'm missing?

It still feels like my "stick vs. automatic" analogy from my other post is apt.