I have tested temperature stability on KvdW Speedster last 3 day, and found pretty long ramp-up time on it.
I wonder why a ramp curve like this was drawn, so I suspected something was up with pre-infusion.

Speedster's has quite long pre-infusion duration with it's own PPIC(Progressive Pre Infusion Cylinder).

So, I did temperature test on my GS/3 MP under two phase ;

#1. Standard extraction.
#2. 15 seconds extraction under full pressure 9bar after 10 seconds 3 bar pressure tap water Pre-infusion.

I got a average data from 10 times extraction test on each situation ;
Totally, it is summarized on same graph "Speedster's, GS/3 Standard flow and with P/I".

On pre-infusion phase, GS/3's ramp curve move toward to Speedster's curve (But not exactly same shape).
GS/3 with P/I shows more slow temperature ramp speed than standard flow, we can suspect that is caused by flow-rate and thermal energy.

If we try make shot using pre-infusion, maybe it was attended with much different thermal energy.

Note :
* Speedster's initial ramp-up duration is about 10 seconds.
* Both Machine are setted to 95 Celsius, but there are different offset. So, for convenience I re-adjust final temperature to 93.0 celsius.

Devin, thanks for doing this very interesting test. For those unaware, Devin did this work based on the Modified procedure for measurement of brew water temperature that I posted about today. I had shared this procedure with him privately last week when I saw that he was using a Scace on his Speedster on Facebook.

I haven't logged the data, but I see a really similar effect with the Slayer, which is to say that using pre-brew when doing shot simulation results in a noticeably slower rise to temperature than going straight to full flowrate ("brew") mode. It's a really interesting dimension to "flow profiling," and I was surprised to see how much it was "baked in" to the Speedster's temperature profile (that is to say, it can't be avoided without disabling the preinfusion chamber).

I definitely feel like this could be one explaining factor in how shots with very long (20s+) pre-brew times don't seem to cook or over-extract-even though the brew time is extended, the average brew time is effectively dropped.

shadowfax wrote: It's a really interesting dimension to "flow profiling," and I was surprised to see how much it was "baked in" to the Speedster's temperature profile (that is to say, it can't be avoided without disabling the preinfusion chamber).

And hopefully no one would ever want to do that. Kees discusses the pre infusion features built into the Speedster here: http://www.keesvanderwesten.com/speedst ... ction.html As I've mentioned a number of times previously many of the profiles I see Slayer owners, modified GS3 owners & others posting about are designed right into the Speedster. A slow, low pressure pre infusion is a key element of the Speedster's design so it shouldn't be surprising that the temperature ramp is slower then what you would see in a conventional pump machine. Whether you select manual P/I or Pump the puck will be soaked with a gentle low pressure shower before the pressure ramps up.

JohnB. wrote:And hopefully no one would ever want to do that.

I love being able to to grind fine and do 20+ second pressure ramps. That said, not every coffee benefits from this strategy, and many are outright terrible with it. There's something to be said for a fast ramp with certain styles of espresso.

In which cause you simple select the Speedster's pump function & get a 4 second ramp to full pressure.