On swarf: does the Fuller brush man sell any 30awg brushes?

(sorry I couldn't resist)

Here's one that surprised the hell out of me, and bound to raise some eyebrows. As I started testing Greg Scace Portaroid (name still in progress), I have learned something interesting: It is reporting a much wider temperature variation within a shot than I expected. A typical variation on the machine I tested was 1.6-2f from the 10th second on. A ristretto will reduce it, but I was checking 60 ml flow in 25 seconds - a double. (it is fixed on the Thermofilter).

However, if I ran the shot on a cold P/F, the variation shrunk to 0.8f. About half. I consistently repeated the experiment, using the same flush routine, and repeated the result. The brew temp I reached was ~1.5 f lower than one I would reach had the P/F been in the machine at all times. However, it was repeatable.

The Thermofilter uses a bottomless P/F that has less mass than a regular one. My conclusion is: at least for a bottomless P/F, on a dual boiler machine, keep it at room temperature at all times.

Abe Carmeli wrote:at least for a bottomless P/F, on a dual boiler machine, keep it at room temperature at all times.

You're assuming that the smaller the temperature variation as measured by the Thermofilter thingie, the better the shot.

Where's the evidence for your assumption? Just cause Schomer says it, doesn't make it so....

So......

How does the coffee taste?

terryz wrote:So......

How does the coffee taste?

Coffee? What coffee? Don't confuse me with the facts Terry, I'm too busy looking at thermometers here.

Abe Carmeli wrote:It is reporting a much wider temperature variation within a shot than I expected. A typical variation on the machine I tested was 1.6-2f from the 10th second on. A ristretto will reduce it, but I was checking 60 ml flow in 25 seconds - a double. (it is fixed on the Thermofilter).

However, if I ran the shot on a cold P/F, the variation shrunk to 0.8f. About half. I consistently repeated the experiment, using the same flush routine, and repeated the result. The brew temp I reached was ~1.5 f lower than one I would reach had the P/F been in the machine at all times. However, it was repeatable.

The Thermofilter uses a bottomless P/F that has less mass than a regular one. My conclusion is: at least for a bottomless P/F, on a dual boiler machine, keep it at room temperature at all times.

your conclusion does not necessarily follow from your evidence. it sounds like you're rejecting the results of the first method simply because you prefer the results of the second method. using a cold test instrument will pull some of the heat from the brew water sample, which would give you both a lower overall temp indication as well as attenuate some of the temp fluctuations.

Barry wrote:your conclusion does not necessarily follow from your evidence. it sounds like you're rejecting the results of the first method simply because you prefer the results of the second method. using a cold test instrument will pull some of the heat from the brew water sample, which would give you both a lower overall temp indication as well as attenuate some of the temp fluctuations.

I believe you lost me there. Do you mean to say that I prefer an 0.8f temp variation to 1.6f, and that preference dictated my conclusion? If so, you are perfectly right. And my followup question is who doesn't?

Abe Carmeli wrote:I believe you lost me there. Do you mean to say that I prefer an 0.8f temp variation to 1.6f, and that preference dictated my conclusion? If so, you are perfectly right. And my followup question is who doesn't?

my point is it is very likely the .8F temp variation isn't. it's very likely you're getting the same 1.6F variation, only attenuated by thermal losses to the <cold> instrument.


--barry "rose-colored glasses"

Abe Carmeli wrote:My conclusion is: at least for a bottomless P/F, on a dual boiler machine, keep it at room temperature at all times.

Abe, are you saying the portafilter AND the thermofilter were cold?

I ask because Terry sent me a thermofilter for evaluation. I started with the bottomless portafilter in the group (hot) with a normal basket. Later I swapped in the thermofilter and noted the readings indicate it has greater initial thermal mass than the equivalent coffee + basket. That is, the readings from the thermofilter were not consistent with the profile I would expect from an "over the basket edge" measurement until the thermofilter was hot. Not surprising given the brass fittings, but implies cold instrument measurements are suspect at best.

terryz wrote:So......

How does the coffee taste?

I am pleased to say this morning's was the best in recent memory. Really, really good.

barry wrote:my point is it is very likely the .8F temp variation isn't. it's very likely you're getting the same 1.6F variation, only attenuated by thermal losses to the <cold> instrument.

Indeed so. But how would the results differ if I had cold P/F and coffee there, instead of the thermofilter puck? Greg says his puck and coffee share a similar heat index. So, wouldn't I get the same results?

HB wrote:Abe, are you saying the portafilter AND the thermofilter were cold?

Yes. I use the Thermofiter (entire unit) cold - at room temperature. If the thermofilter puck has the same heat index as real coffee, shouldn't it behave like it when it is cold as well?

Greg or Barry can answer your question authoritatively, but my first blush measurements suggested that a cold thermofilter profile doesn't accurately model a genuine cold basket + puck. However, it correlated nicely with what I expected once the thermofilter and portafilter were at brew temperature, including an allowance for the normal interval for preparing a real puck. These results make sense to me since the heat index of coffee is pretty much zilch (and presumably the thermofilter's puck substitute), but the brass filter and flow rate valve fittings would draw off a lot of heat.

Abe Carmeli wrote:Indeed so. But how would the results differ if I had cold P/F and coffee there, instead of the thermofilter puck? Greg says his puck and coffee share a similar heat index. So, wouldn't I get the same results?

i don't think it's the plastic which is causing the results you're seeing; rather it's the heat loss to the probe itself, and to the exposed upper circumference of the portafilter basket.

also, i think you're trying to infer more to this device than was intended. it was meant to provide a uniform means for measuring the water temperature at the upper water/coffee boundary. it was not intended to simulate the thermodynamics of a coffee puck during brewing. the coffee does attenuate fluctuations in water temperature; this device was designed to minimize that attenuation to better reflect machine brew water performance, not the whole machine/pf/puck dynamic.

fwiw, the standardized testing protocol for this device calls for the espresso machine to be at operating temperature for at least an hour before the test, with the thermofilter in the brewhead during this warmup period.

--barry "hoping greg writes some instructions for this thing"

barry wrote:it was not intended to simulate the thermodynamics of a coffee puck during brewing.

...and I venture to say it couldn't do that. Not even if it were intended to do it...

What Abe noticed here is IMHO quite a normal - buffering - effect of his brewing environment. And if e.g. setting the temperature 1.5F higher and having the PF at room temperature does the trick of improving his espresso (though I am not sure it will), so be it...

After all coffee brewing is not - yet - considered an exact science, it's only that chemistry and physics apply ... as usual...

It's all about the espresso in the cup! Is it?

barry wrote: i think you're trying to infer more to this device than was intended. it was meant to provide a uniform means for measuring the water temperature at the upper water/coffee boundary. it was not intended to simulate the thermodynamics of a coffee puck during brewing. the coffee does attenuate fluctuations in water temperature; this device was designed to minimize that attenuation to better reflect machine brew water performance, not the whole machine/pf/puck dynamic.

--barry "hoping Greg writes some instructions for this thing"

Ok. I see the error of my ways. I need to re-evaluate the practical usage of this device.

barry wrote:

Abe Carmeli wrote:I believe you lost me there. Do you mean to say that I prefer an 0.8f temp variation to 1.6f, and that preference dictated my conclusion? If so, you are perfectly right. And my followup question is who doesn't?

my point is it is very likely the .8F temp variation isn't. it's very likely you're getting the same 1.6F variation, only attenuated by thermal losses to the <cold> instrument.


--barry "rose-colored glasses"

The cold portafilter isn't how you would normally produce the second and third shot if your preference is to keep the pf out of the machine, so I wouldn't do that. My understanding is that you installed the thermofilter into a cold pf, which is a different heat transfer situation than installing the thermofilter in a hot pf. Should attenuate temperature differences during the shot, so the result is not surprising.

-Greg

barry wrote:

Abe Carmeli wrote:Indeed so. But how would the results differ if I had cold P/F and coffee there, instead of the thermofilter puck? Greg says his puck and coffee share a similar heat index. So, wouldn't I get the same results?

i don't think it's the plastic which is causing the results you're seeing; rather it's the heat loss to the probe itself, and to the exposed upper circumference of the portafilter basket.

also, i think you're trying to infer more to this device than was intended. it was meant to provide a uniform means for measuring the water temperature at the upper water/coffee boundary. it was not intended to simulate the thermodynamics of a coffee puck during brewing. the coffee does attenuate fluctuations in water temperature; this device was designed to minimize that attenuation to better reflect machine brew water performance, not the whole machine/pf/puck dynamic.

fwiw, the standardized testing protocol for this device calls for the espresso machine to be at operating temperature for at least an hour before the test, with the thermofilter in the brewhead during this warmup period.

--barry "hoping greg writes some instructions for this thing"

Yeah, I'm working on them. The device is intended to be installed into a hot portafilter and left in the machine to warm up. The reason is as has been described, which is to say that there is significant heat transfer to the brass filter and flowmeter. It's not a perfect heat transfer substitute for a coffee cake, but once heated it is a reasonable and very repeatable substitute. Remember that the object is to learn about the temperature dynamics of your machine in a way that is very reproducible, then make adjustments consistent with what you know is good practice.

I'm working on the instruction set. I'll get a good set out there real soon, like maybe this weekend, as long as the punkin doesn't throw up on her daddy too much. I'm thinking about including the WBC measurement standard as well because it specifies a systematic measurement protocol that works pretty well. Downside of the WBC standard is that the technical justification for the test protocol is included and perhaps pithy reading, although it clarifies a lot of decisions for why the thermofilter is done the way it is. At any rate, the WBC standard is a good piece of work that can be very helpful because it allows all of us to discuss machine performance in response to a standardized set of tests that test a machine from intermittent usage to full-on line out the door panic.

-Greg

Walter wrote:

barry wrote:it was not intended to simulate the thermodynamics of a coffee puck during brewing.

...and I venture to say it couldn't do that. Not even if it were intended to do it...

This is correct. The thermofilter is a decent first order approximation that isn't too egregiously out of step, once the brass bits are heated up.

-Greg