Hi,
I have extensively searched and read various forums to find info on oscar. Although oscar is quite common machine, very little testing or procedure information is available. I have started testing the temperature stability of oscar with different settings to learn more about its properties.

I have modded my oscar with musica 3.0 and 3.2 mm gicleurs and sirai pressurestat. The brew pressure has been adjusted to 10bar with a cold water side OPV. The temperature measurement device is a diy-scace built to a blind filter. Thermometer is a fluke 51 with a thermocouple and the flow rate has been adjusted to 50ml in 30s. No data logging so I have read the temperature to nearest full degree at 5s intervals.

Here are the tests and conclusions. The machine was on 45 minutes before I began the test.

Test 1:
Idle 5min, boiler pressure 1.5bar (max), brew pressure 9.5bar, volume in 30 seconds: 50ml
Test setting: Steam+3s extra flush (Waterdance+3s, I will shorten this S+3s), 10s recovery, brew
First flush from idle (walk up): Steam+3s, wait 60s, S+3s, 10s recovery, brew, 2min intervals between shots

So first two test runs would go as follows:
Walk up: Idle for 5mins, remove PF, flush the steam+3 seconds, wait for 60s, flush steam+3 seconds, wait for 10 seconds (PF attached during this time), brew for 30s and record temps, remove PF wait for 2 minutes, begin run #2, flush the steam+3 seconds, wait for 10s and brew for 30s and record temps.

Data for the first test:
Time Test number #
(s) 1 2 3 4 5
05 91 91 91 91 91
10 91 90 90 91 91
15 91 90 90 91 90
20 91 90 90 91 90
25 91 90 90 91 90
30 90 90 90 91 91

This was my old routine and I was amazed at the stability of Oscar. Clearly I was running few degrees on the cold side.

Test 2:
Test setting: Steam+0s, 10s recovery,
First flush from idle: Steam+3s, wait 60s, S+0s, 10s recovery, 2min intervals

(s) 1 2 3 4
05 93 93 93 93
10 92 92 92 92
15 92 92 92 92
20 92 92 92 92
25 92 92 92 92
30 92 92 91 92

This time I shortened the purge which worked nicely. Pretty good stability still and no overheating that some oscar owners have reported (I have the gicleurs installed).

Test 3:
Test setting: S+3s, 40s recovery,
First flush from idle: S+0s, wait 60s, S+3s, 40s recovery, 2min intervals

(s) 1 2 3 4
05 91 91 91 91
10 91 91 91 90
15 91 91 91 90
20 91 91 91 90
25 92 91 91 90
30 92 91 91 91

I increased the rebound time to 40s but kept the total interval the same. Surprisingly, this produced colder extraction temperatures. I think the group in the oscar cools too much in the rebound time.

***********
Brew boiler pressure adjusted to 1.55bar

Test 1:
Test setting: S+3s, 10s recovery,
First flush from idle: S+3s, wait 60s, S+3s, 10s recovery, 2min intervals

(s) 1 2 3 4
05 92 92 90 91
10 92 91 90 90
15 91 90 89 90
20 91 90 89 90
25 91 90 89 90
30 91 90 89 89

No meaningful change from test 1 with 1.5bar pressure.

Test 2:
Test setting: S+0s, 10s recovery, First flush from idle: S+3s, wait 60s, S+0s, 10s recovery, 2min intervals

(s) 1 2 3 4
05 92 92 93 93
10 92 91 92 93
15 92 90 92 92
20 92 90 92 91
25 91 90 92 91
30 91 90 91 91

I guess the pressure change does not directly increase brew temperature as I thought it would.

In all the previous tests I had simulated shot prepping with the portafilter removed for the 2min interval time. In the last test I repeated the previous test but kept the PF attached for the first minute and the removed it for the second minute.

Test 3:
Test setting: Steam+0s, 10s recovery,
First flush from idle: S+3s, wait 60s, S+0s, 10s recovery, 2min intervals, filter removed for 60s

(s) 1 2 3 4
05 94 94 94 94
10 93 93 94 94
15 93 93 94 93
20 92 93 93 93
25 92 93 93 93
30 92 92 93 93

This clearly had the intended effect and the temp increased by about 1-2 degrees Celsius. Still nice stability with the gicleurs. Much better than I expected. The tests were conducted with a naked PF, but with a heavy commercial PF the heat loss during shot prep would be less. I did not realize earlier how important the PF away time from group is to the brew temperature.

The musica gicleurs seemed to be effective since I had no temperature rise problems in any of the tests. Also, installing them was very easy since my oscar had the threads already in place.

Clearly I will adjust my techniques and see how it reflects to my espresso.
I hope this information is helpful to other Oscar users.

Cheers,
Mikko

Here are some pics of my diy-scace:







I did the thermocouple attachment with an old bicycle tube valve and epoxied the wire permanently.

Thank you

On you procedure from idle and inter shot what does S represent ?

Thanks again for sharing your work

^I thought that was clear from the first test where it says (S)team+3s, which means the water dance plus 3s extra flush. On the other tests S stands for steam. I will edit the first post to make it clear.

Great, thanks so much! I'll keep this as I've done this mod to the 2 Oscars I restored and I always have questions about the temperature profile and I don't have data, although I did tests using a Scace and kept in my mind the results (very similar to yours).

Only one question: how long does it take to get rid of steam? I remember I got barely 1-2 seconds. But I did the tests using 1.3 bar (and the boiler controlled by a PID, around 123°C). I remember I could get 94-95°C shots with a tiny flush (max 1 s). Indeed this group and HX setup (no injector, around 120ml HX, with the restrictors) is a killer one.

Márcio.

The initial flush after idle was about 2-3 seconds steaming plus the extra seconds for cooling. The consecutive test runs after the initial flush gave very little steam during flushes, maybe 0-1 seconds plus the extra cooling seconds of each test.

Clearly, the Oscar hx does not retain as much thermal energy as many other hx-machines. I used to flush way too much with oscar and ended up brewing little cold. I was using flushing routines that were tested with better machines like the Rocket Giotto, etc. No wonder my SO light roasts often felt little too sour to my palate. Mottling is also a good indicator, since now with corrected routine I am getting mottling on my espresso.

Carneiro, I had been reading your posting on Oscar with great interest, but as the machines were modded little differently, I wanted to perform my own tests. I haven't tested my oscar without the gicleurs, but that would be interesting to everybody else who has a stock machine. Maybe I will get my self around to that some day. The assumption is that the temp profiles might be rising and, then, a lower start temp of 92, e.g., is ok. But we will see.

Now I remember, when the machine was idle in fact the steam flush was a little bigger. Nice to know your results are very similar.

I've run some tests without the restrictors and the temperature profile was always a rising curve. Even shots starting around 91°C ended around 97°C. Maybe the OPV regulated for 10 bar after the HX (the port at the left of the group) could help on this.

BTW, I think on the second machine I modded I kept the OPV after the HX, with no problem. This keep the flow inside the HX similar from ristretto to lungo shots, as the OPV at the cold side not. Maybe you could measure different temp profiles between a very slow flow (1 ml/s) and a fast flow (3-4 ml/s).

I will do a test where I modify the flow from normal as you said. I will post back in couple of days when I have time to complete the test. Any other good ideas for testing now that I am able. I ran the pressure at 1.3 bar first but I was getting very sour shots.

Nice to know that without the gicleurs the profile was always rising. No need for me to test that. With this information what I don't understand is that why they have not installed the restrictors in Oscar in the first place. It hardly can be a cost issue since they virtually cost nothing, even as spare parts. Maybe it's because they had to make Oscar somehow inferior to Musica? Who knows.

One thing that is very clear though. Oscar is very different to most prosumer hx-machines from what I have gathered. My friend has a Rocket Giotto and he runs it at 0.9bar boiler pressure. He was flushing something like 200-300ml of water initially before starting to brew. Very different animals.

I am very happy that we are learning more about the oscar. If I would have known the temperature stability and behavior of oscar earlier I would have saved myself a lot of grief.

Ok, I did some more testing. First I varied the flow to slow and then fast, then I increased the boiler pressure for the last test and tried to maximize the brew temperature.

Test setting: Slow flow 30ml/30s, Steam+0s, 10s recovery,
First flush from idle: S+3s, wait 60s, s+0s, brew,
2min intervals (PF 1min attached, 1min removed)

(s) 1 2 3
05 92 93 94
10 92 92 93
15 92 92 93
20 92 92 93
25 91 92 93
30 91 92 93

I was little generous on the first cooling flush, thus the first set is a bit low on temp, but more importantly the temps are stable, not rising. Then the fast flow.

Test setting:Fast flow 120ml/30s S+0s, 10s recovery,
First flush from idle: S+1s, wait 60s, s+0s, brew
2min intervals (PF 1min attached, flush s+0, recover 60s)

(s) 1 2 3
05 95 94 94
10 95 94 94
15 94 94 94
20 94 93 93
25 93 92 93
30 92 91 92

Because in the first test the temps were a little cool, I was careful with the initial flush on this one. Especially, because the flow rate was going to be high on these runs. The first run started one degree hotter than usual. The other two ran a bit cooler. The most important thing here is the profile. Even with the high flow the brew temp is relatively stable within the shot. Only 2-3 degrees of variance. So the Oscar brew head design is certainly very capable in terms of varying brew flow.

Then in the last test I tried to increase the brew temp a little more so that I could brew in 92-96 degree range by controlling the hx-flush. I increased the Sirai pressurestat by turning it 3/4 turns hotter. This increased the boiler pressure by little more than 0.05bar. Also, I tried to minimize the flushes so see what effect that has. I only flushed the steam from the group.

Boiler pressure 1.63bar (max), vol in 30s: 50ml
Test setting: S+0s, 10s recovery,
First flush from idle: S+0s, wait 60s, mini flush, brew
2min intervals (PF 1min attached, 1min removed)

(s) 1 2 3
05 96 94 94
10 96 94 94
15 96 94 94
20 95 94 93
25 95 94 93
30 94 93 93

The first run produced the results I was aiming for. The 2 other runs showed that, the oscar needs more than two minutes recovery time for higher brew temps. I guess I will still need to do a last test where I vary the recovery (interval) time between shots.

I am beginning to like my Oscar more and more

Great info FinnLight! Thanks for taking the time to do this. Helpful info as I get to know my new to me Oscar.
-Daran