Sunbeam EM6910 vs Rancilio Silvia

21 April, 2007

Sunbeam EM6910 and Rancilio Silvia

Contents:

Introduction

For years the Rancilio Silvia was arguably the most popular top-end home espresso machine. Websites, forums and newsgroups were devoted to discussions, and it would be in the top bracket of machines with the information on tweaks and tune-ups in its sector of the market, if not entirely.

Until 2005.

In conjunction with world barista champion Paul Bassett, Sunbeam designed the EM6900 – a machine with computer-controlled dual thermoblocks, a semi-commercial portafilter and 58mm filter baskets, resulting in undoubtedly their closest product to a commercial machine by a long shot.

As with any new model, the 6900 was problematic and forums were ablaze with comments such as “keep the box, you’ll need to return it” and stories of faults in almost every area of the machine. Inside information from Sunbeam suggested the demand for the 6900 was far beyond their initial expectations, which resulted in production and repair facilities being overstretched to the point they were unable to maintain a decent level of quality or reliability with the product.

It was good to see Sunbeam obviously took a lot of the negative feedback they received in a professional spirit and released the 6910 towards the end of 2006 – a model with all the good points of the previous model but improving on what let it down. It also equipped the advanced user with the capability to tweak some of the internal settings to optimise performance.

So now the Rancilio Silvia has a real contender in the top home espresso niche of the market. Or so we’re told.

After spending a few minutes helping a friend with milk on his 6910, my interest in the machine that had aroused so much lively forum activity was definitely piqued. I was subsequently fortunate enough to secure a 6910 for a few days to compare alongside my Silvia. It should be noted from the outset that I am a Silvia owner, however the 6910 wasn't on the market when I purchased my Silvia so the outcome isn't pre-determined in the Silvia's favour. I have attempted to look at both machines as a new owner would and report accordingly.

First Impressions

When one looks at the specification sheets for the two machines, it’s clear their internals are quite different, creating plenty of areas for potential differences to arise. The 6910 has two aluminium thermoblocks, one for steam and one for water (which are apparently stainless-lined); the Silvia has a single 300ml boiler made from brass. The 6910 relies heavily on electronic control; the Silvia has traditional thermostats and 240v toggle switches. The 6910 has plenty of user-friendly design innovations, the Silvia appears to have followed an ‘if it ain't broke don’t fix it’ philosophy and stayed entrenched in the 90's.

EM6910 and Silvia control panels
Sunbeam control panel (top), Silvia's (bottom)

Both units appear impressively solid and well built. Side by side, the 6910 is a larger unit than the Silvia. Including the hot water and steam knobs, the 6910 is 330 x 330 x 375 and weighs in at 10.2kg. The Silvia is physically smaller at 240 x 290 x 345 but slightly heavier at 11.5kg. (width x depth x height in millimetres, weight with the water tank removed). The 6910 employs soft-touch buttons on the front top edge of the machine while the Silvia relies on traditional 240V toggle switches. Both are easy enough to use, with the 6910 being the more intuitive to an uneducated user.

The Silvia’s been around for a number of years and next to the 6910 it shows its age. Its styling is notably rectangular and boxy compared to the more contoured profile of the 6910, and the 6910 has a stack of user-friendly features lacking on the Silvia:

quote about sunbeam user friendliness

Aesthetics are a personal thing so I won't go into that, but as far as user friendliness is concerned, the 6910 certainly wins the day. The Silvia’s not poor, it’s just that the 6910 is better. As an appliance manufacturer, Sunbeam probably have more experience with what the average customer wants and it shows. Being a much newer design this is to be expected, but Rancilio can be accused of having rested on the laurels of Silvia's success, despite having had plenty of time to make any number of improvements.

Edit 21 September, 2009: Since this review was written, Rancilio have updated the Silvia to include (among other features) a ball-jointed steam wand, a revised group shroud and a user-adjustable over pressure valve.

Image of portafilters
Silvia (L) and EM6910 (R) - portafilters and accessories

Portafilters are often an indicator of the quality of the rest of the machine. 6910’s weights 526g with no baskets and the handle is shiny hard black plastic. Silvia’s weighs 584g with no baskets and handle is matte semi-soft black plastic. Both feel solid. The 6910 portafilter includes a black false floor which is designed to keep the espresso away from a cold portafilter, on the assumption most people are home users who wouldn’t bother to warm it up properly. Contrary to initial perceptions, it’s not a crema enhancing device, but does collect some nasties if not cleaned regularly. The spring on the 6910's portafilter is a lot thicker than that of the Silvia, which is probably because greater tension is required to retain the unridged filter baskets supplied by Sunbeam.

One thing I like about the 6910 is the separate hot water outlet, though being thermoblock-driven the water comes out in squirts rather than a continual flow. Ultimately it gives hot water though which is what matters. The Silvia uses the steam wand to dispense water, which it does smoothly, but the separate outlet on the 6910 is a nice touch.

Of course you don’t get something for nothing, and there are some cut corners on the 6910:

Power up

The 6910 definitely heats up quicker than the Silvia, and this is probably for a couple of reasons.

The net result of this is that you can feel significant heat coming off the 6910's showerscreen within a minute or two while the Silvia is only just not cold. Supporting the 'thermal inertia' argument is the fact that 2 hours after turn-off, Silvia was still warm around the group while the 6910 was cold.

Is that a good thing or a bad? The time it takes to get things to operating temperature is critical for some in a domestic environment, where you may bring friends home for coffee and want to serve within 10 minutes of arrival. If it’s cold 20 minutes after turn-off, that’s also probably a good thing, to minimise accidental contact burns.

So from a practicality side of things are all well and good for the 6910; but is that the end of the story?

I'm no expert in thermodynamics so I won't endeavour to go into the details, but aluminium has a higher heat conductivity and specific heat capacity than brass, which makes it favourable for this sort of purpose as opposed to brass. However water can hold far more heat energy than brass or aluminium, so the fact the Silvia eventually has 300ml of very hot water inside makes it very stable to any temperature changes, such as the inrush of cold water during brewing. In theory an identical assembly to the Silvia's but made from aluminium may be the ultimate winner, but that introduces corrosion issues with water, heat and electricity inside a 'soluble' metal.

So the 6910's assembly essentially creates a smaller heat 'battery' than the Silvia's, which accounts for its quicker warm-up time (bigger batteries take longer to charge). However that smaller heat 'battery' also means it's less inherently stable to change of temperature, such as when cold water suddenly arrives during brewing. This in turn means if the 6910 is going to have a better thermal stability than the Silvia, it has to do it some other way than relying on a hunk of hot metal around a lot of hot water. The 6910 attempts this through its electronics; we'll see shortly how good a job it does.

Good practice would still be to leave both machines on for 45 minutes or so, and after being on for about this long, the group surround on the Silvia sits around 60 degrees C. By contrast the 6910 sits just under 50. I expect real values were higher than this but it was difficult to get enough surface area on two round surfaces to effect enough heat transfer, but the relative difference is both likely to be accurate and felt so to the touch.

The EM6910 also provides the advanced user with the capability to adjust certain parameters. Unless otherwise stated, the 6910 tests were run with the following settings which give the highest temperature water and steam, as well as the wettest steam but the greatest volume:

Some early tests were run with the steam pump rate at +0.2 before it was apparent it was set at that point, and mention is made of this.

One thing that did annoy me a bit about the 6910 was the noise; apart from the louder noise when brewing, there were the idle characteristics of both machines when up to temperature - the 6910 clicks away to itself every minute or so; the Silvia's thermostat comes on for just 50 seconds every 17 minutes. That could well be attributed to either better thermal characteristics of the Silvia or the oft-reported wide deadbands in its thermostats, but either way it's more pleasant company.

Brewing

Unfortunately I wasn’t able to complete the side by side comparison with best attainable shots. I was using some 5 Senses Colombian and basically exhausted my stock before I had quite finished. Both machines required similar grinds for best performance, but my previous experience indicated the Sunbeam required a slightly finer grind than the Silvia. I used a Mazzer Mini to grind the beans just prior to brewing.

An interesting note is that the Silvia baskets are smaller than the 6910’s – Silvia typically holds 14g, the 6910 about 20g. These figures are a little different to some have mentioned; these were based on filling the baskets above the rim, tapping three times to settle, then smoothing off with the back a knife. They then received a 10kg tamp. After tamping there would be room to add more to the 6910 to presumably arrive at the 25-28g figures others have suggested but for the sake of consistency I used the same process on both machines. The smaller baskets in the Silvia should put it at a disadvantage in the taste area; we’ll see what happens.

I happened to have a few Pullman tampers around the place, which was just as well since this machine didn’t come with a factory tamper, being an in-service machine. From past experience the factory tamper is a whole lot better than most (including Silvia) get – while it’s plastic, it’s 58mm in diameter and at least looks something like a tamper. It also has a thin stainless steel cap over the tamping surface, but this has a linear brushed finish (like all other stainless surfaces on the machine) rather than a radial finish. That means the brushed surface runs across the face of the tamper as on all the panels, rather than in circles. What this means is that instead of polishing the puck you tend to rip it open! But it’s certainly a better fit than the plastic thing which comes with the Silvia, which is about 3mm undersize.

6910

Brewing on the 6910 can be via manual or volumetric modes, which is another great feature of the 6910. While most purists won’t use the volumetric settings, they’re probably useful to have if you’ve got the other factors about right and just want to mass-produce coffee when perfect espresso isn’t essential. The 6910 uses a form of pre-infusion by pulsing the pump for a few seconds before running continuously. 6910 recommend having the coffee ground so the needle on the pressure gauge stops at the top of the gold V. I found best results came around the middle of the brown band, even though this resulted in slower pours. Annoyingly, though understandably, brewing will still automatically shut off in manual mode about 45 seconds after starting. I guess by this time the pour’s pretty overextracted anyway, but it means ‘manual’ mode isn’t quite ‘manual’.

The 6910 is noticeably noisier than the Silvia when brewing, which is surprising since they presumably use the same ULKA pump (video of Sunbeam in operation). It lends a cheap feeling to a machine that is in most other areas well-built.

Silvia

The Silvia only has one option for brewing, and that’s hitting the ‘Brew’ switch! It’s then up to you to turn it off when you think the pour’s finished. There’s no pre-infusion on the Silvia, nor is there a pressure gauge. The only tools one has is looking at the pour, tasting what comes out and correctly interpreting it all. A tricky exercise for a home user, but something which should become second nature for an espresso fanatic.

The sound while brewing is more pleasant than the 6910; the pump is quieter and with less rattly accompaniances.

 

Temperature stability in brewing

Absolute temperature and thermal stability in brewing are both critical factors to perfect espresso. I ran two separate tests here after the machines had been on for at least an hour with both portafilters in position. Temperatures were measured on a digital multimeter with a K-type thermocouple sitting in the exit spout of the portafilter. Thermocouples have a few degrees of error compared to an absolute standard, but they’re relatively linear so comparative readings can be taken even if the absolute temperatures aren’t quite right. It should also be mentioned that I took two or three sets of readings which yielded different absolute results, but the same patterns were evident in all tests.

The first test measured the water temperature exiting the spouts when no coffee was in the basket, i.e. a free running pump.

Free flowing water graph

The Silvia hits 101 degrees C within three seconds of pump turn-on. Over the next 15 seconds from 3-18 seconds, the Silvia changes temperature by 6 degrees (101-103-99) within a 4 degree range. Remembering the Silvia has a 300ml boiler and that in free flow the pump delivers about 10ml per second, that’s half the boiler replaced with cold water for a 3 degree drop from maximum – not bad, and it's something which can probably be attributed to the thermal energy stored in the group, boiler and portafilter which gets transferred into the water to maintain its temperature. As can be seen, temperatures start to drop off a bit quicker after that.

The 6910 initially lags behind the Silvia by about 1 second at the start due to the pre-infusion phase, but after that very roughly follows the same pattern. Over the next 15 seconds from 3-18 seconds, the 6910 changes temperature by 19 degrees (81-89-85-93-89) within a 12 degree range which isn't particularly good. Interestingly, further along in time it's actually more stable than Silvia; allowing the same 4 degree C range for consideration and starting at the midpoint translates into a starting point of 91 with a range of 89-93 degree C, which covers 7-26 seconds. That’s actually a 19 second range of stability within 4 degrees which is better on paper than the Silvia, but Silvia definitely has the stability early on. The temperature is also 10 degrees lower than the Silvia, which may or may not be a good thing depending on the absolute accuracy of the measuring device.

However that’s not a real-world test since far more water is pumped out than would ever be required for espresso, so the second test measures the temperature of the espresso coming out of the spouts. If our measurements of the first graph were correct, we would expect to see some similarities in the characteristics of each machine between the two experiments.

Espresso temperature graph

Both machines appear to have started pouring espresso around the 7 second mark and had a relatively flat curve from 12 seconds onwards.

The Silvia’s espresso temperature steadily climbs from 7-14 seconds at which point it stabilises around 92 degrees until turn-off at 30 seconds. Over the 18 seconds from 12-30 seconds the Silvia changes temperature by 5 degrees within a 3 degree range.

The 6910 doesn’t reach its maximum until 20 seconds, and then it’s 8 degrees cooler than the Silvia. Having reached the maximum it immediately starts to drop off. Over the same 18 flattish second period, the 6910 changes temperature by 9 degrees within a 7 degree range.

quote about Silvia's thermal stability

There are certainly some obvious similarities between this graph and the first 10 seconds of the free running graph. If the 6910 was going to be more stable, unfortunately by the time it would get there the shot would be thoroughly overextracted, so early stability is clearly of critical importance; the Silvia may take longer to heat up, but its larger thermal inertia makes a difference when it comes to keeping temperatures on the relatively straight and narrow.

In both cases, the espressos had the most body on pours which matched the ‘drip then slow run’ profile. Since the best shots from each machine were from different runs I couldn’t do a side by side taste test. I’m sure some could pick the difference, but both had plenty of body and mouth-feel to my uneducated palette.

So for espresso, Silvia is the winner by a small margin, if only on the numbers.

Texturing milk

If you wanted a machine for espresso only then either machine would do the job in my opinion. However you don’t buy a dual-thermoblock machine if you only drink espresso, and the 6910’s capacity to do both at the same time has been marketed as a big plus. In practice I found the main advantage of the dual thermoblock was that I could start the milk pretty well straight after the espresso pour rather than waiting for the boiler to get up to steam temperature.

One benefit of a thermoblock, at least in 6910’s configuration, is that it gives steam on demand, rather than by storage as with a boiler. It’s exactly the same principle as instant gas hot water systems vs storage systems. So long as there’s water in the tank, the 6910 will keep pumping steam out forever, presumably, whereas the Silvia has a finite capacity within its boiler, after which it must be refilled and reheated. It’s this reheating process which really slows the Silvia down in texturing milk.

The 6910 is a mixed bag in this regard. The capacity to steam immediately and continuously is a big plus over the Silvia's system. Unfortunately for the 6910, the quality of its steam is probably the biggest thing that lets it down. When I tentatively opened the steam valve for the first time I subconsciously expected a rush of steam based on the solid appearance of the unit. Instead I got nothing. I opened it further and got a horribly reminiscent ‘dnk … dnk … dnk … dnk …’ sound of water being pulsed into the thermoblock, after which I got a few dribbles of water from the wand. In fairness, when the steam eventually came it was better than any thermoblock steam I'd seen in the past, but certainly nothing in comparison to the Silvia.

The following video clips show the comparative steam output of the two machines, as well as giving an idea of the different sounds they make:

Video clip of EM6910 steam output (pump +0.2)
Video clip of Silvia steam output

Although my technique improved with time, I found microfoam more difficult on the 6910 due to the lower steam volume. The reduced volume makes it harder to incorporate air into the milk without producing big bubbles – it seems to reduce the range you have to play with to get the ideal air incorporation point. With the steam set to the driest setting it was almost impossible to incorporate any air without generating big bubbles; Silvia users can equate this to opening the steam wand perhaps 1/8 – ¼ turn and trying to stretch the milk. At the wettest setting of -0.2, the volume was passable but obviously a lot wetter and this affected the end result, which admittedly was still an improvement on the dishwasher foam I'd gotten initially! However Silvia has both a greater volume of steam and it's pretty dry, if you're prepared to wait for it.

Here’s some comparisons:

Condition EM6910 time (sec) Silvia time (sec)
240ml milk from 10 - 65 degrees C
Pump +0.2: 105
50
300ml water from 27 - 65 degrees C

Pump +0.0: 83
Pump +0.2: 95
Pump -0.1: 69
Pump -0.2: 64

52

However, given one reason people buy the 6910 is for its capacity to heat more milk in the same time, it seemed fair to test the time taken to steam consecutive jugs on both machines.

In the absence of a dairy sponsor, I used 4 jugs each with 250ml of water for the next test. In each case the wand was immersed in the water to ensure all the steam went into heating the water – stretching wasn’t a priority here. The stopwatch was started after the wands had been bled and as soon as the wand was opened in the water. In the case of the 6910 I paused the watch while I changed jug; this wasn’t necessary in the Silvia’s case as the changeover could be done while the boiler was recharging.

graph showing consecutive jugs of water

The lines are all roughly straight as we're performing the same operation on each jug, so we wouldn't expect one to take longer than the next.

There are two things which should be noted from this graph.

  1. quote about 6910 steam timeConsider the solid purple and solid pink lines; they indicate that while the Silvia gets the first jug of milk out the door quicker having been brought up to temperature prior to the test (perhaps unfairly), the 6910 quickly catches up on account of the zero recharge delay and opens the gap wider and wider as the jugs are heated. The jug to jug time on the 6910 was about 73 seconds (1:13) in all but the last case (for some peculiar reason); the Silvia took about 35 seconds longer per jug at around 108 seconds (1:38). For every two extra people you have round for coffee, the Silvia will take at least half a minute longer to satisfy, so a 6910 will have its owner enjoying the fruits of their labours with a room-full of friends long before the Silvia. This is probably the most meaningful and real-world-relevant result from this graph
  2. Consider the solid yellow and solid blue lines. Forgetting the recharge delay of the Silvia, Silvia leaves the 6910 in the dust for heating the water. Silvia averaged just over 37 seconds per jug; the 6910 was exactly double this at 74.5 seconds. The greatest disparity between them was a whopping 45 seconds to heat the same volume of water.

Conclusion

Sunbeam have done a great job with the EM6910 by using modern technology to address the complex variables in espresso production, and it has some clear winning points over the Silvia. It’s essentially an idiot proof design – there’s no boiler to refill after steaming or if the steam valve leaks; it turns itself off if unused; there are unmissable visual indicators of water levels; there's a false floor in portafilter to prevent its presumed non-preheating from affecting the espresso quality; you get both pressurised and unpressurised baskets to allow people to use freshly ground or pre-ground supermarket coffee. In almost every sense it appears to be designed for a wide range of people coming home with friends, want to get up and running quickly to make coffees. If I was regularly hosting dinner parties or the like, the 6910 is a tool I'd love to have in my kitchen. For under $700 the 6910 makes an unbeatable case of bang for buck.

Yet it’s impossible to be all things to all people, and some of what the 6910 gains on being user-friendly it loses on the quality of the unit and what it puts out. Despite being at the expensive end of the consumer price bracket, I found it hard to block out the cheap heritage evident in the pump hammering of the thermoblocks. While this may be overlooked in time, the steam performance is still less than impressive and will be something that will never fill me with joy. The quality of the steel panels is also a concern, as is the overall longevity of its internals. These factors suggest this certainly isn't a machine I'd be wanting to buy for the long term, and for that reason I imagine it will fetch a lower equivalent product resale price than the Silvia.

The Silvia by contrast is a time-honoured design which uses traditional principles to produce its product. Modern electronics may have come a long way, but the laws of physics never change, and the huge lump of steel that is the Silvia's internals do a better job of what's required than the brain of the 6910. In varying degrees, it will produce superior espresso and milk and will still be doing it just as well in 20 years' time, something which few would attribute to the 6910.

On the downside, the Silvia's certainly not as user-friendly as the 6910, and it requires more ongoing level checks than the 6910 which will alert you in some form when it needs something. It's certainly not the machine to have if you regularly want to produce large quantities of milk unless the consumers are pedantic about the quality of what's in the cup. Furthermore, the Silvia will set you back $60-$100 more. Rancilio would do well to look at the user-friendliness of the 6910, particularly in relation to the water level indicators, and make the next upgrade more than minor cosmetics.

So which is the winner? That depends on how you judge it. If you're likely to be regularly catering for more than half a dozen people who aren't extra-pedantic about their coffee, if you're prepared to accept slightly inferior results yourself and you're on a budget, then the 6910 is the clear winner. No other machine for less than $700 provides the capacity to steam and brew simultaneously, and the 'bang for your buck' is a major drawcard of the 6910. However if "it's what's in the cup that counts" and the other factors aren't quite as important, then the Silvia retains the crown of the sub-$1k home espresso fanatic's dream machine. As a bonus it'll keep doing it for years to come, will provide plenty of scope for tinkering with component upgrades and if upgraditis sets in in a few years' time you'll likely be able to sell it without losing more than about 30% of its retail value.

Neither machine will be ideal for every person, so which one you choose will be best decided by reviewing the list of pros and cons (below) and deciding which set of compromises you're more prepared to live with. Whichever you choose, you will have the capacity to produce very good quality drinks and commence your journey to coffee snobdom.

 

At a glance list of pros and cons

Sunbeam EM6910 - RRP $699

Pros Cons

Quick warm-up and cool-down time
Good espresso quality
Capacity to steam multiple jugs of milk simultaneously
Ball-jointed steam and hot water wands
Water tank low level light and viewing window
Brew pressure gauge
Larger drip tray, easy to clean
Level indicator on drip tray
Manual or automatic brewing modes
Excellent feature set for the price

Cheaper build materials – unlikely to last long-term
Poorer steam volume
Poorer likely resale value when it's time to upgrade
Poorer temperature stability
Non-commercial portafilter with non-standard lugs limits ability to upgrade components

Stereotypical buyer: Someone who wants a machine to produce low or high quantities of very good milk or non-milk drinks without being as pedantic about details

Rancilio Silvia - RRP $759

Pros Cons

Excellent espresso quality
Excellent steam volume
Better temperature stability
Solid build materials inside and out – will still be going in 20 years
Basic technology less prone to failure
Commercial portafilter with standard lugs
Lots of resources, spare parts and upgrade components
Good resale value

Slower to warm up
Slower to steam consecutive jugs of milk
No water tank level indication
No brew pressure gauge
Drip tray smaller and harder to clean
No level indicator on drip tray
Manual brewing mode only

Stereotypical buyer: Someone who wants the absolute best espresso and milk available at the price point, is prepared to wait for it and to spend a bit more, and may like to play around with component upgrades

* Thanks to Epic Espresso in West Perth for providing the Sunbeam EM6910 for this review.

View or post comments about this review

Valid XHTML 1.0
Valid CSSBest viewed with Mozilla Firefox, a standards-compliant browser. Click for free download