Clarification on Decent Espresso design

Need help with equipment usage or want to share your latest discovery?
Marco_83

#1: Post by Marco_83 » Sep 11, 2019, 7:49 am

I read the following post :

Decent Espresso v1.1 shipping!

and I have some question concerning the diagram of how the coffee machine works.

1/ On the auxiliary manifold and the mixing manifold the internal chamber are in communication. Why ? for example, in the first manifold, the cold and hot route are not separated. A small internal pipe, a sort of flow restrictor, make communication with cold water route and hot water route ?
In the mixing manifold why there is a flow restriction inside ? For me a simple communication chamber for mix hot and cold water is sufficient because the flow cold or hot can be regulated through pumps ?

2/ Another thing concerning the position of the flush valve. The position of the valve on the groupheat output is not as usual position. Normally the dip tray flushing valve is a three way group valve at the input of group head. This allow to release pressure in pipes and water in puck is eliminated through depression. Here, with the position of the flushing valve in Decent design, the puck is wet (some water stay on coffee and the end of shot).

3/ Concerning the post in reference, I notice that the 100 W electrical heater on the grouphead is working when the flow rate of water during the shot is low, and off most of time (with classic shop where water flow is consequent). According preheat A and B as explained, the manifold and group valve are in temperature, but if you are in a normal flow mode, the water enter in a cold grouphead (never heated by the 100W heater)... Is there something wrong in my reads...

I don't understand why they don't make a circulation of water at temperature in group head (by using hose of flush route) to preheat perfectly the grouphead ; for this they have just to found the way to include a sort of valve inside the group to prevent the supply of water to the coffee grounds (during this heating step).

do not see here any criticism from me, it's just to understand and bring about discussions :D

User avatar
Jake_G
Team HB

#2: Post by Jake_G » Sep 11, 2019, 8:38 am

I'm posting the image in question to make it easier to follow along with your questions:
Image

I will do my best to answer the questions shortly, but I've got to run and someone else may chime in first. I'll edit this as time allows if any of your questions remain unanswered by the time I get a few spare moments.

Cheers!

- Jake

Marco_83

#3: Post by Marco_83 » Sep 11, 2019, 4:21 pm

Thanks Jakes. Waiting for your or other person replies.

User avatar
sweaner
Supporter ♡

#4: Post by sweaner » Sep 11, 2019, 8:29 pm

Pretty colors!!

I am not sure of the point to these questions. This is what it has evolved to. There are likely many ways to accomplish the same goals.
Scott
LMWDP #248

User avatar
Jake_G
Team HB

#5: Post by Jake_G » Sep 12, 2019, 7:05 am

Marco_83 wrote:1/ On the auxiliary manifold and the mixing manifold the internal chamber are in communication. Why ? for example, in the first manifold, the cold and hot route are not separated. A small internal pipe, a sort of flow restrictor, make communication with cold water route and hot water route ?
Indeed you are correct. I will gladly be corrected by a Decent guru, but my understanding from a systems standpoint is that this first communication allows both pumps to be protected by the single OPV, while still allowing a proportional flow from each pump, based in the desired output of each. That is, the restrictor between the cold water path and the soon to be hot water path (both are unheated at this point) does allow cross-talk between flow paths, but the amount is proportional to the difference in output of each pump and well-understood. I suspect this also plays a small roll in smoothing the pulses from each pump. But I'm speculating at this point.
In the mixing manifold why there is a flow restriction inside ? For me a simple communication chamber for mix hot and cold water is sufficient because the flow cold or hot can be regulated through pumps ?
The restrictor is necessary to get the pumps operating in a controlled region of their performance. The way the machine controls temperature is to first pump 100% hot water through the group to warm things up. The design informs me that most of the water to the group remains heated, with a small amount of cool water added to trim the temperature. The port in the axillary manifold and the orifice in the mixing manifold both serve a common purpose of sharing the overall flow demands between both pumps while also making the output of the cold water pump less sensitive to changes in duty cycle (more precise control) than if the water supplies were segregated prior to the mixing manifold and shared a common cavity therein.
Marco_83 wrote:2/ Another thing concerning the position of the flush valve. The position of the valve on the groupheat output is not as usual position. Normally the dip tray flushing valve is a three way group valve at the input of group head. This allow to release pressure in pipes and water in puck is eliminated through depression. Here, with the position of the flushing valve in Decent design, the puck is wet (some water stay on coffee and the end of shot).
The flush valve operates very similarly to a standard 3 way except that the dirty path between the group and the flush valve never sees fresh water. This also means that fresh water entering the group never sees dirty water. This is by and large a good thing. Air in the group is still pressurized when it fills with water and this pressurized air still evacuates the water atop the puck after the shot is complete. The stories of wet pucks are largely due to the increased headspace of the DE1 relative to more traditional machines. I've talked with folks who have the spacers that have been developed on the private DE1 forum that report dry pucks after installation.
Marco_83 wrote:3/ Concerning the post in reference, I notice that the 100 W electrical heater on the grouphead is working when the flow rate of water during the shot is low, and off most of time (with classic shop where water flow is consequent). According preheat A and B as explained, the manifold and group valve are in temperature, but if you are in a normal flow mode, the water enter in a cold grouphead (never heated by the 100W heater)... Is there something wrong in my reads...
Couple notes here:

The heater is 250W, not 100W. See below:
MrEd wrote: For very slow flow rates, the group heater is actually a pretty good way to control the temperature [This is what JayBeck was referencing above, and the group has a 250W, not a 100W, heater]. As the flow rates increase, we end up in a situation where the water mixer and group controller can fight each other, and I haven't found a way around this yet. This is one of the reasons why the group heater is turned off for faster shots.
The issue isn't that the brew head is cold, it's that the heat coming in from the water has a higher impact on the group temp than the 250W heater and the control schemes don't get along with each other when the group head is actively heated. (This is a common problem with cascade control loops that share a common measuring point and not a problem unique to Decent.) It is my understanding that the group is actively heated during warmup and also in between shots, so a cold group is not a concern with Decent.

Hope this helps,

Cheers!

- Jake

Marco_83

#6: Post by Marco_83 » Sep 12, 2019, 4:29 pm

Thanks a lot :D
For flow restriction and internal communication inside manifolds sorry but you are difficult to understand. Me I read your reply several times and I don't really understand the goal with these internal communications.

Is it for simply equilibrate pressure ?

For the question about heater I note that the most important thing is to manage correctly the temperature of water going to the group head. And also to ensure correct regulation of temperature during the flow.

If the 250w heater is really put on during warmup, or other time when no shoot is operated, this is a really good thing to ensure optimum thermal stability of the system ; and guarantee a perfect coffee extraction.

User avatar
Jake_G
Team HB

#7: Post by Jake_G » Sep 12, 2019, 5:57 pm

Marco_83 wrote:For flow restriction and internal communication inside manifolds sorry but you are difficult to understand. Me I read your reply several times and I don't really understand the goal with these internal communications.
Imagine how my wife feels trying to understand my disorderly ramblings on a daily basis :lol:

Here is a MUCH better explanation of what I was trying to say (and failed) from John Buckman:
decent_espresso wrote: - 1st advantage: the machine needs a lot more hot water flow, than it needs cold water flow. The flow constrictor between the pumps is there so that a good part of each cold water pump stroke goes to the hot water side. Thus, this little hack allows us to increase hot water pump power by about 50%

- 2nd advantage: the flow constricted bridge reduces the amount of cold water that is introduced to the mix, with each cold pump stroke. Since we need very little cold water (normally, we're dropping 110ºC hot water to ~92ºC) we can be more precise if each cold water stroke gives us very little cold water into the mix.

Naturally, the physics for all this are more complicated, but we like the advantages, and we seem to have adequately modelled the physics of this.
Marco_83 wrote:If the 250w heater is really put on during warmup, or other time when no shoot is operated, this is a really good thing to ensure optimum thermal stability of the system ; and guarantee a perfect coffee extraction.
This is my understanding.

Cheers!

- Jake

Marco_83

#8: Post by Marco_83 » Sep 13, 2019, 4:39 pm

Jake I'm really sorry for your wife because with explanation of John I have understand :lol:

But...by considering this... that means pumps are not individually regulated to offer for the cold pump only required cold flow of water.

Me if I have to consider this, I install 2 different circuits (hot and cold) ; and I make regulation on cold pump in order to obtain the smal needed quantity of cold water.

After Decent design is really well thought out if they needs an important power on pumping function. ...they could also select pump with higher flow.

For my point of view, Decent design is fun. But machine is too expensive. 2000€ max will be a more raisonnable price.

User avatar
Jake_G
Team HB

#9: Post by Jake_G » Sep 13, 2019, 7:49 pm

Marco_83 wrote:But...by considering this... that means pumps are not individually regulated to offer for the cold pump only required cold flow of water.
This is technically true, but not really relevant. See below, and I'll try better to make sense :wink:
Marco_83 wrote:Me if I have to consider this, I install 2 different circuits (hot and cold) ; and I make regulation on cold pump in order to obtain the smal needed quantity of cold water.
This would work, but it would be likely end up more complex and no more precise than what the Decent team came up with. Take for instance an imaginary situation where you need a maximum flow of 8ml/s hot water. The makeup of this hot water is 85% heated (too hot) and 15% cold. In your machine, You wouldn't want to use 2 of the same pump (each capable of 8ml/s) because the hot water pump would run at 85% of its rated capacity (6.8ml/s) and the cold would be only at 15% (1. 2ml/s). Trying to run any device (pump, valve, sensor, etc.) above 85% or below 15% of its rated capacity is not ideal as the output becomes much less predictable than when they are run in the "controlled region of their performance", which is generally between 15% and 85% of the rated output.

Instead, you would source 2 different pumps; one large hot water pump to feed the heater and one small pump to feed the cold water side if the mixing manifold directly. The big pump is now bigger than ideal and a small change to control signal gives you much more total flow than a similar change in signal to the small pump. Now you see that you have 2 unique pumps on your bill of materials, and you have 2 very different response curves to model when it comes to controlling the mix temperature and total flow through the machine.

What decent has done is allowed 2 identical pumps to be used. Furthermore, the clever sizing of the ports between the hot and cold water passages has allowed each pump to be operated closer to 50% of its rated capacity (4ml/s from each pump gets you the total 8ml/s). Since the pumps are the same size and running in a sweet spot for response to changes in the control signal, they are able to model the total system flow of the combined output much more accurately and they are able to consistently hit target temperatures by adjusting the ratio of the two pumps. In this example, you could precisely adjust the temperature up or down by running a 45%/55% ratio, while maintaining the same target flow. While this isn't a perfect split between pump output and temperature management, it is modeled well in the decent software and provides consistent results with much less complexity that trying to run two different sized pumps with segregated flows.
Marco_83 wrote:After Decent design is really well thought out if they needs an important power on pumping function. ...they could also select pump with higher flow.
By selecting the pump that they did, they are able to deliver a wide range of flow rates with excellent accuracy and repeatability. Going the route of segregated flows would require a very large pump to compete with the two evenly sized pumps that decent uses.

Marco_83 wrote:Jake I'm really sorry for your wife because with explanation of John I have understand
Words cannot describe the level of patience she has for me. :wink:

Cheers!

- Jake

Marco_83

#10: Post by Marco_83 » Sep 14, 2019, 6:46 am

if they use pump at 50 % of their rated power that means they pilot them through PWM signal.

Their solution (Manifold with flow contributor) is amazing and well think.

Concerning now the flow rate. You mention 8 ml per second with 2 pumps working at 50%.
This flow is it the flow value attached to DE1 or is it as example ?

I read data for a typical ULKA pump : for 9 bar the flow is about 250 cc/mn.
Espresso is done in 25 s for 6cl (60cc). So required flow is 144 cc/mn.
If they pilot each pump at 50% then total flow with 2 pump is always 250 cc... and we need 144 cc ?

Algorithmes are needed for the management of 250w heater to heat or not the group head according the flow needed during a shot. Algorithms are need for the pump control and water temperature regulation due to their design (Manifold with flow contributor). Algorithms are necessary when you put a shot in flow profiling mode. So this is a big calculator :D
With all these algorithms are we sure, as consumer, that we will obtain a perfec espresso :?:

On your profile you have a La Marzocco. I love this coffee machine, but expensive...
Have you test some coffee from a DE1 ? What to you think
I know all people will says it is different. Ok but what else ?