topbanner.gif
Login
Username

Password




Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
Shoutbox
You must login to post a message.

snwcmpr
03/20/2019 4:17 AM
Eth Yirg Nat Idido Gr1 today.

snwcmpr
03/18/2019 8:15 AM
1 lb roasted lasts a week. I can taste the flavor increase to peak, then drop after. Still very good, but interesting to follow the wave of flavor rise and fall.

snwcmpr
03/11/2019 4:10 PM
Roasted Panama Gesha Esmeralda yesterday. Yummy.

Beebee74
03/11/2019 3:10 PM
I seem to have lost any roasting touch I thought I had. Very frustrated to be wasting time and money. I’m hoping someone can provide some insight on roasting at high altitude. I’m at 4400ft. Thank

aicardo
03/07/2019 8:50 AM
tb

Users Online
Guests Online: 5

Members Online: 1
baldheadracing

Total Members: 5,867
Newest Member: SONDR
In Memory Of Ginny
Donations

Latest Donations
renatoa - 2.00
JitterzZ - 50.00
PRABHATH COFFEE WORKS - 10.00
John Despres (Scene... - 25.00
snwcmpr - 10.00

View Thread

Who is here? 1 guest(s)
 Print Thread
Massive Fluidbed Heat Controls
Dowst
I posted this over on Home Barista, but figured you guys over here on the Fluidbed forum would be very helpful as well.

I have been scheming an entrance into the coffee business for several years now, but have found the possibility to present itself only recently with the acquisition of an enormous, one-of-a-kind, untested and incomplete fluidized bed roaster. This roaster was given to me by a good friend and exceptionally bright and eclectic engineer and machinist, who originally designed and built it for roasting Cacao. Unfortunately, the customer disappeared near the end of the build, leaving his creditors and the machine behind.

Fabricated entirely of TIG welded 304 Stainless Steel, the roaster is about 25" in diameter, 50" tall and about 300lbs. A 5HP 3-phase electric motor drives the 15" impeller, and heat is provided by a 230V NiChrome element of unknown Wattage (I estimate around 7500W based on arbitrary comparison with a pottery kiln I have), wrapped around a machinable-lava insulator core. The untested design premise is that the beans will levitate and spin inside the roaster, and will be shot out of the laser cut aperture in the side when the roast is complete and the door is opened. The design allows for up to 100% recycling of air, and the roaster will be insulated from the outside to prevent excess heat loss. Two doors on the top allow for the loading of beans, along with viewing the roast through machined but yet to be installed high-temp glass windows. I am working out the details for chaff extraction (perhaps an additional aperture will be laser cut near the top to provide a vacuum outlet) as well as preventing beans from being trapped in the louvers. I may need to fabricate a new louver plate with smaller outlets if bean capture becomes an issue.

I am quite capable when it comes to mechanics and DC based electronics (restore vintage cars for a living), but industrial electrical controls are new to me. I have figured that I will be running the 5HP motor on a VFD (variable frequency drive), which will convert single phase 230V to 3-phase 230V and allow me to control motor speed/airflow within the unit. I expect that at full speed, the 5HP motor could launch 20LB clear out of this solar system, which I may end up needing to do if I can't get a decent handle on temperature control!

I have spent countless hours reading about PID control, thermocouples, TC4 Units, Artisan, Arduino and the like, but am having difficulty assembling this information into a cohesive and actionable understanding about how I will go about applying heat controls to this roaster. I think that I have an elementary understanding, but know that it is not nearly complete enough to begin buying parts.

Understood that I will be driving the element with a PID temperature control unit (such offerings I see can be sourced from Fuji and others), and will be monitoring environment and bean temps via thermocouple for roast control purposes (Inlet Temp, Outlet Temp and Bean Temp if feasible with beans moving at high velocity), but how all this interfaces is a confusion to me. It appears that many roast control softwares (Artisan) have internal PID control, doesn't this make the Fuji controller redundant, and just an expensive place to plug thermocouples into for data packaging and export? The TC4 seems to be the unit that brings all the data together to allow for a seamless control, but it is not available. What is the current substitute? What is the practical difference between a $40 Inkbird PID (maybe no data outlet?) and a $400 Fuji? That may go without saying, but I am under the impression that the SSR paired with a PID will pulse current to the element continuously, as opposed to a rudimentary oven controller that is full on/off and produces large swings. Full power to the element for too long will cause it to overheat and self destruct, I was under the impression that PID control will prevent this sort of thing?

I would love all your input--thoughts, insights and commentary. Do you guys think this thing will even work?
Dowst attached the following images:
3_10.jpg 6_5.jpg 1-2-dsc_0910.jpg 2_12.jpg

Edited by Dowst on 05/17/2018 3:08 PM
 
JSA Coffee
Check out www.phidgets.com for USB based controls as an alternative to Arduino. Dig into the forums there, plenty of good information to point you in the right direction. I am not running them, but have done a bit of reading about it..

I can't answer if that machine will work or not, smarter people than me will be here to answer that question. I am using an 1800 watt heat gun element for a pound on my little fluidbed, and that is push the limits to the edge.
 
renatoa
Welcome !

You wrote us some impressive details about this build, but... before digging more, what are expected capacity of this machine ?

The Artisan PID is not that great, at least in my experience, and I can detail reasons, if interested.
That's why people prefer either manual control, either an external controller.
With that massive size and probably thermal inertia you have there you can afford to roast manually, imo, will be about 4 power changes during the whole roast cycle.

Fuji can be found used for less than $100, it does a lot more than a simple C100 style PID, things like ramp/sink and manual PWM % control, that are a must in coffee roasting.
Plain PID is a no go, imo, especially for fluid bed, could work for drum oven, where thermal inertia is bigger. It is basically also an on-off operation, just smaller swings, 2 seconds cycle instead 20-30.
 
btreichel
All electric heating elements are switched on and off. The fastest is zero crossing detection with a traic being fired at a specific time within the ac power cycle. Doing this makes it seem like 0 to 120 volts to the heater. Same applies to 3 phase, just more complicated. This basically runs at 120HZ per phase. An ssr you can control (normally at slower speeds) to get the same effect, controls are easier.

The trick is to switch the power faster than the heating coil can respond. Just like a human eye and a tv screen, it just seems like continous picture because the switch happens faster than the eye can detect.

So it depends on your heating element, and the controler can limit the maximum power that it passes to the element. A PID controller works to maintain a specified value. most of them allow you to change the value 'on the fly' So, once you get the heating element being controled to a desired temperature, you can let the pid control the desired temp.
 
Dowst
Thank you guys for the replies!

I have looked into the Phidgets, it looks like the 1048 temp module is often used for this application? I have read that Omega Thermocouples are the way to go as well, any thoughts/preferences. It seems that this would only provide data-logging capability, which may not be an issue if I'm not looking for complete automation like Renatoa suggests.

I am not sure about total capacity of the equipment. Physically, I imagine that it could fit about 30-40 pounds of green beans with headroom to spare, but I am almost certain that the heating element is much too small for this capacity if I am shooting for around 15:00-20:00 roast time, but I am not sure given the convection nature of this machine. Less than 5 pounds seems like it would get lost. I would like to aim for 10-20 lbs per roast. I did some Btu/hr conversions and compared to a Probat 11lb drum roaster, which uses 48,00btu/hr, or 14.1kW/hr to get a sense for what may be possible. Of course with a gas powered drum, I would imagine many of those Btu's are lost to atmosphere, while this roaster could be insulated and recycle the majority of the heat in a convection cycle. To be sure, I do not actually know the output of the element. I am guessing in the neighborhood of 7,500W. I have posted a picture of the element below.

I had read about the ineffectiveness of a cheap PID to control a meat-smoker (excess overshoot and poor control), so it does not surprise me that this is not an option. That would be way too good to be true. Btreichel's description of the SSR switching quickly to effectively create a larger/smaller voltage confirms my thoughts about how the system works to control heat. I assume that this switches very quickly to avoid large temperature swings and overheating the element? Overheating probably wont be too much of an issue with this roaster since the CFM past the element will be large. I read there are two main types of SSR, one that starts again where the Sine wave was stopped, and one which always starts at the crossing point of the Sine wave. Which is preferred? I imagine the former would be better? Any suggested manufacturers/models? Probably in the neighborhood of 30-50 Amps.

So if manual control was the choice (which I have no problem with, especially if I am using datalogging to watch RoR etc), I would use the Fuji controller to do this, or some other manual method like a variac? Variac seems to be the electric equivalent of the gas control needle-valve. I am thinking it would be better to use the Fuji, as this would allow for a more repeatable roast profile and eliminate large swings/inaccuracy created by adjusting by hand via variac. What are the Fuji models to look for, I would love to pick one up used. Trying not spend $1000's to get this thing operational before I know if it will even work to provide exceptional quality coffee.

If I use a Fuji, do I need separate thermocouples for that unit and the Phidget?

Bean temperature would be great to measure accurately, but I wonder how possible this would be. I expect I would place a probe at the elevation where I would expect the beans to be rotating. I have heard mixed reviews about accurately measuring BT. I think I would place the Inlet Temp thermocouple down below the louver plate and one before the air is reheated. What are the best types of thermocouples for this sort of application?

So far my parts list is:

Fuji Controller of some type
Phidget 1048 Temp module
3 Omega K thermocouple (one to measure BT, one pre-heat and one post heat)
Dowst attached the following images:
img_0509_1.jpg img_0513_2.jpg
 
greencardigan
I'd be finding a multimeter to measure the resistance across the element leads. That will let you calculate the power.

A random fire SSR combined with zero cross detection will give you a smoother power delivery. However most of us only use that method for controlling a motor which don't run smoothly otherwise. A zero cross SSR can be easier to control but could provide more of a pulsed output. My 750g roaster used a zero cross SSR switching at 1 Hz. So for 50% power it's on for 1/2 a second then off for 1/2 a second. This seems to work fine. I'm not familiar with the fuji style PID controllers so not sure which sort of SSR they can switch.

The other thing I'd be doing first is getting the motor running and getting some beans in there to see how it circulates them. Maybe measure some airflow rates to help determine if you have enough heating power.

If it were mine (and I wish it was!) I'd use a TC4 (or something compatible) with a zero cross detector. That gives much more flexibility and customisability but requires a bit more of a learning curve probably.

EDIT: I just checked the fuji PXG4 data sheet and it does down to 1 second switching cycles. But maybe there are other models that can do phase angle control.
Edited by greencardigan on 05/18/2018 1:13 AM
 
Dowst
greencardigan wrote:

I'd be finding a multimeter to measure the resistance across the element leads. That will let you calculate the power.

A random fire SSR combined with zero cross detection will give you a smoother power delivery. However most of us only use that method for controlling a motor which don't run smoothly otherwise. A zero cross SSR can be easier to control but could provide more of a pulsed output. My 750g roaster used a zero cross SSR switching at 1 Hz. So for 50% power it's on for 1/2 a second then off for 1/2 a second. This seems to work fine. I'm not familiar with the fuji style PID controllers so not sure which sort of SSR they can switch.

The other thing I'd be doing first is getting the motor running and getting some beans in there to see how it circulates them. Maybe measure some airflow rates to help determine if you have enough heating power.

If it were mine (and I wish it was!) I'd use a TC4 (or something compatible) with a zero cross detector. That gives much more flexibility and customisability but requires a bit more of a learning curve probably.

EDIT: I just checked the fuji PXG4 data sheet and it does down to 1 second switching cycles. But maybe there are other models that can do phase angle control.


Just measured with a low quality multi-meter, 17.7Ohms across the two leads. So If W=(A^2)*R, and I assume that I'm going to put 5 Amps across the element, this means the power should be around 11,000W. I will have to remove the core to confirm, but the NiChrome wire looked to be about 16GA in size. So it seems the element should be up to the task. If I could put 10A across the wire, I would be looking at nearly 18,000W. That should allow for a pretty decent capacity, my guess. Could also reconfigure to a larger diameter wire if I needed more capacity.

I agree that next step is getting the motor turning. I have already spec'd a VFD, but was waiting to purchase because I wanted to order both controller and VFD at the same time from the same supplier, but it is probably premature and I think I would rather get a Fuji off eBay anyway.

I see that the Fuji PXG is the preferred model (also most expensive), but am seeing lots of PXR's and PYX's on ebay. I suspect that these have slower cycle times, and may not be suited to this application. Are these other units an option if I do not mind running manual, or is their cycle time going to be too slow to precisely manage heat?

I had originally planned to control this machine with a TC4, but it seems that these are not available. I have a good amount of experience soldering automotive DME's and the like, so could probably build the board if I could get all the parts.
Edited by Dowst on 05/18/2018 2:09 AM
 
allenb
Hi Alex and welcome to HRO!

I agree with Greencardigan that the first thing to be doing before investing in any controls is to prove the system will function as a coffee roaster as a fluidbed and provide adequate bean movement. Unfortunately, determining if your heat source is up to the task without having a typical intended batch of beans circulating in the roast chamber is very difficult if not impossible. One cannot make accurate calculations as to pressure drops and resulting air flow without actually running it with a batch of beans in the roast chamber.

Once you know that this machine will adequately move the beans, my choice for control would be to use the Fuji PXG with 0-10 volt proportional output controlling a Crydom proportional solid state relay. This will give you a continuously variable output. Roast profiles can be programmed via ramp/soak segments and works great. I've used the Fuji controllers with ramp soak for a long time and love them. Easy to set up, robust and can handle industrial conditions without crashing in the middle of an expensive 10 to 20 lb batch roast.

https://www.instr...ure/17396?

https://www.mouse...egMA%3D%3D
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
allenb
Dowst wrote: Just measured with a low quality multi-meter, 17.7Ohms across the two leads. So If W=(A^2)*R, and I assume that I'm going to put 5 Amps across the element, this means the power should be around 11,000W.


In doing some calcs, at 17.7 ohms at 220 volts, you will have around 12.5 amps and 2734 watts. (E/R=I) amps = 12.5 x 220 V = 2734 watts. Not sure that even with recirc if this will do 10 lbs but I could be wrong.

Please check the coil resistance again to be sure.

Allen
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
renatoa
Me too I doubt, in a TO oven we have total air recirc, and 1300 Watts are good for at most 800 grams for FC at minute 12.
 
Dowst
Thanks again to everyone for your interest and thoughts on this project. Your support and input makes me feel like I have a much better handle on how to approach finishing this project.

I am also concerned about heat output of the element. I had originally miscalculated the estimated wattage of the element by using a bogus online calculator. 3000W is no question much too little heat output. I am concerned that the fluidbed action will take place best only when the roaster is very full. A 5LB bag from Sweet Marias looks lost when placed inside--I worry that it will take 20LB to achieve even fluidbed movement, and it is impossible that 3000W will roast this in adequate time without baking the beans. There is of course the fact that there will be heat recycling within the unit, as well as a decent thermal mass to help maintain heat, but not enough to make up for much too small of an element. I suppose I will be looking into a different element winding material. I believe I will need at least 10,000W for this machine to function ideally. Perhaps it will not be too big of an issue, I imagine there are larger output winding available to increase the power.

But first step is getting a 230V outlet near the machine and wiring up the VFD. Will keep everyone updated!
 
renatoa
First step would be, imo, to see this machine moving even 5 kg of greens...
It should be huge for this task, so far I don't have a clear idea about roast chamber dimensions.
 
allenb
Chart compiled from various commercially available roaster specs for fluidbed and drum.
Fluidbed--------------------Drum

lbs--- BTU's---KW---------lbs---BTU's---KW
1-----10200--3kw---------1----6800----2kw
2-----17000--5kw---------2----10200---3 kw
5-----25500--7.5kw-------5----17000---5kw
8-----34000--10kw--------8
10----40800--12kw-------10---35700---10.5kw
12----47600--14kw

Looks like without any recirc you would want around 12 kw for a 10 lb batch. With recirculating you will obviously need much less, maybe start with 6 kw?
I'm assuming you have dug around the element enough to rule out the possibility of the heater being composed of two sections and fed from two separate power feeds? Looks like a third wire below the coil.

One thing to keep in mind with this design is that the beans aren't intended to become airborne as in a spouting bed fluidbed roaster but are to hover slightly and travel around the roast chamber. See the bean movement in this roaster for an example:


Edited by allenb on 05/20/2018 12:46 PM
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
Dowst
Have an electrician coming to get another 230V outlet in the barn. Thinking I will be going for a 60Amp breaker, at least. We'll see how much can be achieved with the current panel/wiring. Crossing my fingers that it won't be a major operation.

I pulled the heating element again this AM. There is indeed a third pole, but it is unused. Not sure what it was originally intended for.

In any case, I measured the coil diameter to be .057" which works out to AWG15. Based on a spec sheet from WireTron (https://wiretron.com/wp-content/uploads/2017/04/NiCrTechTips.pdf), AWG15 NiCr wire is about .20ohms per foot, so I guess if there's 17.7 ohms total resistance we are looking at about 88.5' of straight wire, or roughly 20' coiled wire which appears to be accurate based on the size of the core and basic math.

If I wanted to increase the wattage of the element, and still have it be the same size, I could re-wire the core to AWG8 wire, which provides twice the output per foot (.4ohm). This is apparently the biggest off the shelf wire, and would increase the elements resistance to about 35ohms. This seems like it would be better suited for the capacity of the roaster, but we will see.
 
chaff
Hi,
(edited , I just read the wire gauge link, did a decimal get missed ? The 8AWG looks like 0.041 per foot, about a fifth of the 15AWG at about 0.21 so, as you say, more power per foot but resistance a factor of about 5 less. ( I'll have to read some more comparing wire size of my clothes dryer, which is 5600w 24A three wire 110/240v rated.
Edited by chaff on 05/25/2018 10:57 AM
 
Want to Help?
By donating you can ensure that HR will continue to be open for years to come. Expenses are 100% transparent and can be viewed publicly at all times. You can find more information on donations and why we request them by reading up on the forum topic here. Thank you.

Dowst
chaff wrote:

Hi,
(edited , I just read the wire gauge link, did a decimal get missed ? The 8AWG looks like 0.041 per foot, about a fifth of the 15AWG at about 0.21 so, as you say, more power per foot but resistance a factor of about 5 less. ( I'll have to read some more comparing wire size of my clothes dryer, which is 5600w 24A three wire 110/240v rated.


I am such a knucklehead and indeed missed the decimal when comparing! Thank you for noticing that! I guess that I need to read that entire PDF. You'll have to forgive me, as this AC current thing and general electrical design are new to me with this project. I am learning quite a lot!

So basically If I am hoping to make say, 7000W from 230V, I know that I will need to draw 30A across an element with 7.5Ohms. I guess the chart will help me figure out what spec I need to achieve this resistance, but it doesn't seem likely that it can be achieved with the same length of coil. Are there other materials available to achieve this goal?
 
chaff
Those power and current levels are more Topsy than popC, this is a 'could it be done ? ' question,
I've seen a dryer replacement element come as two coils joined, the joined centre tap being the third wire. If your 17.7 ohm coil were centre tapped you'd have two 8.8ohm coils joined.
putting 230v from end to end, 17.7 ohms, results in a flow of ~13amps, about 3kW

The pdf suggests nominal max current flow for 15AWG is at 240v, 16ohms: some 15amps

Applying 230v across each of your 8.8 ohm half coils, with 26amps flowing in each coil, would be about 6Kw per coil, 12Kw and 52amps total. Would it be practicable to limit current to alternate each coil every half cycle for an average power of 6kW, what's the effect of 26amp half-cycle current on 15AWG wire and how do you prevent what could happen if 230v gets applied to both coils together while there's not enough cooling air flowing ?
 
Dowst
chaff wrote:

Those power and current levels are more Topsy than popC, this is a 'could it be done ? ' question,
I've seen a dryer replacement element come as two coils joined, the joined centre tap being the third wire. If your 17.7 ohm coil were centre tapped you'd have two 8.8ohm coils joined.
putting 230v from end to end, 17.7 ohms, results in a flow of ~13amps, about 3kW

The pdf suggests nominal max current flow for 15AWG is at 240v, 16ohms: some 15amps

Applying 230v across each of your 8.8 ohm half coils, with 26amps flowing in each coil, would be about 6Kw per coil, 12Kw and 52amps total. Would it be practicable to limit current to alternate each coil every half cycle for an average power of 6kW, what's the effect of 26amp half-cycle current on 15AWG wire and how do you prevent what could happen if 230v gets applied to both coils together while there's not enough cooling air flowing ?


This is an interesting thought that did not occur to me. It sounds like it would be pushing the bounds of what is possible, as you say, which demands the question of "why?" to be asked. My goal here is to simply start roasting coffee on a larger scale, not to "re-engineer the wheel" or take home the ribbon at the local science fair!

In any case, the core does have a third pole that could be wired as you suggest.

Looking forward to seeing how many LBs of beans this this can move, I loved that video!

Why not bail on the electric element and switch to a propane heat source? Surely this would be easier and provide plenty of headroom.
Dowst attached the following images:
2-dsc_0034.jpg 1-dsc_0035.jpg
 
btreichel
reasonable idea since 230 circuits are scarce in us residential, and roasters that size are more commonly gas fired.
 
Will2
I do not see any problem in how to make a new winding for the required power.

Here in a similar post I described how to make instead of 1.7kW winding from wire 0.5mm new element with power 4kW of wire 0.7mm.

https://forum.hom...post_64427

up.picr.de/31130974us.jpg
Viliam
 
Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
Fluidbed roaster build... questions... Fluidbed Roaster 36 03/14/2019 1:16 AM
Heat Gun Driven Fluid Bed Roaster Fluidbed Roaster 26 03/07/2019 2:24 AM
Blower > Heat > Min distance??? Fluidbed Roaster 6 03/06/2019 8:46 AM
EMT Connector Heat Rating Building a Coffee Roaster 1 03/01/2019 3:34 AM
Gas Fired 1 lb Fluidbed Build Fluidbed Roaster 42 02/19/2019 4:39 PM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2019 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX