BrunDog 50A eRig - no HLT for me!

[BrunDog]

Thanks. I have some work to do to get it where I want, but I have learned a lot and I am making progress.

I also had a problem finding a relay and in truth, have not found the right one quite yet. I wanted an enclosed DIN rail mount relay and did not find one easily. You need a power relay that is DT (dual throw), can handle 30A, and has a 120VAC coil voltage. Relays like this: http://www.digikey.com/product-detail/en/T92P11A22-120/PB492-ND/365926 have a 30A NO rating and a 3A NC rating. I wasn't initially sure why there was such a NO/NC difference (if it was real or just a rating thing), but I didn't want to be a tester.

So I selected this relay: http://www.alliedelec.com/schneider-electric-magnecraft-300xbxc4-120a/70184975/

As it turned out this relay wasn't perfect either. I learned this the hard way, because despite the 30A rating, it probably should have a lower NC rating like the relays above because it will not successfully disconnect the load when it is running in sparge mode. So, if I release the relay when applying 240V (sparge mode) and the SSR is on, it will trip the breaker supplying it. I believe an arc is forming in the air gap between the NO, common, and NC contacts. Since the NC contact is tied to neutral, there is an electrical pathway there. So, my work around has been to always make sure the output to the SSR is off before turning the output for the relay off. This is easy enough to do in programming, so it has not been an issue.

I may look for a decent replacement relay someday (not a priority right now), or I may wire the other pole in parallel to reduce the current in half through each pole, as this may resolve the problem.

BTW I have been intermittently successful using PID on the BCS for sparge water heating on the fly. Sometimes it dials in and other times it is unstable. I am going to test a new tuning method but will likely just go to a duty cycle, since the flow is very consistent, and the output temp is not super critical. It sucks that the BCS doesn't have autotuning, and I am convinced it is also not calculating or applying output periods correctly.

 

[3683Brewing]

Thanks for the info BrunDog. I'm going to do the same. Great info and great build!

Cheers!

 

[crane]

It sucks that the BCS doesn't have autotuning, and I am convinced it is also not calculating or applying output periods correctly.

I agree that not having autotuning is a pain. I previously used an arduino to autotune and then copied the coefficients over to the BCS. That worked as a decent starting point, but still needed further tweaking.

Have you tried using the PID tuner that is available from the utilities Web page? I found this extremely useful when tuning as it will show you the current outputs of all 3 gain stages along with the resulting duty cycle. It seemed to always correctly calculate the output based off of the error input in my experience. If it is actually miscalculating the output it will tell you part of the algorithm is wrong. In the end it seemed best to set the min integral to 0 and the max to something that is about 50 to 100% higher than the integral value it settles to during steady state. This along with setting the proportional gain high enough to negate the max integral when you are above the setpoint resulted in very little overshoot (0.5F). Take this as a grain of salt as I was not tuning a RIMS tube but rather my HLT filled with 12 gallons of water.

 

[BrunDog]

Thanks crane. I have not tried the tuner as it was not working often with the beta. Using a fast responding system I have the parameters set very tight: output period 1 sec and sample period .1 sec. According to ECC, the temps are read that fast. However, I do not believe that the output successfully works with that small an output period. For example, a calculated 10% output period would be an output of .1 seconds, but I never see the output pulse this fast.

I will give the tuner a try. Ideally, I would have the output set at a certain duty cycle, then have it slightly adjusted as needed to increase or decrease the output temp. Because the inflow rate and temp are consistent, and because the sparge water temp is not super critical, using duty alone would really be enough.

 

[JonW]

For example, a calculated 10% output period would be an output of .1 seconds, but I never see the output pulse this fast.

What is your method to "see" what the output period is?

 

[BrunDog]

My eyes! I know what a very short output pulse looks like, and I never see it. I suppose it's possible the LED tied to the element has a certain on delay, but I would say it never lights for shorter than a third of a second.

I could hook up a scope or other logger to the input of the SSR but I am confident with my experience.

 

[thekraken]

4. The RIMS element runs at 240V (5500W) for heating strike water (via recirculation) and sparge water (via one pass). However, I did not want to have the element run that power for mash recirculation. I felt this was too much heat and could cause scorching even if correctly pulsed to a very low duty cycle. So I implemented a relay to switch the element’s voltage between 240 and 120VAC. At 120V, the element only generates about 1400W. Therefore the relay provides 240V at first for strike water heating, then is switched to 120V for mashing, then is switched to 240V again for sparge water heating. BTW, since the BCS uses different PID parameters for each output, I am using two outputs, which then power the same SSR. Finding a DIN rail mounted relay to handle 30A was a little tricky, and I am still not thrilled with the one I have (relay is circled in wiring construction shot here).

I had been thinking of doing this exact same thing!
How does the on demand hot strike and sparge water work out for you? What %duty cycle do you typically use and what temperature delta do you usually get in one pass?

Also, I use that same flow meter for my RIMS tube, I've programmed hard limits in my controller to shut the process down at >172F for fear of damaging it at it's 80C max temperature rating (also, my mlt as plastic). Do you have any checks to prevent passing too hot liquid through the meter? Or do you ever even come close? I ask because for me at 120v it doesn't heat up fast enough for me to worry about me or the controls not catching it overheating, at 240v I wouldn't know...

Nice setup btw!

 

[crane]

Keep in mind most SSRs have a min pulse threshold in the hundreds of milliseconds. The datasheet for the ones I use state 350ms. So a pulse of 100ms is too short to turn on the output. This correlates with your observations of only seeing the Led light if the pulse is more than 1/3 of a second.

 

[BrunDog]

I had been thinking of doing this exact same thing!
How does the on demand hot strike and sparge water work out for you? What %duty cycle do you typically use and what temperature delta do you usually get in one pass?

Also, I use that same flow meter for my RIMS tube, I've programmed hard limits in my controller to shut the process down at >172F for fear of damaging it at it's 80C max temperature rating (also, my mlt as plastic). Do you have any checks to prevent passing too hot liquid through the meter? Or do you ever even come close? I ask because for me at 120v it doesn't heat up fast enough for me to worry about me or the controls not catching it overheating, at 240v I wouldn't know...

Nice setup btw!

On demand works good but as mentioned above the tuning can be tricky. Because I am using ambient temp water in my garage in FL, the water is over 75 degrees to start, and I can easily get it to boil at 1 qt/min. I can get it to sparge temp and almost 2 qt/min. That would be 100% duty. Going from ~75 - 170 takes about a 75% duty cycle.

I have the flow meter measuring the fluid into the RIMS tube, not out of it. This way, it never sees over 170 degrees (mash out temp). During mash it sees that temp, during sparge, it only sees the ambient water temp.

 

[BrunDog]

Keep in mind most SSRs have a min pulse threshold in the hundreds of milliseconds. The datasheet for the ones I use state 350ms. So a pulse of 100ms is too short to turn on the output. This correlates with your observations of only seeing the Led light if the pulse is more than 1/3 of a second.

Where are you getting that information? The SSR spec I show has a transition time of <= 10ms: http://www.ebrewsupply.com/40a-solid-state-relay-ssr/

I suppose the specification is not correct, so the best thing to do would be to just test it.

Speaking to JonW's point about how do I know the BCS is not putting out short pulses, my experience showed that with an output period of 1 second and a duty of 10%, the output never turns on (LED never lights) and with an output of 90%, the output never turns off (LED always stays on).

 

[crane]

Where are you getting that information? The SSR spec I show has a transition time of <= 10ms: http://www.ebrewsupply.com/40a-solid-state-relay-ssr/

I suppose the specification is not correct, so the best thing to do would be to just test it.

Speaking to JonW's point about how do I know the BCS is not putting out short pulses, my experience showed that with an output period of 1 second and a duty of 10%, the output never turns on (LED never lights) and with an output of 90%, the output never turns off (LED always stays on).

Total brain fart. I think I need a beer. Yeah looking back I see its 10ms. Not sure where I got the 350ms from.

 

[BrunDog]

It's all good! I get those!

 

[BrunDog]

Well, on to the next upgrade... Because I despise cleaning, I decided to implement an "AutoClean" system/program. On the new incoming water manifold is a takeoff port for water, piped with a 1/4" poly tube.
View attachment ImageUploadedByHome Brew1449460338.916611.jpg

That feeds a 24VDC valve which is powered by a relay by the BCS. Zip ties are my friend, especially as I keep adding and changing things!
View attachment ImageUploadedByHome Brew1449460417.291031.jpg

The output of that tees into the line where my water tanks feed in.
View attachment ImageUploadedByHome Brew1449460483.000261.jpg

And here is a look at my cheap/testing CIP sprayer (yeah, it's a sprinkler head). This will be replaced by a legit spray ball. I use the camlock inside the locline sparge ring normally connects to. I used Saran Wrap to view/tune the spray pattern but the photo didn't show well so here it is temporarily off.
View attachment ImageUploadedByHome Brew1449460666.490687.jpg

I made a program in the BCS to perform the process. First a "rinse" cycle starts that fills the MT based on time. Then that water recircs through the pump/sprayer for a couple minutes, then the appropriate valves open and the water is pumped out through a discharge tube. It will do this a few times, then start a "clean" cycle.

For this cycle, the MT fills up to the sight glass sensor. I use that here so I have a consistent and known amount of water. I don't use it for the rinse cycle because the sight glass sensor measures about 3.5 gallons at the lowest level, I only need about a gallon, and the exact amount is not critical for rinsing.

Once filled, the RIMs tube will heat the water to 130 degrees then prompt me to add PBW. It will recirculate, pause for a soak, then dump. After that it will run a few more rinse cycles.

I will add the BK next now that this is proving worthwhile on the MT side. Also, I am going to add some small inline strainers (only for this cleaning sequence) in front of the pump inlets to pull out junk rather than recirculate it.

 

[iijakii]

Looking good man. That's a pretty complicating cleaning cycle for me, but I suppose it all being BCS is nice and hands off.

 

[jcav]

Looking very good indeed! (Wish you could come over and fix my pump issue!!)

John

 

[jcc4507fly]

Keep going...I love what you're doing over there. Fantastical!

 

[jcc4507fly]

Are you using your chugger to power this? If so, how is the pressure with the sprinkler head. Do you think it will it give you enough psi with a cip ball? I've been wanting to add a cip cycle to my bcs controlled brewery but haven't got to that point yet. I was curious how I would go about it and if my chugger could handle it or if I'd have to add an additional pump to the mix...

 

[BrunDog]

Yes, my chuggers are doing the pumping. This is because it's easier, they are in the loop, and they need cleaning also. They create plenty of pressure and flow for the sprinkler head. I tried a cheapo rotating head which threw water more erratically but the pattern was more upright which made it difficult to use. Upside down and it hit the bottom, and right side up and it tried to come past the lid.

My understanding, after reading and speaking with the folks at Brewers Hardware is that they need much more pressure and volume than the chuggers can provide. Realistically, they are made to clean large tuns and vessels, not 15 or 20 gallon kettles.

I will try this after a brew and see how it goes. If it doesn't work, I will look for other random/rotating sprayers.

 

[jcc4507fly]

I'm interested in your findings. Good to know about the one from BH. I will continue to research this as well.

 

[BrunDog]

Did my next brew last night. Made a nice Pale Ale/Session IPA, depending on how you look at it. Of course, each time I brew, I learn new items/issues with my system. Here are the problems/issues that popped up that I need to address:

1. My efficiency has not been stellar, so I decided to tighten up the gap on my mill a scosche... Well, that resulted in much better efficiency but it also resulted in a stuck-ish mash. I needed to stir it in the beginning to keep it going. After that it was ok. I have a NorCal false bottom, so not sure what the deal is. My recirc rate may have been a little high, but not out of line IMO. I had a stuck mash on my pumpkin beer, but was never expecting one here with just malt. I am going to leave it and make sure I do a very good grain condition before crushing next time. I was in a hurry and did it kinda half-assed because I didn't rest the grain after conditioning it.

2. During sparge, on occasion the flowmeter/Arduino said the flow was zero. That caused the Arduino to temporarily open the proportional valve, which solved the problem, but not sure what the deal is. I suspect either noise or just a cheap-ass flowmeter=cheap-ass results. Later on, the Arduino locked up, but only during the AutoClean cycles. I think this is due to EMF/voltage spikes from the inlet water valve solenoid. I will add a suppressor/flyback diode to remedy this. Also I got my eye on a non-contact magnetic/inductive flowmeter on eBay...

3. During the sparge, the RIMs tube water temp was all over the place. This has often been an issue with such a quick responding loop that occurs from direct sparge water heating. It has proven difficult to tune. I am also convinced the BCS PID has errors in it and cannot correctly time fast output cycles. I changed it to duty mode at 65% and it held like a champ. Going forward I will just use this mode. With consistent water temp and tightly controlled flow rate through it, there is no reason not to. The good thing is this will free up a BCS output (was using two to take advantage of different PID parameters) because I have no more outputs left! Thinking a valve to control the oxygenator or an automatic hop/adjunct adder. Any votes?

4. During whirlpool, my Chugger decoupled again. I learned that without any restriction (head pressure) and near boiling fluid, this is causing it. Not exactly sure why. I didn't see bubbles so I don't think it was cavitation from the pressure change. When I added a small restriction by placing a clamp on the hose leading to the whirlpool arm, it resolved. I need to add a nozzle to the whirlpool arm for more velocity anyway, so this should take care of that.

5. The AutoClean rinse process worked but didn't exactly clean out the debris. The reason is the junk drawn through the bottom drain recirculated right back into the kettles because it is a closed system. I have two filters on hand but did not have the fittings yet, so I could not install them. I am going to try them, and if they do not work, I will come up with a drain/feed bucket system to strain out debris using a bucket of water and a whirlpool effect. Also, the sprinkler heads worked but they aren't "violent" enough. I need something rotating that directs more water in one place. I would like to make a rotating T-bar with a nozzle on one or both ends to make it spin, but a quick search revealed not much in the way of cheap rotating fittings. Any feedback is appreciated.

6. The real shocker of the night was when I went to run the AutoClean process. I use the MLT sight glass sensor to measure the fill height of the kettles. Since the MLT and the BK are connected during the fill, they automatically level. Once the sensor indicates there is enough water (to cover the BK element), it switches the valves separating the two and begins heating both 5500W elements to draw the full 50A. While it was filling, I started to smell something... I went over to take a look thinking it was something in the control box as I haven't pulled 50A since my initial tests following the build. Well, it was the BK element which was not under water. It was dry-firing and had a nice, bright, cherry glow to it! Slammed the E-stop and let it cool down. Smelled like crap but it sure took care of the gunk that was on it following the boil! As it turned out, the MLT sensor tripped on a water droplet or two that was running down the inside of the sight glass! This was enough to tell the BCS to go to the next state, which was the heating state. Fine tuning it's sensitivity, and adding one (was already on order from eBay) to the BK sight glass as a safety will prevent that from happening again. The element worked fine after that but I will probably order a replacement just in case.

Of course in the end, I made wort! It was bubbling hard 12 hours later, so I will have beer. The sample was delish and I hit the numbers dead-on. I am happy to be making progress, have the big stuff out of the way, and only have some final-ish programming and minor updates. My system is almost as I originally visioned it. I don't really know what left I have to add or upgrade. So what am I going to do now?!? Building is 90% of the fun!

 

[JonW]

Ahh.... the trials and tribulations of a new brew rig... lol.

Sounds like you're making good progress and getting it dialed in though.

 

[BrunDog]

Ya where is your new rig's thread. Let's go, you're late!

 

[JonW]

Ya where is your new rig's thread. Let's go, you're late!

It's coming! Instead of posting as I go, I've taken pics along the way and I'm documenting certain aspects in detail to include with the write-up. I dropped it off at Brewers Hardware last Friday for a couple of small welding additions and I'm hoping to stop and get it back tomorrow. Then it's just some final wiring (which is plug & play) to do before the test runs. Did I tell you it requires 2 BCS's ?

 

[BrunDog]

I actually saw that in your teaser post. Too bad you need two controllers to get the I/O. I guess your interface makes it a moot point.

I am out of outputs so may have to go the same route. Either that or switch to a PLC or Arduino Mega.

 

[thekraken]

Regarding item #2, how many batches has the meter been through? I haven't experienced this zero reading my self.

 

[jcav]

Looks like your working out the kinks just fine. Glad you got another brew on her already. Going for brew number 2 on mine the week after Christmas, and I hope to solve my pump issues! Learned a lot from your build, and from the issues you had come up, and how you tackled the solutions to fix the problems. Keep us posted on the final tweaking!

John

 

[BrunDog]

Regarding item #2, how many batches has the meter been through? I haven't experienced this zero reading my self.

Sorry overlooked your question... Probably 6 batches? Each had a PBW run after so who knows what that may do to it. What I really don't know is... Was it an actual zero flow reading or an electrical issue? It only happened 3 or 4 times and self-corrected quickly. It could be something or nothing. It could have been bubbles in the feed line, dunno. Could have been aberrant grains wreaking havoc. It is also possible the meter is a POS and just came to a halt with such a slow flow rate. I may do a water only run and see what happens. If it happens again I will debut it live.

 

[BrunDog]

When you skip an HLT and sparge on the fly (starting to wonder why any indoor brewer wouldn't), using a standard foldback element inside a standard RIMS tube, the fluid that flows directly near the element is laminar and takes on a lot of heat while the surrounding fluid does not. This can create local boiling if you don't regulate the duty cycle within a confined window. I learned this early on and wanted to make some baffles to create turbulence, but these were tricky to make without a legit machine shop handy.

What I had done was made a mesh tube and put it around the element, knowing it was not ideal, but was easy as I had the stainless mesh and could roll it and slip it around the element and up into the tube. This helped, but what I did not contemplate (like an absolute moron!) is that the cage would also catch grains that slipped past the false bottom. And no amount of PBW is going to clean those out. I had a rude awakening when I took my RIMS element out today only to find some nasty gunked up grains in there... which smelled really funky - yech!

Attempt two... I bought some stainless wire a while back and today wrapped the element in the wire in a zig-zag fashion. I think this will do a better job of creating turbulence inside the tube and hopefully not catch grains. I will do a brew and check it after. If this doesn't work, I will need to find a way to make the baffles. If I were to build another RIMS for this purpose, I would use a iRIMS like trimixdiver invented. But I have these parts so I need to make it work.

 

[Mountainsax]

Nice! Simple. I'll be curious to see how it works.

 

[thekraken]

When you skip an HLT and sparge on the fly (starting to wonder why any indoor brewer wouldn't)

My problem is I use store bought RO water, I'd need an "LT" to feed the RIMS tube anyway.

Attempt two... I bought some stainless wire a while back and today wrapped the element in the wire in a zig-zag fashion. I think this will do a better job of creating turbulence inside the tube and hopefully not catch grains. I will do a brew and check it after.

Anxiously awaiting your results