• Please visit and share your knowledge at our sister communities:
  • If you have not, please join our official Homebrewing Facebook Group!

    Homebrewing Facebook Group

Home made peristaltic pumps

Homebrew Talk

Help Support Homebrew Talk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
One of the limiting factors in getting a HERMS system to ramp quickly is the rate of flow.
Assuming grain bed compaction doesn't replace the flow rate as the limiter, this peristaltic pump could be a great solution to quick ramps using HERMS and RIMS.
In my system, my March manages about 8 litres / 2 US Gall per minute.
I would like 3-4 gpm. from a peristaltic pump to make it a significant change from the March for the above purpose.
The priming advantages aside of course.
So Silverbrewer have you arrived at any design specs we can pass to the CAD guys?
Don't mean to rush you....

Cheers.
 
Yorg....the mk1 has 1/2" pipe on a wall that is 160 mm radius. The construction leaves a bit to be desired, but once primed it will run between 0.55 litres pm (0.5 us pint) at 8 rpm to the max I have run it at so far that was of the order of 120 rpm and 4 usgpm. I have just had it pump vertically up a 15 ft length of 1/2" stainless tube with no problems at 105 rpm, so that is quite a big flow range.

My main concerns at the moment are how it will perform when asked to pump hot liquids, (as that is what it is going to be dealing with, even though it can be placed after the cooler during boiler to fermenter operations, it will still see mash temps in my rims) and how the tube will stand up to the rollers giving it a constant hammering....

I cannot really test the tube longevity in the mk1 setup, as the MDF crap it is made from gives off a lot of wood and glue dust that settles on the outside of the pipe, so it would not be a fair test until it is made from something nicer like nylon and is constructed in such a way as the tube and walls etc are sealed from outside contaminants.

That said, I have it set up to run at a constant 46 rpm, (that will give 4.6 litres per min.) and over the next week I will be running it all the time I am at home, to give it the nearest I can to a continuous test. So far it has run for a couple of hours at that speed setting.

When I get a free weekend, I will set up a similar test, but at high temperatures.

The Mk 2 is taking shape, and might be a single wheel, running in a 3" length of 1/4" wall thickness alloy tube of about half the size of the Mk 1. I saw a whole bunch of alloy tube remnants when I went and bought a length of stainless 1/2" tube last Friday, so I will be buying a bit of something suitable as soon as I can. I have bought three conveyor wheels from RS, item no 2352224 ...... that has good solid ball races and is 2" wide, 1 7/8" dia and runs on a 15 mm shaft. http://uk.rs-online.com/web/search/...od=searchProducts&searchTerm=2352224&x=15&y=8

Having my lathe run the pump is a brilliant idea.....I have fitted a potentiometer into the lathes existing drive electronics that gives zero to max revs in whatever gear setting I put the lathe in. From outside the shed, on a brewing day, the potentiometer is a pump control knob, and from inside it gives the lathe infinite speed control when used as a boring old lathe (Pun intended)

I have taken pictures and a video, and will post them when I get a bit of time and also work out how it is done!!
 
I get the impression that running mk1 at something like 100rpm wasn't giving you the impression that this would be a normal running speed. Would that be because of the MDF construction, or inherent in the design?.
Is mk2 an improvement for same purpose as mk1 or is this version for something else altogether - I think you had mentioned something about a spererate pump for sparging?

Since I don't, and i suspect most others don't have a lathe hanging about, what power/torque specification would a motor have to be? I thinnk you mentioned something about a wiper motor - are you thinking it would do the job.
Would a wiper motor be speed controllable via pulse width modulation do you think?
Cheers.
 
How about picking taking a piece of 6" schedule 80 pvc pipe and cutting off a 1" slice of it. then you could sandwich that between two pieces of mdf, steel, etc. I can't figure out how to make the roller assembly without access to a machine shop.
 
These pumps may well work fine at 100 RPM, or even faster? Now that I have shown that they may have the capability, it needs someone to make a proper one, then use it for a year or so before we can be absolutely sure. (I am going as fast as I can) The same goes for the sizes of motor required. Using wiper motors would enable everyone to have a cheap power source, but that may be wishful thinking. Reducing the speed of an electric motor electrically reduces it's torque proportionally (I think) so it is much better to get the speed down to whatever you need using reduction gearing of some sort, be it chains, belts or gears. The good news is that if you reduce the speed of a motor mechanically you INCREASE the torque proportionately! If the dérailleur gear sets from push bikes can be made to work back wards, that would do it and give you a selection of gears to choose from! Some sort of simple chain tensioner should keep the chain from falling off.....Pulse width modulation would do it as well (but with the reduced torque problem), so get a sizable scrap motor from somewhere and then gear it down, then speed control it for fine adjustment. It most likely won't be pretty, but pretty costs $1800! so settle for ugly.

As soon as I get my 15 mm drive shaft coupling from RS I will connect my Mk1 pump to my large motor rather than my lathe. (this motor is in the pics and is about 4 times bigger than a wiper motor, it is geared down to 120 rpm mechanically and also has electronic speed control on the 240 volts) If this works, I will then try it on a 12 volt wiper motor and see how that manages. My guess is it will get very hot, but who knows. In the meantime, I am endurance testing the MK1 at 40 rpm on the lathe. It has done 8 hours so far. (2X4hrs)

The Mk 2 is being made mostly because of the crap MDF construction of the Mk1 and the fact that the skateboard wheels I used need to be a tad wider to fully squeeze the tube completely, this lack of squeeze at the very edges is causing air to pass the rollers slightly and so it needs a few revs before it primes. Fully squeezed, it will prime at any speed.
Sjlammer...please note that the 1/2" tube needs rollers of at least 27 mm wide, so a bit over 1"

I don't see the need for such a high flow rate as 4 USG PM, but feel free to enlighten me if I am wrong...I was going to make two pumps using 1/2" silicone tube and a bit smaller wall radius, say 100 mm, using 1/4" thick walled aluminum tube, or some of the large plastic tubes used for water mains under the street! Each pump will be made wide enough to let the 1/2" tube overlap, so it then only needs a single wheel, but that wheel needs to be at least 60 mm wide. So that will probably be two bits of pipe about 6" long. Each bit of pipe will have to have an arm to hold the wheel and two solid end plates to hold the bearings and seal out any dirt, unlike my Mk1

The plan would then be to use one pump to put hot liquor into the mash tun, say 60 rpm to get about 6 litres per min) one pump to circulate the mash liquor during RIMS, say 20 rpm to get about 2 litres per min, both pumps to sparge, at say 5 rpm to get 0.5 litres per min. Then either one or both pumps to circulate the boiled wort through the plate cooler and then back to the boiler to swirl the cold break at 60 rpm and 6 or 12 litres per min, and then the same to transfer the cold wort to the fermenter.

My biggest fear is that the pumps have the ability to suck at a stuck mash until it solidifies into concrete, so hopefully the rims process will settle the mash out to a nice filter! March pumps will just sit there waiting for liquid in the same situation....anyone wish to comment?

Do the flow rates above make sense? If you want higher flows to get HERMS fired up fast then just use two or three pumps on the same shaft, or two or three tubes in the same pump...just make everything wider and use a strong motor!
 
Fantastic news guys....The Mk 1 pump turns just fine with a windshield wiper motor.

The thing is out of a Range Rover Discovery, and has two speed settings which under no load give 60 and 40 RPM. driving the pump on the faster setting the pump slows down to 37 RPM and pumps 3.55 litres per min, which is 0.937 US Gallons.

The motor is going to get mighty hot, but hey, you've just got to make a small water jacket for the motor case and either cool it with tap water or knock up another tiny peristaltic pump to flow some water through it !!!! I will run it until my leisure battery gets low, or the motor melts, whichever happens first.

Houston, we have cheap pumps......
 
I'm currently working on drawing up a model in solidworks, after it's done I can export it to other formats for whatever cad software people may have.

Soon as I have it finished and find some material I am going to cut one out on the cnc and see how it works.
 
Yea I'm dead chuffed with it! I have had the wiper motor running for a hour and it does indeed get hot! Smokin would be a better term to use, so anyone going down this route had better (read must) source a wiper motor with a steel motor body rather than an aluminum one so a copper water jacket can easily be soft soldered on, and set to work.

To connect it to my 15 mm dia pump drive shaft, I made up a 2" length of 15 mm dia steel and drilled and tapped a thread concentrically in one end. This just screwed onto the motors output shaft until it hit the start of the splined cone that seems to be universal on these things. I then clamped the two 15 mm shafts together using a proper shaft coupling with decent clamping screws. Jubilee clips and a bit of old hose pipe will not stand the torque!

An hour long test at that temperature means that the motor is basically up to the job and we just need to stop them dying by getting the temperature down to reasonable levels. From previous work, water cooling the motor bodies of electric motors "in distress" I can say that water cooling is mighty effective and would expect the motor to not get past luke warm if a water supply is rigged up.

Yeaaaaaaaaaaaa!
 
So are some photos of my pump design. Now these are based on Jonathans pump design in the other thread. I really liked the look of the pump and how compact it was.

This is designed for 1/2" tubing.

Here you can see the main case, you would just cut two of these out and then sandwich them together.

maincase.jpg


These are the covers, again you would cut out two of them. I'm planning on cutting them out of 1/4" acrylic

cover.jpg


And here is the beginning of the assembly, I still need to design the rotor and figure out how to mount bearings.

pump.jpg



If anyone wants a copy of the drawings please let me know, or I will be posting a download link once I have the rotor finished.
 
It's been worrying me that during sparging, with two of these pumps working in unison, the mash could get sucked onto the false bottom, as these pumps won't "cry uncle" like the March pumps would.

It has just dawned on me that if a flow loop is provided so the flow can get back to where it came from if there is a restriction then the flow can be controlled with a couple of taps, or even a couple of clamps to just squeeze the silicone pipe!
 
So what's the thinking on hose compression:
If I used jdcillin's design, I would still have to design the roller assembly ( it would be great to have a CAD of that to discuss) and I would have to decide the measurement that would result in the hose being compressed by the wheels. So would I compress the hose so that it just closes off and use that measurement as the width/thickness of the compressed hose, or would I give it a good press and use this smaller measurement as the thickness?
 
Compress the hose just enough to give it an air tight seal. (check by blowing down it) Any more than that will only cause unnecessary strain on the entire assembly. I think I got my gap right, but my wheels were about a 1/16" too thin so there is a slight air leak and I cannot do anything about that until MK 2 The air leak past the wheels causes a reluctance to prime in air unless the revs are increased. Once it is pumping water, everything is fine. If a single wheel system is employed then you only have one wheel to position correctly, so you could even include a bit of adjust ability in the system. Note that with a single wheel, the tube must overlap itself before it leaves the circular wall and the wheel width must be twice as wide to accommodate two squeezed tube widths.
 
Peristaltic pumps are notorious power consumers. If the wiper motor is getting that hot a cooling jacket will have little effect. It is the windings that are producing the heat and they are insulated with lacquer as they are wound. This thin layer will burn off under high heat causing the motor to short. Cooling the case will do little to reduce the winding temp.
SB, try wrapping the rollers with a couple layers of duct tape to increase the diameter and create a positive seal. I know its ghetto but they don't call it 100MPH tape for nothing, its tough stuff.
Many peristaltics use spring loaded rollers to account for minor machining and assembly descrepencies and reduce the chance of blowing tubes from over collapse or pressure spikes while still keeping a positive seal.
 
Those seem to be sound observations and suggestion Beerthirty.
Any suggestions/ scanned hand sketches on a design for spring mounted rollers - one for a CAD talented brewer to draw up for the rest of us to critique?
(Open source peristalitic pump design :) )
 
When designing the Mk 1 pump I did indeed consider making the radius that the wheel was set at adjustable. The initial thought was to make the adjuster so wheels clamped tight in various positions, like the adjustable chain tension on an MX bike..... but then having it spring loaded crossed my mind. I was thinking along the lines of a trailing arm pivoted where my wheel is now and about two inches long. The wheel is then at the end of this arm and an adjustable link, or a spring loaded link goes across to the main arm from the wheel end of the trailing link. I ruled out these options as un-necessarily complicated, and un-needed provided the pump was made accurately. The air problem with my pump is because the wheels are not wide enough. It is not diameter, concentricity or gap related.

My Mk 2 pump is going to be made out of 1/4" (or thicker) walled alloy tube. Two alloy, or clear acrylic end plates are going to be turned up to fit perfectly into each end of a 4 inch (or longer) length of this tube and give support for two sturdy sealed ball races with a ID to suit my 15 mm drive shaft. This way concentricity should be held within a few thousandths of an inch, and the tube will be in a sealed cavity, so it can be lubricated with something, perhaps food grade white grease? I will have only 1 wheel to reduce wear on the tube.

I fully agree that using a windscreen wiper motor to run this pump is pushing your luck. As soon as a properly engineered Mk 2 pump is in existance, I will run the thing with a watercooled wiper motor 24/7 just to see what reliability we can expect from such a cheaply avaliable motor. Mine ran fine on a small 4amp battery charger

With all these things sorted, it could still go pear shaped once we try to pump hot liquids for any length of time, but I am crossing my fingers and hoping that empirical testing proves all our worries are un-founded :eek:
 
I've been trying to draw up a design for different spring loaded versions and from what i've been doing I think it makes it all a bit complicated.

Wouldn't it better just to have different sized rollers that you could put onto the shafts that way you could make adjustments just by changing the size of the roller.
 
Here is an idea you may want to play with. wheel carriage made like a caster(trailing axle) and instead of a single post use 2 hollow tubes that would slide over posts on the end of the rotor arm and install the springs in the tubes to act against the end of the posts. As long as the clearance between the post and tube was minimal it should slide easily without binding. If going with a multi roller where the tubing doesn't overlap, guides could go straight off the arm to help keep the tubing aligned just before the trailing caster so multiple tubes could be laid in the same housing and run by the same roller.
 
Interesting thought, I am going to give that some thought.

Today I plan on cutting out the housing just to see how it comes out.
 
Good luck, hope it comes out alright! A good supplier of small bearings in the US is boca bearings. Check out their website, the 99 cents page is exactly that ! Every bearing is 99 cents.
 
By the way, how loud are these pumps, compared to a March?
 
I have not heard a March pump, but it is revving fast so I would say a peristaltic would be much quieter.

JCD, I just saw your pump design. You should be able to fit the bearings into the side plates if you use thin bearings, or make the plates a bit thicker. You will need at least two rollers at 180 degrees separation as you have not overlapped the tube on a 360 degree wall ( IE 380 degrees of tube before the tube starts to leave). Three wheels (at 120 degrees) would be better to ensure it primes easily, as with two wheels you are bound to get a bit of back flow as one wheel arrives and the other leaves your 180 degree "wall". Looking good though, when will it be getting made? Did you see my link to wheels that already include two bearings? You must be able to find similar stuff over there....

I have just made a water jacket for my tiny windscreen wiper motor from a dog food tin!! I have tried it out and there are no leaks, so I will be endurance testing it soon. I am aware that the armature and windings will still be getting a hard time, but a lot of heat will be getting removed as the pump circulates water through the jacket and back to the large bucket. The motor was free, so I have no qualms about testing it until it fails! This test will just be for fun as I would say that these small geared motors are not really up to the job. If anyone is serious about going the peristaltic route they must source a decent size geared motor, or come up with a gear reduction system to suit an ordinary motor. Shop about, as there are hundreds of redundant motors getting thrown out all the time.....
My idea for getting home made gear reduction would be to knock up a couple of sturdy plates like the side walls found on a "monster mill" have three shafts set at suitable spacings to accommodate push bike pedal sprockets and rear gear sets. Arrange the drive of the motor to drive the smallest rear gear on shaft 1 with a chain to to a big front sprocket on shaft 2. Fit another small rear gear on shaft 2 and drive shaft 3 with a chain to another big front sprocket mounted on shaft 3 etc etc. chains can dangle down with a weight hung on one of the tiny sprockets from the dérailleur mechanism to get a chain tensioner If the sprockets can be clamped to line up with a selection of gears then various ratios will be possible. You just get whatever sprockets you want lined up, drop a chain onto them and off you go.....Then your large motor, which in my case is 150 rpm flat out, can be set to give virtually any lower speed and therefore any lower flow rate, without reducing the speed electrically, and loosing torque.
 
That was my plan, I am going to cut almost all the way through the plates and then press fit the bearings into place. I think they will probably last just fine.

Yeah on my rotor design i'm going with a 3 wheeled design just to keep it simple. How much clearance did you leave between the wheel and the wall?

I am waiting for material but I have programmed it up to cut out on the cnc and i'm tempted to do a version out of wood just to see how it works. But the final version will be made out of something better like derlin or nylon.
 
I would think the clearance would be twice the wall thickness of the chosen tubing minus a couple thousands. (this is where the spring loading comes in, different tubings and elasticity of the tubing.) Of course the elasticity would play a roll in the spring rate.
 
I would think the clearance would be twice the wall thickness of the chosen tubing minus a couple thousands. (this is where the spring loading comes in, different tubings and elasticity of the tubing.) Of course the elasticity would play a roll in the spring rate.

A couple thousands of an inch would be within the tubing wall differences of the tubing production runs besides different brands of tubing selected.
I was thinking .010" to .015" preload against the tubing as the two walls of the tubing will compress half that amount per wall thickness allowing for any wear thus preventing any future backflow leaking problems as time is put on this pump.
 
For my 2 cents worth, I think .010" to .015" extra compression is a bit too much. I think that as soon as the wheel squeezes the tube enough to prime at near zero revs (IE by hand) and stay primed if you stop rotating it, then that is enough. Any more pressure will just make the pump harder to turn.

Remember that from day 1 if needs be, or after a bit of wear has set in, more squeeze can easily be had by fitting a bit of flexible shim stock behind the tube, for the entire width and length of the wall. In fact, any adjustment required can be done this way, as long as the wheel is made wide enough in the first place, unlike mine, which was limited from the outset by a bad wheel purchase (£10 skateboard)
 
Back
Top