PWM pump control

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MrNatural

The Original, beware of imposters.
Joined
Dec 31, 2010
Messages
531
Reaction score
91
Location
Elverta
Finally getting around to arranging (and rearranging) guts of my control panel and have come to a fork in the road.

Have room on relay board for pumps, and also have a couple of spare SSRs. Want fly sparge ability and wonder if there is any possible use of PWM for flow control of pump reacting to MLT level switch? Or during mash recirculation?

I realize I likely have a couple more layout revisions ahead of me, but want to get components figured out first.

Single tier, 3 keggle, two pump, BCS 462.

Experiences, theories or opinions welcome.
 
If your pump is brushless, like most are, PWM is a bit hit and miss. In my experience with a little tan pump PWM around 1 kHz does work but makes a horrible noise. At higher frequencies (like 20 kHz) it doesn't work.
 
I put a sanitary diaphragm valve in to throttle my flow from the mash tun to the kettle during sparging. It works much better than a ball valve. Lots of breweries will use a Variable frequency drives to modulate motor speeds, but these work with 3-phase motors.
IMG_2980.jpg
 
I put a sanitary diaphragm valve in to throttle my flow from the mash tun to the kettle during sparging. It works much better than a ball valve. Lots of breweries will use a Variable frequency drives to modulate motor speeds, but these work with 3-phase motors.
View attachment 176615

Looks like that conical is going to a good home....:mug:
 
I put a sanitary diaphragm valve in to throttle my flow from the mash tun to the kettle during sparging. It works much better than a ball valve.
Interesting, where can you find them? Guessing they handle hot liquids fine, what material is the diaphragm?

Are these AC or DC pumps

Pumps are 110V March 809.
Sensor on MLT is http://www.mouser.com/Search/ProductDetail.aspx?qs=TwPrcXezm73yZ6LcOrsVFg==
e2k-lf.jpg


Setting it up for auto-fill and got thinking about using it to trigger HLT output during sparge.

I was going to get solenoid for output with bypass back to HLT for de-stratification, but believe it or not, I'm already over budget :fro:

Kind of funny. Spent (or should I say $pent) last six months trying to go from "zero to beero" building stand, keggles, auto gas or electric HERMS, CFC, SS fittings, sanke fermentor, 14 cornies and keezer, stir plate, yada yada - and now I'm getting shriveldick over a solenoid :eek:

Dale
 
I'm going to be working on constructing a pwm using only integrated circuits from Texas instruments specifically for March or chugger pumps this semester. I'll keep you posted

Sent from my DROID RAZR HD using Home Brew mobile app
 
I got mine on ebay for ~$40, I've seen them that cheap recently but they range from 30-200 on ebay. Search ITT or Saunders Diaphragm valve. The Valve is all sanitary 316 stainless steel with a high temp silicon valve seat. I coupled it with a modified autosparge and my sparge now is close to set it and forget it till the kettle is full.

I think automating the sparge/transfer step is definitely the most challenging thing to automate without unlimited funds, and even more challenging to do it with electronic controls. I went the analog route, but I think others here do have fully automated setups, so it can be done for sure. I'll Like to see what you come up with.

IMG_2978.jpg

IMG_2939.jpg

IMG_2940.jpg
 
I got mine on ebay for ~$40, I've seen them that cheap recently but they range from 30-200 on ebay. Search ITT or Saunders Diaphragm valve.
Thanks, will check that out.
I coupled it with a modified autosparge and my sparge now is close to set it and forget it till the kettle is full.
That is my ultimate goal. If I can get the sensor and BCS to make nice, I can set another sensor on HLT to trigger after calculated sparge volume has been delivered :rockin:
I'll Like to see what you come up with.
That makes two of us................:)
 
Just FYI, I have used a SCR "router speed control" module to reduce the power of a PSC pump motor. It seems to work ok - my concerns with it are overheating due to slowing down the motor fan and if the chopper circuit damages the pump electrics somehow. From what I read it is a normal way to control the power to the pump. The other issue is that this only controls the power - so if more head gets put on the pump it will slow down even more, I have seen some advanced chopper ciruit designs that somehow utilise the back EMF to control the power/speed independent of the load but can't find anything off the shelf that uses this method.
 
Funny thing. I'm currently going through the same process right now.

I ended up getting a 12 volt 809 March pump that I intend to regulate with a pwm circuit run off a pot and a 555 timer. The intent of this is for mash recirculation immediately.

Ultimately I want to create an auto sparge utilizing the same pump method, and balancing flow using flow meters on the input and output of the mash tank. Though I think I will be using a micro controller for that.

Here is a PWM Control Example which you could build for about 15 to 20 bucks if you were interested.
 
I control my flow with a valve on the outlet of my MARCH pump & the pump runs full speed al the time.
 
If anyone is still interested, there exist driver boards for single phase ac motors. However, when looking at the cost of buying a march or chugger ac plus the control hardware, its far more expensive than buying the 12 or 24 V DC march pump which can be had for about 190 for the polysulphone (sp?) Head. The DC motors actually put out a little more power than the ac counter parts. Anyway ask questions if you have any.

Sent from my DROID RAZR HD using Home Brew mobile app
 
If anyone is still interested, there exist driver boards for single phase ac motors. However, when looking at the cost of buying a march or chugger ac plus the control hardware, its far more expensive than buying the 12 or 24 V DC march pump which can be had for about 190 for the polysulphone (sp?) Head. The DC motors actually put out a little more power than the ac counter parts. Anyway ask questions if you have any.

Sent from my DROID RAZR HD using Home Brew mobile app

Can you post a link to the driver boards?
Cheers
 
Maybe I am confused on this but what is the point of all of this? I get what PWM is and the idea of pulsing the power to slow a motor. Heck I use 12V DC Motor Controllers on my motorcycle to get a PWM for my heated grips.

Where I am confused on this is the point of it with a March (Chugger, ect) pump. Being they are magnetically driven why not use a ball valve to restrict the output flow from the pump. This is what most people do from my understanding. The automatic method could be something like the Blichman Autospage that was posted in an earlier post.

So is the goal better automation, to save wear and tear on the pump (don't see that as an issue as the impeller isn't direct drive) or some other reason?

Just trying to understand it. I am in the middle of a BCS build and have SSRs controlling my pumps (March 815).
 
Maybe I am confused on this but what is the point of all of this? I get what PWM is and the idea of pulsing the power to slow a motor. Heck I use 12V DC Motor Controllers on my motorcycle to get a PWM for my heated grips.

Where I am confused on this is the point of it with a March (Chugger, ect) pump. Being they are magnetically driven why not use a ball valve to restrict the output flow from the pump. This is what most people do from my understanding. The automatic method could be something like the Blichman Autospage that was posted in an earlier post.

So is the goal better automation, to save wear and tear on the pump (don't see that as an issue as the impeller isn't direct drive) or some other reason?

Just trying to understand it. I am in the middle of a BCS build and have SSRs controlling my pumps (March 815).

From a fully automated standpoint, it's easier to control flow rate by directly applying PWM to the pump rather than pulsing an electronic ball valve. Valves are more prone to failure than pumps. A 12V DC motor is obviously more simple to control than an 120V AC motor, because you don't have to reduce the voltage to increase slip or slow the frequency of the applied voltage.

The end goal, at least for my senior design, is to have a system where I can adjust flow rates during sparging and mashing to maintain a constant water level above my grain bed, and to also maintain mash temperature. I can easily do this by controlling my pump speed, but there is no good way to restrict the ball valve because they don't have reliable feedback control.
 
From a fully automated standpoint, it's easier to control flow rate by directly applying PWM to the pump rather than pulsing an electronic ball valve. Valves are more prone to failure than pumps. A 12V DC motor is obviously more simple to control than an 120V AC motor, because you don't have to reduce the voltage to increase slip or slow the frequency of the applied voltage.

The end goal, at least for my senior design, is to have a system where I can adjust flow rates during sparging and mashing to maintain a constant water level above my grain bed, and to also maintain mash temperature. I can easily do this by controlling my pump speed, but there is no good way to restrict the ball valve because they don't have reliable feedback control.

Ok, that makes sense but why not use a manual ball valve where the wort enters the MT? I once you work out the amount of restriction needed, it would be fairly constant. Though the issue may be that from one batch to the next the needed restriction may be different. Even if brewing the same beer recipe the grain bed may set up differently to where it flows faster or slower between batches. This is where something like the Blichman auto sparge might work well. It would take some setup but once done it should maintain the proper volume of wort above the grain bed.

Like I said, just trying to understand the why behind what people on this thread are trying to do.
 
Ok, that makes sense but why not use a manual ball valve where the wort enters the MT? I once you work out the amount of restriction needed, it would be fairly constant. Though the issue may be that from one batch to the next the needed restriction may be different. Even if brewing the same beer recipe the grain bed may set up differently to where it flows faster or slower between batches. This is where something like the Blichman auto sparge might work well. It would take some setup but once done it should maintain the proper volume of wort above the grain bed.

Like I said, just trying to understand the why behind what people on this thread are trying to do.

For most systems that would probably work. I'm trying to design a clean in place system though, so having the ability to open the valve all the way when transferring 160 degree water mixed with oxiclean or pbw to recirculate is a huge benefit. Another reason could be that I want to recirculate slower or quicker than how fast I lauter

Sent from my DROID RAZR HD using Home Brew mobile app
 
...A 12V DC motor is obviously more simple to control than an 120V AC motor, because you don't have to reduce the voltage to increase slip or slow the frequency of the applied voltage.
...
Doing some extra reading into the 120/240VAC single phase pump control, and it looks like the main method of controlling PSC/Shade pole motors (the typs on march/chugger/etc.) pump is using a Triac and Phase angle control. So that gives me much more confidence that using the router control isn't going to damage my pump (as long as it is running fast enough for the fan to cool it effectively :))
Thinking about how to implement that into a control by a uC; I either have an addresable (1-wire if they still make them) digital Pot to replace the manual pot on a SCR control board or implement PWM of a standard SSR. The second route seems from my rough thoughts to only give you an finite number of steps based on f_mains / f_PWM (since you want to use a zero crossing SSR that will only switch off why V=0) - so at a PWM frequency of 5Hz you would have 10% steps - going any lower/higher with the PWM frequency you would large expect surges in the pump speed or the # steps would not be enogh to control well enough. To get going, since I have all the parts I need for method, I will go The #2 route to prove it could work then I will look at the digital pot/SCR circuit method.
I agree a fair bit more complicated than just PWM the DC feed to the pump, but in reality not much more :D By the way feedback will be provided by a flowmeter for my circuit.
Ok, that makes sense but why not use a manual ball valve where the wort enters the MT? I once you work out the amount of restriction needed, it would be fairly constant. Though the issue may be that from one batch to the next the needed restriction may be different. Even if brewing the same beer recipe the grain bed may set up differently to where it flows faster or slower between batches. This is where something like the Blichman auto sparge might work well. It would take some setup but once done it should maintain the proper volume of wort above the grain bed.

Like I said, just trying to understand the why behind what people on this thread are trying to do.

Just like any question in life - sometimes the answer is just because you can :D.
I guess it comes down to where you are using it - the blichman product works good at keeping a set volume above the grainbed of the MLT but it will only work in that cirumsatnce - whereas the pump control works to give you the flow you want but you'd need extra sensors to set exact GPM / keep a level above the grainbed / etc.
Since more an more advanced automated homebrewers have started to appear I would say no-one has really address how to adjust flowrate - it is just a matter of full bore or nothing. Secondaly there were concerns about it not being great to continously pump the duration of the mash as proteins get damaged - I would expect less damage from a pump running slower that one going full bore and throttled at the valve.
But I for one just want to do it because no one else seems to have completed a full feedback AC pump control - can't find any YouTube videos, blogs, etc. on it!
 
Thanks for the info.

Makes sense. A matter of what level of automation one is looking for.
 
Thanks for the info.

Makes sense. A matter of what level of automation one is looking for.

Pretty much. And some people just can leave well enough alone :D (me included)

As for the other part of my above post - I found some cheap non-zero crossing SSR for around US$4. Big enough to handle pump motor currents. Going with one of these will hopefully get me 0-100% control digitally.
 
Pretty much. And some people just can leave well enough alone :D (me included)

As for the other part of my above post - I found some cheap non-zero crossing SSR for around US$4. Big enough to handle pump motor currents. Going with one of these will hopefully get me 0-100% control digitally.

What pump are you using?

Sent from my DROID RAZR HD using Home Brew mobile app
 
What pump are you using?

Sent from my DROID RAZR HD using Home Brew mobile app

I have a Resun (Iwaki knock-off!) MD-40. It is 80W and puts out a bit more than the standard march/chugger. When I bought it I purchased the MD-30 which is more comparable to the norm, but the Chinesse seller was "out-of-stock" even though the website listed more than 20 in stock. managed to get "upsized" to the bigger pump for no extra cost.
 
I have a Resun (Iwaki knock-off!) MD-40. It is 80W and puts out a bit more than the standard march/chugger. When I bought it I purchased the MD-30 which is more comparable to the norm, but the Chinesse seller was "out-of-stock" even though the website listed more than 20 in stock. managed to get "upsized" to the bigger pump for no extra cost.
I can't seem to find temperature ratings for either of them?
 
I can't seem to find temperature ratings for either of them?

From memory the temp rating is 80°C - but I have had the thing apart and check all the materials. The housing is Glass Fibre Reinforced Polypropylene (GFRPP) which is good for above 100°C as well as al the other materials used. I also asked a compititor about the pumps rated for under 100°C and his response is the ratings are what the pump is guarenteed to run at for it's design life - and for homebrewing this was way over the top, i.e. it is fine.
The little giant pumps that used to be popular but now don't seem to be (I think because of Chugger) were also not rated for 100°C but plenty of people used them for years without issues.
iwaki manual here http://iwaki-pumps.com.au/manuals/MD instruction manual.pdf
By the way I wouldn't recomend this pump over a March/Chugger - I went with it because it was the best value for money for a pump landed in NZ. If I was lving in the states I would be going the Chugger SS route.
 
By the way I wouldn't recomend this pump over a March/Chugger - I went with it because it was the best value for money for a pump landed in NZ. If I was lving in the states I would be going the Chugger SS route.

Gotcha, thanks.
 
I have an automated fly sparge on my system. Its brew troller based and I wrote the code for brew troller to support it.

Sadly the fastest PWM output I could get with the brew troller is 10hz because they didnt hook up the hardware counter timer output lines to the PWM control pins on the board, so I had to use an ISR generated by the counter timer to bit bang the GPIO's. Thus I took a pot controlled PWM controller off ebay and made a really low low pass filter and tied its output to the center pin of the pot, with no power applied to the pot. This then let me control the voltage just like a pot would but using PWM, which it then gives an 8khz PWM signal to a DC motor that controls a peristaltic pump head.

I first thought about PWM controlling the AC motor on my little giant pumps but realized that the RPM to PWM curve is non linear. Further the flow rate to RPM curve on centrifugal pump heads is VERY non linear. So in the control system you just compounded two non linear curves... that's bad news. DC is much more advantageous in that the RPM to PWM curve is almost perfectly linear, but your flow to RPM still is not. Trying to use a PID control method on either of these options just spelled trouble.

That's why I went to a peristaltic pump head, the flow rate is 100% linear with RPM, thus making it much easier to control. I got the pump head off ebay for 25 bucks, use a 30kg/cm torque geared down 12 volt 10 amp DC motor that is only 35 bucks off a battle robot web page and then made the mount for the pump head, motor, and drive shaft my self.... the other nice thing, its self priming.

So the peristaltic pump does the pumping from the MLT to the kettle by PWM control from the brew troller. I added code to the brew troller to measure the volume difference in the kettle every 45 seconds so I know the flow rate averaged over that time. I then wrote a by hand controlled algorithm (PID still didnt work great) which changes the duty cycle by 10% if the flow rate is off by more than an amount, 5% if less than that, and 1% if less than that, so it really is just changing the proportional gain in a stepwise linear way with no integration or derivative.

Then I have code in the brew troller that once a programmable amount of liquid has been put into the kettle, it turns your pump that pumps into the MLT from the HLT on until an equal amount has left the HLT. So its a "pulsed" fly sparge on the inlet to the MLT, and that works great.

With that setup on low gravity beers (pilsner at 1.048 OG) I've hit 90% efficiency with that setup. Then again I also have a V-wire false bottom that was custom made..... I really should post picks of this whole thing sometime.

Anyways, that's my experience with automating fly sparge.
 
...I first thought about PWM controlling the AC motor on my little giant pumps but realized that the RPM to PWM curve is non linear. Further the flow rate to RPM curve on centrifugal pump heads is VERY non linear. So in the control system you just compounded two non linear curves... that's bad news. DC is much more advantageous in that the RPM to PWM curve is almost perfectly linear, but your flow to RPM still is not. Trying to use a PID control method on either of these options just spelled trouble.
...

Did you try PWM with the AC centrifugal pump or just write it off from the start?
While I agree that the RPM - Flow curve is also not linear; the power - flow is reasonably straight though, and with PWM of the AC pump you are not looking to regulate the motor speed but the pump flow. the question I guess is would the power output of the PWM circuit be linear?
 
I havent seen any data correlating pump power input to flow rate so I cant comment on that. I suspect, that given the fact that AC motors like to run at a fairly constant RPM, and that the RPM will change with power input (how much? that's a very good question and I'd like to know) I think the whole thing is just suspect. There's a reason they just go frequency drive for most applications, or DC.

I wrote it off after looking at all the AC control options I could find (not that many for single phase that I could find) but this was also nearly 3 years ago when I made this. One thing while looking that really bothered me was the lack of good information any of the options gave. You'd pretty much have to buy one of the solutions and just try it, and I wanted it to work the first time so I just threw the idea out and went with what I was much more confident would work.

If they could give more info on RPM vs power vs flow rate on the pumps, I'd have been more likely to give it a try. Given how single phase AC motors work though (how they like to run at a specific RPM for the most part)... my intuition as an EE just says "not worth trying", but I could be wrong... Also note that a true PWM control of an AC motor doesnt go well, its really more of what has already been described where they use a triac or the like to give a kind of "variable" resistance given a PWM input (over simplifying the idea, but gets the point across), thus controlling the "power" like you're talking about, but you already covered this in an earlier post.

I'll also give you a heads up on the flow meter. The cheaper ones are turbine based and give a pulsed output based on turbine speed and that gives you flow rate, but with all the crap that is in wort I had a lot of issues with them clogging, so that's why I use a change in volume over change in time method. If you're using a much more expensive flow meter (and if you are let me know where you go it! I want to try one out but not willing to spend 1000 bucks on it from what I've seen) that uses capacitance or ultrasonic then you should be fine. (though I'd worry about it being off with change in specific gravity, I already had to take that into account with my volume measurement)
 
http://www.mcmaster.com/#9663k93/=qvd4b1

They are a little hard to snip, but work great. I think if I need some more I'm going to use a dremmel or a cut off wheel to cut the springs.

I bought one of these as suggested and I have two words for you, "Lineman Pliers." They worked like a charm. These springs are made from 14 gauge SS wire, but a decent set of linemans will shear right through them with little to no effort.
 
Finally getting around to arranging (and rearranging) guts of my control panel and have come to a fork in the road.

Have room on relay board for pumps, and also have a couple of spare SSRs. Want fly sparge ability and wonder if there is any possible use of PWM for flow control of pump reacting to MLT level switch? Or during mash recirculation?

I realize I likely have a couple more layout revisions ahead of me, but want to get components figured out first.

Single tier, 3 keggle, two pump, BCS 462.

Experiences, theories or opinions welcome.

I sort of do this myself however with a sligtle less complicated method. I use a floatswitch on my spargearm to turn the pump on and off and manually turn down the flow level of the pump via pwm control so the pwm speed is not really controlled by flow but the way I do it works very well... Due to the speed of the pump being set so low at this stage the cycling on and off is minimum... I also set the pump speed pumping out of the MT at a very low rate so it drains slow. (I do use 24V dc pumps)

Just an alternative simplier method for folks out there who are looking I guess..
 
Doing some extra reading into the 120/240VAC single phase pump control, and it looks like the main method of controlling PSC/Shade pole motors (the typs on march/chugger/etc.) pump is using a Triac and Phase angle control....

I've been working on automating an old hotplate + stirrer and ran into a similar problem of how to variably control the speed of the AC stir plate motor. After some research is appears the best (and possibly only?) way to control most AC motors requires complicated frequency/voltage control. An exception are shaded poll motors, found in most fans (and strriers!). These can be controller via phase angle control, as mattd2 mentions. As I understand it, this trick only works with some load types. I'm not sure if a pump is similar enough to a fan to drive it correctly.

(since you want to use a zero crossing SSR that will only switch off why V=0) - so at a PWM frequency of 5Hz you would have 10% steps - going any lower/higher with the PWM frequency you would large expect surges in the pump speed or the # steps would not be enogh to control well enough....

I'm not sure burst fire control will work for a pump, it is different than phase angle control. Burst fire uses bursts of full AC sine waves, always turning on/off at the zero crossing. Phase angle control turns on the AC some time after the zero cross and off again at the next zero cross. Phase angle provides much finer control, but causes more interference (such as the whine you hear when using a cheap light dimmer). Typically phase angle control is used when a high frequency of pulsing is required, like lighting, in which we would see flicker if burst fire was used. Burst fire is used when high frequency is not required and causes less AC interference since it turns on and off at the zero cross, for example when controlling a heating element.

A zero cross SSR will not work with phase angle control, it has to be random fire (i.e. will turn on anywhere, not just at the zero cross).

I suspect if you try to control an AC pump using burst fire it will be to low a frequency to control the speed. It would probably cause the pump to either be fully on, or fully off, not some intermediate value.

I have no idea if phase angle control would work for a pump, and motors are trickier to control via triacs than lights and heating elements (inductive vs resistive loads). What you could do is buy a cheap light dimmer and see how that works, they use phase angle control. If it works, implementing phase angle control in a microcontroller is slightly complicated by the need to detect the zero cross of the AC wave, you can look up microcontrolled light dimmers for the general idea.

These problems are probably why we do not see many people controlling AC motors, since true AC speed controllers are expensive and Triac based methods are only feasible in a few situations.
 
I got mine on ebay for ~$40, I've seen them that cheap recently but they range from 30-200 on ebay. Search ITT or Saunders Diaphragm valve. The Valve is all sanitary 316 stainless steel with a high temp silicon valve seat. I coupled it with a modified autosparge and my sparge now is close to set it and forget it till the kettle is full.

I think automating the sparge/transfer step is definitely the most challenging thing to automate without unlimited funds, and even more challenging to do it with electronic controls. I went the analog route, but I think others here do have fully automated setups, so it can be done for sure. I'll Like to see what you come up with.

View attachment 176667

View attachment 176668

View attachment 176669

Bsquared, I'm interested in your set up for sparging. I am using an autosparge from Blichmann and interested how you've modified it, etc., to come close to "set it and forget it." Any photos, and the like, would be awesome.

Thanks, Cioffi
 
Back
Top