Last time I outlined a number of changes that would improve the working of HERMAN 7. Changing the valve configuration and building an immersion chiller coupled with a busy work week meant that it was a push to be ready for brew day.

Despite this it was a pretty successful day, and definitely some huge steps forward. I forgot a spring while re-assembling one of the valves so I couldn't use the 'liquor in' line at all. It was inconvenient (especially after the work that went into it) but the workaround was simple enough. I'm glad that the system is modular and the camlock fittings work a treat.

The immersion chiller worked really well. It is nice to have proper control over the chilling process after battling for so long with an inadequately sized plate chiller. The immersion chiller takes away sanitisation anxiety, especially compared to the plate chiller. It also means that cleanup is much simpler because I don't need to rinse/PBW/hot water flush the chiller like I needed to before. The immersion chiller is aided by a whirlpool paddle on a motor that agitates the wort and whirlpools it.

The plate chiller has been re-plumbed as a pre-chiller. I didn't use it today but it is a simple matter to pump ice water through it to lower the ground water temperature. It seems to be a much better use of that piece of gear - no sanitation questions and still helping the chilling process.

I realised while brewing that I no longer had a need to recirculate liquor from the kettle drain via the chiller and back into the kettle. As the kettle is higher than my fermenters, I can gravity feed from the kettle once the wort is chilled. This eliminates the pump from the system and makes it far easier to sanitise. As an added bonus, it means one less solenoid valve is needed. This is handy because I needed an extra for the 'water in' modifications.

I don't yet like the way the BT is coded for running an electric kettle. I will speak to Matt about this on the BT forum but may end up doing some custom code to suit my system.

I still had one switching event that upset the controller during this brew. Last time I had at least half a dozen induced resets so again a significant improvement. The difference was having all three tuns run with shielded cable compared to only the HLT for the first brew.

Overall the brew went much better than last time. I started it later, had more time to do other things while brewing and had a lot of fun brewing with the BT. Next week it will be time to make another lager, and there are not too many things that need to be modified. The spring is already back in the solenoid and the only urgent thing is to hook up the ice water pump which is easy.

The laptop to IDE hard drive adaptor arrived today and to my relief the old laptop drive fired up without trouble. I now have a backup on my backup and the HERMAN project is good to resume again after only a minor hitch!

Next on the software side of things:

• implement checksums to make comms between picaxe and pc more robust
• check why the picaxe clock is about 25% slow
• wait for RevEd to release their TCP/IP virtual serial port and figure out ways to implement such a groovy idea into the HERMAN project

And on the hardware side of things:

• finish the work on the stainless steel fermenter from an old converted LPG tank

With the laptop dying unceremoniously, I've responded quickly and am now the owner of a refurbished X40 thinkpad. It is so small and light .... It makes the old machine look like a brick. At a mere 1.22 kg (43 ounces), I love it.

If I was superstitious, I would have checked the calendar today because it seems like it should be Friday 13th. It has been quite a day of bad luck - at least when it comes to technology I normally take for granted.

The day began well. We had planned to brew a Belgian Wit, one of our favourite styles and one that has been elusive to get right.

The first news of the day was that the automatic startup system actually worked. HERMAN is designed to run on its own with a picaxe 28x1 microcontroller. I have coded a companion VB program that adds some extra functions such as graphing data and for the first time today, performing an automatic startup at 5am.

I awoke to the brew machine just about ready to go - success, the VB code had done its bit, and remote automation had taken a step forward. The odd thing was that the laptop had frozen. Thinking it was a Windows crash, the brew day continued...

At this point I was glad the picaxe code had been substantially upgraded since we last brewed. The LCD now shows good information which means no more running blind. The new kettle power code also eliminated surging during the boil. At last the system was working as planned - free running without a laptop. For the previous 5 incarnations of HERMAN I'd been a victim of the dreaded blue windows screen of death.

When I had some spare time, I looked a more closely at the laptop. I began to get that sinking feeling that it will never boot up again. Nothing I tried would get the machine to fire up. As it dawned on me that I've been lousy doing backups, I did some quick ebay work for an adaptor to check the laptop drive my main machine.

Why didn't I do backups? My work computer gets a backup weekly. My recreational laptop on the other hand ...

In the last few weeks I've done a lot of coding in both VB and picaxe. The most recent backup of the picaxe code was uploaded to SkyDrive on 5th February. I think I can remember most of what I've done since then - but it is damn annoying. The last backup on the VB program was late September last year! Ouch!

If these things weren't bad enough, there was more to rub salt into the wounds. As the brewday progressed, we realised that all the recipe details were on the laptop - I'd not done a printout! The most recent recipes on my main machine were 2003. OK, so I could remember how I put the recipe together for the wit and recreated it on the main PC. It then crashed in a really odd way and wiped clean the promash recipe I'd just created for the second time. I didn't have the heart to try a third time.

OK, so the good points. Both picaxe and HERMAN were fabulous. And we have had all sorts of trouble brewing wits in the past but this one went without a hitch.

Now ... I am just hoping I can retrieve some data from my once trusty laptop. This week I will be planning a serious backup regime.

Melbourne is currently experiencing its hottest run of weather for 100 years or so. For most of this week the temperatures have been peaking between 43 and 45 degrees C (>110 F).

The ice bank I recently bought to use for cooling in the home brewery is essentially a very small refrigerator (20 litres of water/ice mix) with a thermostat that can be set between 0 and 6 degrees C. Being small it is efficient, but it also means that its cooling capacity is less than a domestic refrigerator.

A few weeks ago I found out that the ice bank was able to maintain a cellar temperature of 11 C unless the temperature in the garage got quite warm. I don't remember exactly where the capacity ran out, but I think it was around 30C.

Since then I have been running the cellar at 16C. I'm not trying to make lagers this time of year, and 16 was a nice cool environment to ferment out our recent hefeweizen. This worked very well. Not only is the flavour note on the hefe as good as we have ever made, the ice bank showed no signs of stress maintaining that temperature, and the font flooding pump ran continuously during that time as well.

With the extreme heat this week I've been keeping a close eye on things. We had close on 36 degrees on Tuesday and everything worked well. With the 43 on Wednesday, all was good until late evening. By this time the garage was sitting at an ambient 35 C, with the outside temperature not dropping much all night. At this stage all the ice was melted and the cellar was at 18C and slowly warming. I turned off the font pump and kept the cellar control running overnight. By morning there was still no ice, but the cellar was at 16C, with the garage temperature at 30C.

Thursday the temperature hit 45C outside and 40C inside the garage. During the day the cellar was maintaining 18C - but this was the absolute limit of the ice bank. Friday morning the garage was 30C and with another 43C day it will be a further test of limits. I've adjusted the cellar set point up to 18C to give the ice bank a chance to recover, as there was still no ice this morning. I chose 16C as the lower end for fermenting with Wyeast 1056, and it is good up to 22C. We are now 4 days into ferment so the yeast profile ought be pretty much what we were targeting by now.

Back to the ice bank and its capacity - it looks as though a rough rule of thumb is that it can maintain the cellar at 20 degrees C (around 40F) below ambient. Thankfully our weather in Melbourne is mostly reasonable, so ale temps in the summer and lager temps in the winter will be the norm.

When the HERMAN project was first conceived nearly a decade ago, the main idea was to create a machine that would free up time so that I could keep brewing all-grain beers despite the time pressures of my work.

As the machine became reality, it was a delight to have mash temperatures and steps automated, and even have hops drop into the brew kettle on cue without human intervention. As we have gotten use to this technology, we have used some of the in-between bits of time to do things like catch up on housework, take the dogs out to the local sports ground, and take a lunch break at our favourite local creek.

I've long dreamed of being able to check in on HERMAN via mobile phone or an internet connection. We are now getting closer to fulfilling that dream.

A few months ago I upgraded my PDA because the old one was no longer working. I use it mainly as a portable diary that syncs with my office computer. While looking around at the various models, I decided one that had inbuilt wireless would be useful for connecting into the home wireless network. These PDAs are powerful computers in their own right, and so I did a little playing around in December with Visual Studio 2005 coding simple programs and interfaces on the PDA. With a colour 320x240 pixel screen, touch sensitive control of the screen and wireless, it seems there is a lot of potential to press the device into the service of making beer.

I'm just dreaming now, but I can envisage a control panel on the PDA that has a wireless connection (locally and via internet) to the machine. It is a project that is for doing just because I can, and so really has no pressing timelines.

It was this dreaming that prompted me to give the arduino a go. I've had my eye on the 'Making things talk' book for some time now, and the online preview suggested it was the exact missing link that would help me patch all these bits together and brew via remote control.

I can just imagine it now ... HERMAN V 2011 informing me that the boil addition has just happened successfully while I am down the street at a local cafe sipping a flat white.

I'm not really one for New Year's resolutions, but an overview of directions for the coming year seems in order.

• make better beer

I am guilty of not giving enough attention to detail at times and when making beer this can become a problem. Occasionally it has been something relatively minor like not paying attention to water chemistry. In this regard probably the biggest disappointment was the Dortmunder export lager that was simply too sweet because we had not adjusted the water.

The area that has hit us worst though is sanitation. Our plastic fermenters are getting old and it seems they are harbouring more bacteria than they once did. A change recently to a better sanitiser has made a difference, but I've never really liked using plastic. In the big picture of the brewery, it is a weak link. We have recently bought up a couple of stainless steel LPG tanks. Surprisingly they are cheaper than new plastic containers and ought to do the job well. Our first project plan for the year is to cut the top off one and fit it out for a fermenter. Being all stainless it ought to clean up and be a reliable food grade container.

The final part of this picture is improved temperature regulation. The old fridges that we recently walked out the door were great bug factories - and my cleaning practices were not the best (attention to detail again). We are now able to ferment in the cellar box (effectively a larger automated 'son of fermentation chiller'). We also plan to install a cooling coil inside our new stainless fermenters to regulate temperatures without putting the tank inside the cellar.

• make brewing easier

This one is of particular importance to Leah who is unable to lift and clean many brewery items. We need to work smarter, not harder so that she can still brew. The current HERMAN machine brews 40-42 litre batches. The vessels are lightweight stainless steel urns coupled with an igloo insulated drinks container (38 litres). Even though these are all much lighter than the usual kegs, they remain too heavy for Leah. I also have an interest in minimising lifting and awkward cleaning jobs because of an ongoing back complaint.

The plans are well under way for a 'clean in place machine' based on converted LPG tanks. We intend to have tippable vessels which can be sprayed and wiped in place. Lines and taps will be cleaned chemically, again 'in place'. Brewing has already moved from the laundry down into the garage because there is more space there and it is a better 'wet' area.

I have a terrible tendency to make things too complex, so I'm excited about a number of design features that will simplify the new machine. More on that later ...

• be open to new things rather than comfortable with 'I've always done it this way'

I realised recently that I've tried to push the limits of picaxe processors. That is not a bad thing in itself, but when it comes to creating a moderately complex machine (like HERMAN) that demands a level of automation and a decent user interface, it is not always good to be blinkered. I've just ordered an arduino duemilanove with an ethernet shield, plus the 'Making things talk' book. The arduino promises to do many things better than a picaxe - menu systems, floating point maths, PID, keeping track of time, no issues with serial comms hanging etc.

I don't plan to supercede picaxes - they are simple to program and cheap to buy. I won't be buying lots of arduinos, but I do have plenty of picaxes in my parts bin. Really it is about using the right tool for the right job - and about getting over my phobia of C type languages.

• this week and next

Of immediate interest to us is kitting out an LPG tank as a fermenter. We will install a cooling coil (from copper pipe) inside so that it can be regulated with cold water from the ice bank. We will get a 3 piece stainless valve for the drain tap, and also install a port on the bottom as a 'pseudo-conical' yeast collection point. This may or may not be effective, but in any case this vessel is to be dual purpose - both Hot Liquor Tun and fermenter. The details of that will follow in another blog entry.

The other thing to do is to plumb in a water line so we can brew in the garage, and wire in a 15A circuit as well. By this time we will be busting to brew some beer because our 50 patrons since New Year's Eve have just about quaffed our entire stock.

I bit the bullet this morning and went through the tutorial I mentioned last entry as an attempt at understanding how VB and .NET handle multi-threading. I must admit there are some grey edges to what seemed to work, but after about 6 hours of both learning and coding, I now have HERMAN running serial receives asynchronously. It is very pleasing to see rock-solid performance, with no trouble to synchronise comms and stay synchronised with the picaxe.

There is downloadable code available here which formed the basis of the new HERMAN code. I don't think it is worth posting the current HERMAN code, but if anyone specifically wants a look just let me know. It is still a bare bones product, although looking nicer and functioning better by the day.

Coding is not really that exciting for bystanders watching, or those reading on blog sites, so I will spare the details.

The next step for HERMAN hardware will likely be sorting out the LCD. I had the 40x2 display working on the test bench but for some reason it has not worked yet on HERMAN. It is not as simple as I'd anticipated because it requires a negative voltage on the contrast line. To supply this I'd wired in the -12V line from the PC power supply and with a couple of resistors thought all would be fine. I'm not sure at this point what the problem is, but I was not all that happy with the solution - it is far from elegant.

I found some circuitry on the web the other day that looks like it will be the solution I was seeking.

This circuit snippet is from a driver board for a graphical LCD that Microzed supply for the picaxe. The 7660S IC is available from Farnell Australia for $1.45 per unit, so it looks like an easy way forward.

It has been quite some time since designing my last major circuit board. I've realised again that I'm much better at having ideas than actually implementing them. I think this is why I stumble over things like the false bottom - with its poorly engineered plug - because I'm already thinking about the next idea rather than fixing the current one properly.

Having said all that, the work towards final circuit boards for the HERMAN 6 control system is going well. While it has taken extra effort to climb the learning curve of PicaxeVSM and Diptrace, it has been a lot of fun as well. The opening screen of VSM says "from concept to completion", promising the production process will be easier with software modelling. It is true, and it is much more interesting doing it this way. The devil is still in the detail though, and modelling something on-screen remains a long way from working through the detail towards a finished working board.

Because I find detail challenging, I've discovered a real benefit from using this software pair. It is much easier now to design to a basic level of complexity, then go back and re-design the next layer of detail. To begin with, some basic modular bits of the HERMAN project were tested under VSM. The 595 LED driver was one of these modules. Firstly I got a basic driver circuit to work, and then when that was fine, added a second picaxe as a 'sender' to test the extended bus network. The extended bus is a way to add extra modules to the HERMAN system. Some of these have been designed (like the LED driver), while others are just ideas waiting for time to implement them. An example of one of these ideas includes a hop-dropper drive for kettle additions. Knowing that I can design this 'after the fact' means I can focus more narrowly at the moment.

The extended bus has both send and receive serial lines which means that the controlling picaxe can either control devices or receive data from external inputs. I am sure that I2C could have been used for this, but I'm comfortable with serial protocols and how to implement them on the picaxe.

As the parts have slowly come together, the VSM/DT combination has meant it has been easy to add things like test points and connecting points for adding buttons to the control panel.

I've ordered the 40x2 LCD display for the control panel and have been working through interfacing and controlling ideas, thanks to Hippy. When I have all the components on hand, final checking of the physical layout will be possible. I think it would be wise once the schematics seem to be finalised to go back to VSM, simplify the circuits once again and test them with picaxe code, just to make sure all the bits join up as they were intended.

By that time I should have all my PCB making tools gathered together. I'm hoping to get a set of high speed drills like Garrett from ebay, and I need to buy a drill press for our dremel. I will also do a test print on cheap glossy advertising paper to see if it is suited to the toner transfer method of PCB etching. In 3-4 weeks we should be producing our PCBs.

The plumbing configuration of HERMAN has been tweaked over the years into something I'm very happy about. The thing is, that a change like converting from RIMS to HERMS can raise issues that I'd not thought about before.

Below is the current plumbing schematic of HERMAN as a HERMS. You can click on the image to see a larger version.

The HERMS coil in the hot liquor tun is a simple replacement of the inline heat chamber that was situated immediately after the pump. The problem with this arrangement, is that the normal backflush of the plate chiller, which is designed to be clean of biological matter, runs through the HERMS coil. The path is via the hot liquor in solenoid, down towards the HLT, then right and up in through the pump before a right via the backflush tap, then through the plate chiller and into the kettle (or draining from this point via a hose). The problem has become evident because the old HERMS coil has been slowly releasing biological matter under the pressure of the hot liquor in line.

Without realising it, the plate chiller was partially clogged with this unwanted biological matter. It must have been loosened up with the long vinegar wash a couple of days ago. As it is impossible to see what it inside, I was not aware there was more matter in there, waiting to let go at such an inopportune time.

I only noticed there was a problem when the first runnings from the kettle to the plate chiller released some of this matter. Despite all I tried, more matter continued to come out. Flushing did not help because the flushing just encouraged more matter to flow through into the chiller via the HERMS coil.

One workaround would be to move the 'hot liquor in' connection is to the backflush tap. This would mean then clean liquor can flow directly through the chiller as planned. Hot liquor still has a path into the mash tun and hot liquor tun, via pump and HERMS coil, but these areas are much less critical compared to post kettle. One problem I can see with this is that the other chiller cleaning process I commonly use is to fill the HLT with and oxidising cleanser (nappisan) and pump this through the chiller and then back into the HLT in a closed loop. This means that the HERMS coil should be placed above the backflush tap rather than below it. Doing this would eliminate it from this cleansing loop.

The modified layout is shown below. It might not be all that clear which cross overs are actually joints and which are passes behind, but the hot liquor in is connected just below the chiller, the the HERMS coil is only connected on the outlet of the pump. I will redraw and post a clearer schematic when I get time.

It looks like another extensive clean is in order, of both the coil and the chiller. And with a plumbing modification, I think we will be good to go.

I'm not sure whether to laugh at myself or kick myself. I guess I should laugh because I've never had any tuition in fluid dynamics (I'm not even sure that's what its called). My working with liquids, transfers, pumping, heating, chilling and the like is largely based on what seems intuitively right and seeing what others have done with their brew sculptures ...

Anyway I had trouble getting the recirculation path to flow while brewing yesterday. Actually I've had trouble for a while. I've had air get into the inlet side of the pump. March pumps don't like to suck air as they are not self-priming. The batch before yesterday, I forgot to underlet the false bottom - which potentially left air trapped under a heavier than usual grain bill. The other problem with that brew is that it had around 70% torrefied wheat in it. That is a challenge for any system, let alone a HERMS or RIMS. The other issue was that the connection between mash tun and false bottom was simply a piece of nylon hose that slipped over the copper pipe. Over time this was getting looser, and may have been a good place for air to sneak in.

Anyway I came up with what seemed to be a bright solution. If air is being trapped in the false bottom, why not provide a vent to get rid of that pesky air?

So while I added a hose clamp to keep the connection tight, I also removed a plug from the top centre of the false bottom and created a vent to the top of the mash via the shaft of the mixing paddle (see sketch below - click on it to get a larger version).

When I tried to recirculate (now confident there would be no air under the false bottom), flow was very slow. Obviously something was wrong. So I turned the pump off and began to drain some liquor from the mash tun into a collection vessel so that I could heat it on the stove as a pseudo decoction. As I was collecting, I noticed something quite strange. Liquor would flow then suddenly there was a gurgling that sounded like air and liquor mixing, with a dramatically different flow into the collector. Then things would restore to normal, only to have the cycle repeat. This was the point that I wanted to kick myself - the vent was now sucking air into the false bottom because the pressure under there was less than above while liquor was being removed.

I stopped short of self-harm when I realised that by taking the mixing paddle out of the mash, I would immediately close this vent to the outside world. And once this was done, I had a fabulous flow rate in the recirculation loop. The problem was that by now I was so behind in the little things that are part of a brew day, that I did not have the luxury to record things well.

So it looks like I should put the plug back in for starters. And always underlet the false bottom. And possibly ask more questions from those who actually understand these things.

Anyway, the Belgian Wit is fermenting nicely this morning. It is a bit early to tell if I will have trouble with the matter that escaped from the old HERMS coil with some ending up in the fermenters. I have some ideas about how to fix that problem, but that is for another blog entry. I also have some ideas about making the coil easier to use on a brew day, as it was a bit awkward trying to clean the HLT yesterday.

But after all that, some photos from the brew session yesterday, with my favourite the pile of citrus zest that went into the boil. As a note, last batch I steeped the citrus in a closed kettle for 10 minutes prior to chilling. The fragrance was very prominent during fermentation, but there is nothing left in the final brew. This one does not seem to be 'letting go' of the fragrant oils. The difference was a 5 minute boil before a 20 minute steep prior to chilling.

Above: 12 mandarines and 6 navel oranges zested and ready.

Above: The mash during the protein rest. You can see the mash inlet manifold above the grains.

The kettle during the early part of boil.

All this thinking about HERMAN history has made me think about how well HERMAN has been functioning. I actually think that the first version of HERMAN was in many ways better than the current one.

The first one had plastic buckets and leaked way too often - so I'm not talking about the materials it was built from. Rather the simple heat exchange coil in the hot liquor tun with a recirculating pump worked really well.

Since 2001 I've changed quite a few things - all in the name of progress. While some things have really been good improvements, I'm using this reflective frame of mind to question some others.

The coding of the first machine was okay, but was really inflexible. One of the great improvements to coding was incorporating a HERMS batch language where a full brew session could be controlled by an easily changeable batch sequence.

Moving from an analog interface to a fully digital interface was a great advance. Those early temperature sensors were not well matched to a 5 volt system and there was lots of induced noise that I could not trace with the equipment I had on hand. Using Dallas DS18B20 temperature sensors has meant much better reliability and accuracy.

Having a control circuit that allows for a fully functioning machine without a PC attached is one of the greatest improvements - considering how often Windoze seems to crash. The irony is that with a dedicated picaxe control circuit, I have not yet fully implemented the PC interface to allow the best of both worlds. This means that the PC control programming with all its decision-making power and process control is not currently able to drive HERMAN. In this sense, the original machine was more capable than the current 5.5 hybrid.

The first HERMAN machine was a little slow moving from one step temperature to the next. It is difficult to change mash temperatures quickly when you are relying on a coil within a large vessel of water (in the hot liquor tun). When HERMAN was upgraded to a 40 litre system, a separate smaller heat exchange bucket was added. This meant the system was still a HERMS, but the hot liquor tun was independent of the recirculation loop.

This seemed like a good idea and it did work quite well - but it worked against principles of simplicity. I am quite challenged when it comes to keeping things simple. At the time I did not even have any idea why this option might not be as good as keeping the heat exchange coil in the hot liquor tun.

The extra bucket meant that I had lost the advantages of the original HERMS concept. It is beautifully simple in concept. The hot liquor tun needs to be 2-4 degrees C above the mash temperature at mash out and this is typical of the heat difference needed to maintain recirculating temperatures ...

Given the power demands of changing temperatures on a 40 litre electric system (at least 15 amps at 240VAC), the most recent incarnation of HERMAN was converted to a RIMS system. This uses a heating element in line in the recirclating loop to maintain mash temperatures. At the time I thought this was making progress, as direct heating is more efficient than using the coil in the hot liquor tun. I'm coming to the conclusion that it is much more difficult to control liquor temperatures this way, and that it means there is an extra control system needed as well. Both of these things work against the simplicity cycle.

I can vaguely remember the beauty of the original system. Once the heat loss in the system is known (and it is relatively constant - although affected a little by external temperatures), it is quite easy to dial in a hot liquor tun temperature and then watch the mash hit its target and stay there. With a heat chamber in a RIMS configuration, PID control is needed to try and work out how much heat is required and for how long before an overshoot happens. Timing and calculations are everything - and that is after you have figured out how to tune the system.

So I think the next run of HERMAN will be in a test configuration as a HERMS - just to satisfy my curiosity. I remember fondly of the set and forget process with the HLT. Maybe it will find its way back into HERMAN 6.

Oh well, 6 steps forward, 5 steps back. Even if I reinvent the wheel, so to speak, I have enjoyed the process.

The current HERMAN machine is running a patch-work of bits and pieces from old version 5 control hardware. It is currently being pressed into doing something it was never designed to do and it is literally sticky-taped together.

Version 5 used a picaxe 40X to communicate with a PC and read inputs and send control commands to an 18X picaxe. It was never designed to operate as ‘intelligent’ hardware without a PC coaxing it. This is a legacy of HERMAN history. The first machine used a 'dumb' analog interface hanging off a PC parallel port. All the processing was done in VB code on the PC. All the interface did was tell the PC what was happening or switch according to what the PC demanded. The version 5 interface used picaxes to make the interface digital (easier to use and less prone to analog noise). It was never designed to do its own processing, so the PC was still needed.

A key design parameter for HERMAN 6 is that its hardware will be capable of ‘stand-alone’ control. Too many ‘blue screens of death’ from Windoze and PC lockups meant that this was not simply luxury, but necessity.

The picaxe micros are capable of significant processing power. They can be pressed into PID control, for example. The hybrid patch-work interface we are currently using on HERMAN (5.5) is the old version 5 interface with a pushbutton panel that was prototyped for HERMLET. The picaxe coding is substantially different now, with PID worked out on board alongside other processing decisions.

The ultimate aim for HERMAN 6 (and therefore HERMLET) is for the picaxe controls to provide machine control that does not require a PC connected. A connected PC may be added to provide more complex control decisions towards the goal of automation.

So … some design ideas for HERMAN are on the drawing board.

And despite all that has been said about HERMAN, he still makes damn good beer. Cheers. Arnie

Probably the single most thing that cripples me when it comes to inventing, designing, creating is that I have too many ideas. This means that rather than figure out a design brief and stick to it, I have the tendency to figure out extra things I can add well into the design phase to ‘make it better’. I’ve blogged previously about the Simplicity Cycle but still find it a hard discipline to master.

One problem with having a fertile imagination is that sometimes the number of options available becomes overwhelming. Another problem is the lack of being able to focus on a single or simple plan of attack or order of things to do.

This is one reason I find blogging therapeutic. It forces me to commit (at least in bits and bytes) to an idea, or thread, or line of reasoning. It also helps to figure out what is really important.

Last November I hit a creative wall with HERMLET. Often time away is as good a dose of therapy as anything. One problem I had with the HERMLET project is that the first phase was too ambitious – despite attempts to keep it simple. It is a nice idea to create a machine that makes all-grain beer with the press of a few buttons. My problem was that I had confused end-user simplicity with construction simplicity.

To achieve the ambitious goal mentioned above, the coding demands on a picaxe were becoming overwhelming. It probably is achievable, and may well be the focus of further development, but to do it as a first step was too ambitious.

I need to keep telling myself that small steps are good. Working on a simple problem and getting it to work properly before attempting the next problem is good. Doing proper research (coding, components, asking questions etc.) is good. Working slowly is good.

Having the right tools for development is also a good thing. With the picaxeVSM program, I can literally work on small parts at a time and figure out code, circuit, components all at the same time, with the confidence that it will work right when made in reality.

So with the wall now dismantled, it is time to design and build.

Not a lot has been happening at least by way of construction on HERMLET of late. There have been way too many distractions which have included making beer to keep up with the increasing demand of an Australian summer, working out how to make the most of podcasts and lots of other things.

That is not to say that our inventive minds or making hands have been idle though. Some time has been put into the HERMAN machine which was left in a very raw state in it's most recent configuration while HERMLET took the focus. What we have done in recent weeks is use some of the new HERMLET technology to improve HERMAN. The prototype control panel has now been pressed into service on our brew rig and has already made it a lot more reliable and easier to use. If we can code and solder in the LCD before next brew day, it will also make the laptop an accessory rather than necessity.

The podcasts have been great as well. Now when I am cleaning the house, I am just as likely to be listening to the Jamil Show or something just as useful on my iPod. And it is through listening to Jamil that I've been doing a lot of thinking about chilling from kettle to fermenter.

The original HERMLET was conceived as a very simple/easy to use automated brewing system. It was inspired by (amongst other things) the brew in a bag movement (BIAB) and the no chill movement. Both methods are aimed at keeping things as simple as practicable for the homebrewer.

HERMLET has developed in complexity beyond BIAB but it retains its heritage in that idea. The issue regarding chilling has caused further thought through listening to Jamil.

HERMAN currently uses a plate chiller and while it is very efficient, it is not really possible to dial in a fermenting temperature in one easy step. Currently the chiller gets the wort down to the high 20s Celcius and a refrigerator works hard overnight if we need to drop it to lager temperatures.

When I first began all-grain brewing I used the cold water in the bath method of cooling wort. This is very slow and inefficient and is the likely reason an early beer I was very proud of was marked down significantly due to DMS. I soon upgraded to an immersion chiller and used this for a long time but was not satisfied with its inefficiency and in particular the amount of water it used. Water use is a very real issue in many parts of Australia. In Melbourne our water storage is currently at 42% and we are still in Spring. Many parts of the country have experienced prolonged drought. Water efficiency is a prime consideration at our place in a brewing system.

About 3 years ago we built a counter-flow wort chiller that served us very well. Its major drawback was its size. There are nagging concerns in my mind about both counter-flow and plate chillers because I can never be completely certain they are sanitary. An immersion chiller is boiled before use and the exposed surface is all that touches the wort. It can physically be cleaned of organic matter in a way that plates and inner pipes in a counter-flow cannot.

In talks about Helles and Pilsners, Jamil has raised that other fear for me - DMS. He argues that while a long boil (90 minutes) helps, another significant factor is chilling wort quickly. He argues that an immersion chiller acts on the whole wort and will rapidly cool it to an acceptable temperature regarding DMS production, while plate and counter-flow chillers only flash chill a fraction of the wort at any time. Much of the wort will stay too hot for too long using these last two chillers.

Jamil has developed a whirlpool system using an immersion chiller and a pump that recirculates wort back into the kettle. The whirlpool creates a far more efficient chilling system and has the added advantage of dumping hop debris and trub into the centre of the kettle.  By taking wort from the outside of the kettle when filling a fermenter, this matter is efficiently filtered out.

There is much to like about the idea, but there are a few issues to deal with before adopting it. Going back to the issue of water use, we intend to install a rain water tank that will not only be the source water for the chiller but will also take the "waste" water. This water will then be re-used on the garden, ensuring it is not wasted at all. The whirlpool idea also encourages a greater efficiency than my earlier immersion system.

The other issue I have with Jamil's idea is the method of whirlpooling. We have used a pump on the kettle outlet before and run boiling water through it meticulously as a method of sanitising. One day we did a double batch and both brews were rancid through a small amount of grain matter that was stuck in the pump housing. The housing is not something we want to take apart on a regular basis, so rather than use a pump for the whirlpool, we think a simple motor driven paddle that is incorporated into the immersion chiller is a better option. This would be sanitised as part of the boil in the same way as the chiller.

In the summer our city water supply temperature gets up to about 20 degrees Celsius, so to achieve lager temperatures we will experiment with iced water similar to Jamil's method for dropping those last degrees.

Now it is time to think about rain water tanks as we are experiencing some heavy thunder storm activity. It seems a shame to see all that water running down the drain ...

By the time the sun rose on Monday morning, the itch to restart the Munich Helles brew just needed to be scratched.

But first, we needed to do some repairs to our trusty rig.

It was fascinating to look more closely into the failure. The wort had obviously gotten very hot as there was a lovely caramelised wort plug in both the nylon hose that had failed, and just after it in the copper pipe where the temperature probe was housed.

The nylon hose had failed immediately on the chamber side of the plug. This makes sense, as the super-heated wort would have been under pressure, partly from the pump trying to push it through the obstruction, but mostly from the heating in the chamber.

The hose failed because the nylon was forced outwards away from the reinforcing twine with a combination of temperature and pressure. The main failure point shows a 5mm hole.

We decided to remove the plugged inline temperature probe for now and use the probe where the wort enters the mash tun as the probe used in our PID control. This simply meant installing a slightly longer nylon hose to connect the heat chamber to the rest of the system.

We also removed the heat chamber to clean it thoroughly and to test and clean the element. The result here was surprising.

The element was in very good condition considering what had happened, with only the smallest amount of carmelised wort present. I suspect that the machine had tripped off on an earth leakage well before the element had a chance to burn itself up. This could have been for any number of reasons, but hot sticky wort found itself into several 240VAC plugs, any of which could have tripped the breaker.

We replaced the faulty connecting cable between our control picaxes, coded in a failsafe condition in the control picaxe, and then completed our quickest brew day ever, taking 3 hours 40 minutes to have everything fully cleaned and packed away.

The next step will be to upgrade our control buttons.

cheers, Arnie.

Last night we had a catastrophic failure on HERMAN 

In the space of a few minutes being distracted when a friend arrived, HERMAN went from a normal HERMS mash to spilling all the mash liquor and scorching the heat chamber element.

The first sign that something was wrong was a burnt element smell. The second obvious sign that things were not well was that the machine had tripped off on an overload. The third obvious sign that things were really bad was the absence of liquor in the mash tun.

From there, a bit of CSI-type sleuthing was needed.

The leak was easy to find. On the outlet side of the heat chamber we have a very short length of reinforced nylon hose that joins the chamber to the rest of the plumbing. This is in effect a safety valve - and it had done its job and ruptured.

The pattern of wort spray suggested there had been some pressure behind the rupture, although at this stage we did not know why. The other important bit of information was that the driver circuit to the heat chamber was showing itself to be 100% powered on - where it should have been a relatively low duty cycle like 20%. (The control gear has a battery backup, so was still running).

The penny was beginning to drop. We knew there was an occasional faulty connection in the cable that connects our main picaxe chip to our control output chip. The problem with the picaxe we are using for outputs is that if a serial command is not received properly (as can happen because of the faulty cable), then it will wait and wait and wait until things are received properly.

At best this means that controls are not updated. At worst, controls will freeze in their current state, whatever that happens to be. It just so happens that the coding has not added a failsafe to ensure outputs stay OFF once serial communication has been initiated. Some simple coding could have prevented this disaster ...

At least I can be a little philosophical about it. If I had fixed the connecting cable earlier, I would not have considered coding in the failsafe. The problem could have lurked around for a lot longer because one masked the other.

Anyway, the graph below captured the story of what happened while we were distracted. The purple line shows the liquor temperature just before it is returned to the mash tun. The red line shows the mash tun temperature. You can just see it beginning to rise on the right (it actually rose quite rapidly after this point). The purple line rises dramatically and then almost flatlines as the temperature back into the mash can barely get any hotter. It peaks at 91 degrees C. The green line is the heat chamber temperature. It is easy to tell when the control system failed and turned on the chamber to 100% power. At the 24.5 minute mark (probably the time the doorbell rang) things went terribly wrong. The green line keeps rising until the chamber is boiling. At 29 minutes the purple line drops dramatically. This is the point where the hose has ruptured and the liquor is no long recirculating.

 OK, so we lost all our wort, and we may have lost the heat chamber element (yet to be tested). But we have also gained insight into how important failsafes are when you are trying to automate things like pumps and 3600W heating elements. It highlights how important it is to finish jobs as well. We even have a liquor sensor plumbed into the outlet of the heat chamber, but it is not yet connected to anything.

Having said all that, it will only take a few smallish tweaks, and HERMAN will live again.

In the Make: tools n' tips email that I just received they mention a book by Dan Ward called 'The Simplicity Cycle'. It's thesis is that design ought to be simple and elegant, not too clever or complicated.

As a maker who struggles to keep things simple I was immediately drawn to look further. I was pleasantly surprised to find that it is available for a free download.

The HERMLET brewing machine project was conceived from a simple idea, and I think I'm probably in danger of over-complicating it again. Thanks Dan for making this text available.

"What are we going to do today, Brain?"

"Same thing we do every day, Pinky...
try to take over the world."

The 'Pinky and the Brain' cartoon kind of captures the vision of Google, as their search engine, mapping, picture sorting and 3D software seems to be more pervasive than even Microsoft.

I'm not much of a conspiracy theorist, just observing the ever far reaching and clever marketing of Google. In fact I'm quite impressed by how clever it is.

I downloaded their 3D drawing program, SketchUp, the other day thanks to an article in Make: magazine about it. It is the kind of 3D program I've been looking for for ages. Not only is it a free download (the best price, you could argue), but it is so easy to use. The integration of wonderful tools like online catalogs of structural components, and the ability to draw a building and then place it directly in Google Earth for everyone to see is quite impressive.

I guess they are very clever at understanding human nature and pitching it to their advantage. I am not concerned enough about Google filtering my web searching with their own specific algorithms that might miss some things to do anything about changing my search engine. I like what Google gives me and especially how easy it is to use. And then they make it even easier, and now I have a little toolbar on my web browser so that I can find google anytime.

And their photo program, Picasa, is easy to use, and the kind of photo sorting program I've wanted for ages. I can now very simply catalog the many digital pictures I have and never lose them again through poor sorting. I haven't even begun to look at putting photos online to share with friends.

And then back to SketchUp. How clever to allow a worldwide web of individuals who love to play with their computers and have lots of disposable time to create cities in 3D and add them to Google's own tool to take over the world, Google Earth! I'm not one with enough time or the desire to do it, but what a great marketing angle that has a flow-on benefit to me. A free 3D program that I can use to design anything I like.

So I can admire Google's efforts to take over the world, and in particular their clever understanding of human nature. Now ... it's time to make something ...

It may be mid-winter here in Melbourne, but yesterday the sun was shining, we were brewing, and mostly HERMAN did the hard work for us while we sorted out fermenters, cleaned a few bits and pieces and reflected what to do with a Belgian wheat that was over-bittered.

We were due to make our American Pale Ale last week except that we were distracted by the HERMLET project (see previous blog). But this week we loaded our wort with Cascade hops and enjoyed the sweet fragrances that brewing brings with it. This beer follows the same recipe from one a few weeks ago where we used Nelson Sauvin hops in place of the Cascade. The Lady Nelson Ale is already a hit with all in this household.

Anyway, after a few minor tweaks, HERMAN did his thing again and this time the PID control worked like a charm.

In the graph above, the dark blue line at top is the hot liquor tun oscillating gently around the set target of 80 deg C. It does not have PID control. The dark red line shows the mash temperature and heating is controlled by the PID controlled heat chamber. The red line is smooth and hits target perfectly, even though the green line still oscillates. This means the PID could still do with some minor tuning, but we are very happy with the net result.

Next weekend we will either brew a batch of Belgian wheat beer to revive our last one (which has too much bitter pith from citrus fruit in it), or we will continue work on HERMLET.