The Swagelok salesman . . . 3 o'clock and stop").

His advice was good EXCEPT that for the smaller fittings (less than 1/4"), you would tighten the nut 3/4 of a turn past finger tight per Swagelok's published installation instructions.

I'm back!

The machine has been working very well. Oddly enough, though, I pulled off the group cap nut today (a leak showed up out of nowhere) and there's some white sandy crud accumulating. I have soft water water and am running the tap water through a descaling and carbon filter (as well as the mesh screen at the machine inlet). Any ideas where this deposit could be coming from?

Thanks!

aindfan wrote:Update: I ordered a new safety valve and reassembled the tee. With enough teflon tape around the valve threads, my probe setup is leak-free and probe temperature reading confirmed by my multimeter. Now all I have to do is order an Arduino, design and order a PC board, and order the parts. My order of 4 Auber 25A SSRs and heatsinks already arrived.

Get one of the TC4s. I'm just getting up to speed on Arduino development and have some of the parts on order for the TC4 still. See http://code.google.com/p/tc4-shield/ and a long discussion thread at http://www.homeroasters.org/php/forum/v...ad_id=1774 . I've got a couple of Arduinos, TC4s , a couple of the LCD adapters too. It's a really nicely built board.

Mark

Mark, I'm still trying to decide what to do about the TC4, as I'm considering the idea of designing my own board heavily based on TC4 but with some more SSR outputs.

Thanks!

Well, it's been a while since my last update, and not much has changed. However, I finally got my Arduino Uno and a protoshield, with which I wired an SSR driver based on the design in the TC4 (Arduino digital pin drives the base of an NPN through a resistor, which turns on the SSR). I'll be ordering a TC4 and LCDapter from Jim as soon as he gets them, but in the mean time, I'm going to have a simple setup for preinfusion.

The program on the Arduino will be set up to open the solenoid on first button press, turn on the pump on second button press, and shut off both on the third press. Another option would be to turn on the pump after a set, preprogrammed delay. Either way, I'm basically replacing a simple mechanical switch with $50 of circuits, but it's a fun exercise and a nice way to practice soldering before the real TC4 board is available. I'll try to post a picture of the hacked-together setup this evening.

Turns out I had it wrong for the SSRs that I have. All they need is a voltage above 4.5V across the input to activate the output, so after some pointless troubleshooting, I finally read the datasheet and connected the SSR inputs directly between Arduino digital pins and ground, and they now both work.

This means that I can do a number of interesting things, such as timed (or manual) preinfusion or an automatic backflush program!



I set up another program to actually use the machine, where each button press switches through the sequence preinfuse -> pump -> off. I will probably try to get a bit fancier with this, such as holding down the button to start the backflush sequence, but I won't need that once I get the rest of the TC4 stuff (with LCD and four buttons).

Hi Dan,

It's been several months since I looked in on the forum and your thread....

Have you solved the HX leak yet? How did you do it?

Do you really notice much difference in the shots using the Elektra standard ramp-up vs. your pre-infusion method? Do you think it's necessary, or just "cool"?

A thought about what you are doing with the board programming. As you may recall, my Elektra is an early automatic machine that is controlled not by the amount of water flowing past a dosimeter, but by a mechanical timer that you set to the number of seconds you want the shot to run. With your board, you could set up some pre-timed extractions and make your machine work in the same way mine does (albeit with a programmed digital circuit board timer instead of a mechanical timer). It would free you up while you steamed milk, or did other stuff, or just admired your "cool" converted automatic machine.... I would provide for some kind of "reset" button so that you could end the shot early if you wanted to (if you saw blonding occurring early).

My 1979 Elektra is functioning flawlessly so far.... I decided not to install a pre-infusion circuit. I am interested to hear back from you what your opinion is of the difference in the shots produced. I still do plan to install the manual push-button switch I bought so I don't have to use the timer when I don't want to. The timer activates the pump via a solenoid now, so I can simply wire directly to the switch and don't need to add any additional relays at this point. Another reason for installing a manual switch is that at some point the mechanical timer will fail and I want a way to pull shots when it does. (I doubt the timer is replaceable at this point.)

One last question -- I saw that you had a similar problem with the bypass valve failing as I did. Did you try simply replacing the bypass valve and seeing if it fixed your existing pump (it worked for mine)? If not, you might consider getting it so you'll have a backup pump down the road when you may need it....

I look forward to reading your replies.

Philip

Phillip,

Hello again! As someone predicted a few pages ago, the leak just sort of scaled itself shut, so I'm in the clear until my next descale.

Ah yes, preinfusion... yes, I've realized it's mostly there for the coolness factor. Elektra's flow-restricted pressure ramp works much better than my line-pressure preinfusion. Here's my theory on this: you know that sound you hear at the beginning of each shot that tells you about the pressure ramp... the shhhhh---thhhhtttt? If I line-preinfuse, the group's built in ramp gets "used up" on the line-pressure water, and therefore, when the pump turns on, the puck gets slammed with all 9 bars. Though I don't know how much slower the group's own ramp to 9 bar is, I think it works better than having my own preinfusion. Another thing that I noticed is the huge difference in volume - the same basket prep (in this example, 15g or Terroir Northern Italian Daterra) yields 1oz with no line pressure preinfusion and 2oz with preinfusion until the first drop, for the same pour time.

The current program that I'm running is a hack of the aBourbon program (and the associated LCD, PWM, thermocouple, etc. libraries). The four buttons are:

1. Increment espresso state (off -> solenoid on -> pump on -> off)
2. Backflush (~4sec on, ~1 sec off, repeat 8 times)
3. Screen backlight on/off
4. Increment temperature by 1F (default power on set point is 235F, this button allows a temperature up to 260F and then goes down to 220F)

I also have the LCD showing the "heat level" - the TC4 project includes a pulse width modulation library that allows very, very fast toggling of the boiler heat SSR, so rather than only having an on or off state, this can be set to a "heat level" from 0 to 100, and the LCD shows the letter H followed by one tenth of the heat output (my inspiration for this was the blinking cursor on the Strada that shows when the heating element is powered, and the number is more practical here given the availability of the variable heat setting).

But I digress... the answer to your question is no, I don't have it programmed for a timed shot because I would prefer to stop the shot manually (and switch between preinfusion or none). My group state increment button also allows me to do a longer time line-pressure HX flush to warm up the group (and not the other way around) - note that I mentioned the default temperature of 235F, in line with Ian's 1 1/2 boiler HX machine idea. This slow warming flush and low boiler temperature (~0.5bar according to the gauge) gives me a very stable 200F as long as I don't overflush.

Having that push button in parallel with your timer certainly won't be a problem, especially if you suspect that the timer will fail. Because you have a mechanical switch and not a microcontroller that is allergic to line voltage, you won't need an SSR for every solenoid that you control like I did.

The pump was just plain acting up, and seeing as neither of the two companies out there today confessed to making it, I called FOT and had a replacement in the mail in two days (they're only an hour or two up the highway from me). A healthy dose of teflon tape later (for the pump to mushroom fitting adapter) and I was all set with the new pump in place. The original pump is sitting in the closet, and has undoubtedly siezed up by now. The plastic bypass valve assembly had some chips missing and didn't seem worth the trouble (and I couldn't open the pump to do a full rebuild). In this case, the breaking point that prompted the replacement was that pressure settings started getting unreliable.

Glad to hear your machine is also working well! My only problem is a rattling motor - I will need to either outboard it or get some dampening mounts at some point. The outer case of the machine is just sitting on the floor because it creates too much of a racket when attached to the machine with the problematic pump.

Hi Dan,

Thanks for the response.

I had a couple of thoughts regarding the transferred noise from the pump motor. I noticed on my Elektra that the motor was touching the frame. A little judicious loosening of the mounting bolts and slight movement of the motor made it sit without touching the frame (other than where the mounting bolts are at the bottom) or the exterior case panels. This quieted down the noise produced a lot, -- it went from an irritant to more of a loud purr. If you can't keep it from touching the frame or the sides, try getting some strips of neoprene rubber about between 1/16th and 1/8th inch thick, and attempt to force it in between the areas of the frame and/or outside panels that are in contact with the motor frame. This should also help quiet it down.

It will be interesting to see how you solve the over-temp sensor issue. My machine originally had the same setup yours did -- but the parts are long gone. I also am trying to figure out how to fit a sensor to the heating element flange in such a way that I can run the element power through it. The replacement element I found and installed did not have the place for the original type of sensor probe that was originally fitted to the element. I am not in a rush to install this as my machine is plumbed in and mechanical boiler fill control is working properly and freely. -- so the chances of the boiler element ever getting exposed to air is pretty slight.

Oh, one more thing. I found that the stock feet that screw into the frame were a bit too short to allow the plumbing (especially the drain hose) to completely drain out, which left a low spot that was acting a bit like a trap and I knew it would be a matter of time before it would silt up with misc sludge and cause a difficulty. Sooooo, I went to my neighborhood hardware store and found some all-metal threaded feet that allowed me to raise the unit about an inch higher -- just what was needed to get the drain to completely empty. The threaded "feet" I bought were an SAE thread, but it was close enough to fit into the metric threads on the frame without stripping them, and I used a nut to lock the "feet" into the height I needed. The new "feet" look kind of like a quarter with a threaded metal stud coming up from the middle.

Good luck with the remaining tasks you are doing.

Philip