DIY Thermofilter Device

[Endo]

OK....before I start. Let me say that my first attempt here was a classic "fail". But in any case, it was a fun experiment and you guys might get a good laugh, so here goes:

I rented the Scace from Gregor not too long ago. I used it to calibrate my machine and learned a lot using it. These Scace devices and Fluke Thermometers are not cheap (close to $500 I think), so I have to give big thanks again to Gregor.

Of course when I was renting it, I had a good look at the construction. It's a very simple device. Basically a naked PF with sealed basket, flow restrictor and a thermocouple in the center of a plastic puck where the coffee normally would be. It got me thinking of ways of making my own.

I while ago a friend gave me his cheapo t-type digital thermometer he got off e-bay for $10.

http://cgi.ebay.com/2-K-Type-Digital-Th ... 3a5ad64712

I would have preferred a 0.1C meter like the Fluke 52, but this would work for my first attempt.

All the other parts I needed (a spare double basket and a flow restrictor) I already had. I just needed something for the plastic puck.

I thought I'd use a glue gun with a high temp melt stick to make the puck and seal the basket. The thermocouple would be route up through the hard glue puck and then bent over the center hole I drilled and tapped to install a fitting with a 1/32" pinhole restrictor. So here's what it looked like during drilling:

Basket with glue being drilled

IMG_1695 (480x320).jpg (53.6 KiB) Viewed 39729 times

All looked great and I installed in the machine and gave it a try:

DIY Scace 1st run

IMG_1696 (480x320).jpg (53.8 KiB) Viewed 39729 times

DIY close-up

IMG_1698 (480x320).jpg (54.93 KiB) Viewed 39729 times

Unfortunately, the glue did not hold up well and started to get very soft during running:

Glue meltdown

IMG_1701 (480x320).jpg (46.66 KiB) Viewed 39729 times

I had suspect this might happen, but I figured it was worth a shot.

All good fun. I think it would have worked very well apart from the the glue meltdown since the temps were coming close until the restrictor gummed up.
So next, I plan on making a better plastic puck and put a 53mm dia. o-ring under it (like on the real Scace). Anyone got an idea for the puck material and where to get such an o-ring?

[MichaelN]

It got me thinking of ways of making my own.

Thanks for posting, now I know exactly what I'm going to do!

I'm going to wait unti you finish having your fun building and perfecting this and then I'm going to try and borrow it from you next time I'm in town!

I will keep my eyes open for something more suitable for a puck and I will also look around locally on Kijiji & Craigslist for a 0.1C digital thermometer. I'll PM you if I find anything interesting.

Keep up the good work!

Michael

[Endo]

I'm thinking I'll look around for a small 1/2" thick piece of plexiglass and see if I can also find a 53mm hole saw. It should cut out a perfect plastic puck and it will even have the hole in the center. Off to the store!

[JohnB]

How about a 2 part epoxy?

[Endo]

Yes! You're reading my mind.

Couldn't find a hole saw or o-ring the right size but got a batch of epoxy instead. I had some for my surfboards but it didn't mix well (too old) so I'm redoing it.

I kinda like the idea of having a removeable puck, so I may just line the basket with plastic wrap and use it as a puck mold for the epoxy. Then hold it in place with small wood screws through the bottom of the basket. This is how they do it on the Scace, but they use an o-ring to seal down in the corner between the puck and basket.

[oton]

It's not easy to use a blind basket to avoid make a fake puck?

[Endo]

It's not easy to use a blind basket to avoid make a fake puck?

It's possible, but the epoxy puck is nice since it supports the threaded restrictor and also holds the t/c wire. Also, when I examined the Scace, it seems it uses the plastic puck to fill the basket to simulate the coffee puck and also give the actual headspace in a real shot. The t/c tip sits over a countersunk hole in the center of the puck where the water comes down through the restrictor. Since it's not too difficult to make the puck and it makes the whole thing much more solid, I think it's the way to go. Total cost so far: $10 for the meter and about $5 in brass restrictor and glue......I like.

I have my second epoxy puck hardening now. The thermocouple reads 38C above the hardening puck! (some serious chemical exothermics going on ). I'll send pictures soon.

[oton]

Very nice Endo. Keep us informed, please.

After this and after the DIY drain method, next step is DIY Speedster and DIY Synesso; some bolts & nuts, a few meters of tubing and a can of chrome paint, it will be not very difficult.

[Endo]

Alright. Tried the epoxy puck. Here are my thoughts:

I think the epoxy puck has a lot of potential but... The biggest problem I had was the brittleness of the epoxy. I think it may just be the type of epoxy I used. It was a 30 min type and it hardened up like maple candy. I tried drilling but it fractured very easily:

Epoxy Puck

IMG_1703 (480x320).jpg (41.67 KiB) Viewed 39703 times

I also covered the t/c wire with heat shrink tubing rather than tape this time (sheds water drops better and is more durable). I also sealed all threads on the restrictor with teflon plumbing tape.

Bottom View

IMG_1704 (480x320).jpg (44.6 KiB) Viewed 39703 times

Still, even with a few small cracks in the epoxy, it sealed very well and worked surprisingly well !

Next time I make one I think I'll either try and find a less brittle epoxy (like G/flex), or I'll put the restrictor in first (with a straw breather tube sticking up) and pour the epoxy in the basket last. This way I won't need to drill or tap the hole and risk fractures. Also, I think I'd put the end of the t/c in some kind of bendable metal guide tube so it can be aimed in the center easier.

I tried doing some tests. It seemed to perform much like the Scace, but the temperature seemed to be biased low by about -3C. As with all these devices, it's important to do a boiling water test and calibrate it. The other obvious problem is the lack of 0.1C resolution. Since I'm out of t/c wires to make another, I think I'll just buy another meter. I plan on trying this one this time and get the 0.1C accuracy for $20.

Digital Thermometer with 0.1C

t.jpg (18.48 KiB) Viewed 39693 times

http://cgi.ebay.com/Digital-Thermometer ... 35a93e0839

As Borat would say "great success".

[JohnB]

Looks good so far. I was thinking about something like JB Weld which can be drilled, tapped, ect. You can find large o'rings at most bearing supply shops or McMaster-Carr.

[Endo]

I was thinking about something like JB Weld which can be drilled, tapped, ect.

That certainly seems strong enough.

I think the ideal material would be a polyurethane material. (Something like a grippy skateboard wheel material). I've seen some crafts where they use a 2 part resins in a mold to make plastic parts. That might work well. Not sure where I buy this stuff though.

It would be nice to spray the inside of the basket with mold release then use it as the mold to make a poly puck. I could then easily drill, countersink and tap the center and drill a smaller hole for the thermocouple as well (this would then be sealed). You could then line the bottom of the basket with my innertube baskflush disc and use two small wood screws to pull it down tight to seal the bottom of the basket. This would be almost identical to a real Scace. I could make a bunch of these for anyone interested. All they'd then need then is the $20 digital 0.1C thermometer and a spare double basket.

Of course if someone had a lathe, it would be even easier to turn a bunch of these plastic pucks.

[Richard]

How about sourcing Delrin in a thickness approximating that of the plastic puck? Delrin can be turned, machined, bored, shaped, threaded, etc., and it is extremely durable. The tooled plastic puck could be sealed in place by any of the aforementioned techniques.

[Endo]

How about sourcing Delrin in a thickness approximating that of the plastic puck? Delrin can be turned, machined, bored, shaped, threaded, etc., and it is extremely durable. The tooled plastic puck could be sealed in place by any of the aforementioned techniques.

That would work well, but I've come up with perhaps a simpler method.

Have a look at the rubber puck from the rental Scace in these photos from Feb 18 when I took it apart to fix it:

http://www.s1cafe.com./viewtopic.php?f=57&t=1107

The grey puck is all I really need to make.

Perhaps someone with a Scace can check the dimensions for me again, but looking at a double basket, it appears to be a 50mm diameter and roughly 15mm thick. The material is a flexible plastic (polyurethane?) The hole in the center is threaded for a standard 1/4" NPT fitting (use a 7/16" drill and then tap) and then a large countersink on top. The hole for the t/c wire on the side can be drilled 3/32" and then sealed afterward around the t/c wire. The Scace puck has the lower outer edge grooved so an o-ring can be fitted in the corner so the water can't go around the puck but must instead go throught the middle restrictor. The puck is then pulled down to the bottom of the basket to seal the o-ring with the 2 small wood screws on each side. The restrictor itself is simply a capped 1/4" fitting with a small hole drilled in the end. I used a 1/32" drill to make the hole, but the water still runs a bit fast at about 20 sec for a 1.5 oz shot. I'd go smaller but that's the smallest drill Dremel makes.

So all I really need is that 50mm D x 15mm H soft plastic puck and I'm home free.

What I'm thinking now is that I'll find an old white polypropylene cutting board (melting point 160C), trace a circle on it around my brass dispersion disk and cut it out rough with a coping saw. I'll then mount it on my drill and turn it down to 50mm by spinning it on some 80 grit sandpaper. This material should be perfect to drill and tap. It should also accept the 2 wood screws, which I'll use to pull it down onto the innertube gasket I made as a backflush disk a month ago (in place of the Scace o-ring). I'm confident this will work well (and meet my no cost requirement).

Pictures coming soon!

[Endo]

Puck done.

Used the wife's white poly cutting board (I told her it's a drink holder).

Used my standard 2-1/8" (54mm) hole cutter I use to install doorknobs. A perfect 50mm ID fit!!! Unbelieveable. The rubber disc in the photo is my backflush disc I made out of an innertube. I'll put a 1/2" hole in the center and slip it between the poly disc and the basket so all the flow goes down through the center flow restrictor. I'll add 2 small wood screws through the bottom of the basket to pull it down tight (like on the real Scace).

Now I just need to wait for my $20 meter with 0.1C accuracy to arrive from Hong Kong, thread it through a small hole from the bottom, and I'm ready to rock!

Cutting board puck

IMG_1709 (480x320).jpg (37.63 KiB) Viewed 39632 times

[caf4brains]

Maybe you could fit a tealight holder in board so you could still do vegetable prep during a power failure.

[Endo]

Almost finished !

$10 and a 1 week wait and I now have my Hong Kong 0.1C Fluke replica installed.

Just need to seal a bit better where I put the t/c through the poly puck and then fasten the t/c end somehow so it can be position dead center.

Since the k-type thermocouples are only accuracte to 1.5C, (compared to 0.5C for the t-type), I'm not expecting miracles. I'll do a boiling calibration and then see how close I can get to the measurements I took with the real Scace.

DIY Scace (top)

IMG_1768 (480x316).jpg (56.63 KiB) Viewed 39587 times

DIY Scace (bottom)

IMG_1769 (480x320).jpg (53.8 KiB) Viewed 39587 times

[JohnB]

Why not buy a more accurate T/C from OMega? What is the point of testing if your equipment is that poor? Can you calibrate your meter to correct for inaccuracies?

[Endo]

The inaccuracy comes mostly from the t/c material. For K-type, about the best you can get is +/- 1.5C, but you get a higher top end range in return (not much use for the Scace though). I'd prefer a T-type for +/- 0.5C accuracy, but the t/c and meters are much more expensive.

You can read about the accuracy here:

http://www.microlink.co.uk/tctable.html

I wanted to see what I could do at the cheapest possible cost, but if I decide to spend more, I'll first consider getting the Omega HH506RA datalogger and some T-type t/cs.

http://www.omega.com/ppt/pptsc.asp?ref= ... Nav=teml06

It costs $165 and I'm not sure I want to spend that now.

[JohnB]

It really depends on the quality of the t/c you buy. I've had K type t/cs that were several degrees F off doing a boiling water calibration & others that were only several tenths of a degree F off. With a decent meter that allows offset corrections you can dial in your t/c to be pretty much spot on.

[jbb]

Almost finished !

how did this end up working out for you? i've come to accept that these sort of temp measurements are just part of what comes around now, and, at least for the vivaldi, the over the lip method is pretty half a**sed. so if this does the trick i should just go ahead and build one.

[Endo]

It worked pretty well for a while. But recently the cheapo meter died.

Since I'm now very happy with the calibration on my machine, I don't really use it much any longer. Otherwise, I'd invest in a nice thermo-meter with datalogging capability (like the $169 Omega HH506RA I referenced below).

Only problem I had with my DIY Scace was routing of the little T/C bead to the correct position over the hole (it always spins or straightens out). On the real Scace, they use a covered T/C bar that goes across the brass hole. This makes it more durable.

In general, the white cuttingboard puck, inner-tube gasket and brass restrictor arrangement worked VERY well and was super easy to make.