Is there a consensus solution on erratic readings in artisan

[yukoncornelius]

Hi friends...

Building off a thread I created here
https://www.hukyforum.com/index.php/topic,1166.0.html

I want to isolate and solve one of the frustration points. The inference or erratic readings within artisan that is making roasting and logging seemingly impossible.

After searching I can't really see a consensus on a solution but I see multiple people having the problem.

Apparently the huky motor is not grounded properly? I suppose that's why the ac adapter literally sparks when I put it into a wall socket? Or I've read about laptops causing issues which apparently is pure coincidence or luck of the draw...

My laptop is running on battery. I had no issues with logging in roast logger with my hottop HTC tc4c connected via usb. The huky is connected via phidget 1048.

Can someone explain to me how to ground the roaster? I'm not an engineer or electrician... if there's a simple solution ..I'd love to hear it. It seems grounding the roaster should do the trick but I didn't see how?

Thanks to everyone for their help. Maybe this thread can be stickied in this forum section afterwards.

[yukoncornelius]

Could me having the huky on a cookie sheet cause issues?

Susan - there are lots of thread on this forum where individuals suggest fixes but do not update with resolution or what they actually did. Thanks for the help.

[monkey_lever]

Try to isolate and reproduce your erratic readings.

1) Without heating the huky turn on the drum motor and check your ROR on Artisan if your ROR is erratic the noise is coming from the motor.

2) Same with the  other parts.

You already mentioned that you are running your laptop on batteries. If you want to ground the huky you can attach an electrical cable to a ground in your home and use some of the screws at the back of the huky.

[edtbjon]

Apparently the huky motor is not grounded properly? I suppose that's why the ac adapter literally sparks when I put it into a wall socket? Or I've read about laptops causing issues which apparently is pure coincidence or luck of the draw...

There's a clue here... That spark you're talking about here. Unless you've seen that before with any apparatus you put into that socket (or any socket for that matter), either the AC adapter or the wall socket may be the culprit. The motor for the Huky is a 24V motor which doesn't need that much juice to begin with. The AC adapter is a very standard adapter, commonly found on the shelfs as replacement parts for inkjet printers. (I.e you may even be able to source a second hand one from some type of "printer salvage yard". )

[hankua]

Just for fun I bought a 8' grounding rod and clamp at HD; maybe $10 for both. Had some spare wire so skipped the heavy stuff; no more spikes although they were rare.

If you have hard soil, an impact hammer and a gadget works like magic. Otherwise a ladder, small sledge hammer and impact socket to fit would work. My soil is sandy, went in like butter. 

[yukoncornelius]

Great suggestions...

Hank I saw you prior posts about the grounding rod. Can't do that at my apartment. Seems rather over the top as well... you have a wire running from your huky to a rod buried in the ground outside? Dang...


I'm gonna try to find a different ac adapter. I also saw someone mention putting your finger on the chaff tray and see what happens. I may use a power strip so it can be more nearby the roaster if I connect a wire to that for the ground.

[hankua]

I tried a USB isolator and it was a waste of money. My roaster is on a cart next to a sliding door. The wire is laying outside on the patio and I drag it inside and connect it to either an alligator clip or use a wire nut.

I attached a short piece of wire to the Huky body for the grounding attachment. Most modern US outlets have a bare ground wire, and that's what I'm assuming people attach to. Does the little screw on the cover plate do the job, possibly? That's going to depend on how well your electrical panel is grounded in the first place and if all the connections are solid.

The idea of running a ground out of the outlet such as adding a leg to the ground wire nut sounds like a potential source of a short. I should add a longer screw to the cover plate and cinch it down with a washer and nut. Before that attach a ring connector/nut/washer to the head for a connection point or attach to the post with an alligator clip.

I'm used to measuring continuity with the tone on a multimeter, not sure how to actually measure in ohms how well something is grounded.

Personally I found the fan caused more spiking than the roaster motor; and the dimmer was the main culprit. Adding in the Variac cured it for the most part in addition to adding in the Phidget I already had.

Yes your charts are a little spiky, and no fun to post online; but that has absolutely nothing to do with how the roast turns out. I would try working on more control in the three phases; one at a time. (You can go back and mark EOD, I'd just use some artificial point at your choosing). I use 150c and never actually check as its a consistant point from roast to roast. The first phase is charged to EOD and you have leeway here, in charge temp./air/gas settings (4-6 minutes is a good range). Ramp to 1C @ 4-5 minutes; 1C to drop is your choice 2:00 is a good starting point.

"Plan you work and work your plan" is a good way to approach a roast profile. Lets just say for example your plan is to run a 4-4-2. That would be 4 minutes to EOD, 4 minute ramp to 1C, and 2 minute roast development time. You can also target a finish bean temperature as well.

If you can run a 4-4-2, your definitely past stage 1. Stage one being " how the heck do I make this thing work without screwing up?"

We have all been there, no big deal. 

[yukoncornelius]

I suppose that's why the ac adapter literally sparks

There's a clue here... That spark you're talking about here.

Have you tried plugging into a different socket?

Yes

[yukoncornelius]

Hank - thanks. I need to figure out steps of eliminating variables. When you guys are preheating...do your curves look smooth ...as well as ROR curve?  Do your thermocouple wires touch the roaster? MET seems hard to not have it touch the exhaust chute. 

I went home at lunch to do a quick test. I re-secured the wires in the phidget again. I let the huky preheat and tried to see if me turning on the motor did anything. I wasn't really able to see a difference in how the curves graphed. Gotta think of more tests, and or get an engineering friend or electrician.

Regarding the ugliness of my profiles... I'm not embarrassed to post them because I don't understanding roasting concepts. I've been logging roasts digitally through roastlogger for a good bit of time now. I'm familiar with dry/ramp/development time as % of roast time...etc.  The frustration occurs when you look at the photos I posted on page 2 of the other thread.  The photos of the curves as the roast was going on.... It literally makes roasting impossible as the data being presented to me is horribly noisy.

EDIT: an awesome huky owner sent me a simple guide to ground the roaster!  For the purpose of trying other solutions for the good of the community I also ordered a USB isolator from amazon. If it doesn't work I'll ship it back.

[yukoncornelius]

Well this could be funny... negative wire is magnetically attracted right? And negative wire should be connected to the G terminal... NOT numbered terminal...


Won't be back home till later but I think my wire polarity has been flip flopped.

[yukoncornelius]

I'm asking ... on the phidget terminal what does G and # represent.

If G is negative and # is positive... my wires are flipped.

[jyalpert]

I use this horrifying grounding solution.  Totally eliminated the infuriating spikes for me.

[yukoncornelius]

G = The wire that isn't red (sometimes yellow, sometimes white)
# = The red wire for whichever thermocouple you are attaching.

Every diagram I've seen of thermocouple wire shows the red wire as being negative.

Here's Phidgets' Thermocouple Primer
http://www.phidgets.com/docs/Thermocouple_Primer#Choosing_a_Thermocouple

And here's Phidgets' User Guide for the 1048.  Note the section Connecting the Pieces
http://www.phidgets.com/docs/1048_User_Guide#Connecting_the_Pieces

Thanks  - I saw the user guide earlier today.

The interesting thing is I would think G represents GROUND, which is traditionally a negative connection.  Thats why I initially didn't think twice about ok I have a red and black wire, I'm going to connect black to ground (G) as it almost always represents that. 

If G does represent ground and this post is true - https://www.hukyforum.com/index.php/topic,336.msg3307.html#msg3307

And the thermocouple which is attracted to a magnet is the negative one... that would represent red for me.  Meaning, my red one should be connected to G, and black should be connected to the number.   

I'll be home later where I can finally test this. 

[JavaBuzz]

This was mentioned earlier, but I concur with using the "negative is more magnetic" test to verify the negative/ground (G) wire. From what I recall, the TCs I received with my Huky did not follow normal wiring color codes.

You'll get extremely erratic readings if connected wrong.

A grounding issue can also definitely cause issues. If something else on the same circuit is not grounded properly, this can even cause issues. The Phidget is also much better at dealing with ground issues than a TC4/TC4C.

[yukoncornelius]

Ok couple updates for you all.

Although red wire definitely seemed more magnetic to me...when I changed one of the terminals to black going in # and red going in G (negative/ground) - I had a negative temperature reading in the phidget control panel. 

Then I attempted to ground the huky with jyalpert's grounding suggestion - I used picture hanging wire that I had laying around - I might try a thicker gauge wire tomorrow.   Result unfortunately is no different:

During preheat, I observed BIG swings up in ROR measurement when I turned the motor on, and the ROR would drop big time when I turned it off.  Then I tried dumping in 220g of miscellaneous greens.  Sure enough, jagged lines  on the actual BT curve.  When I turned the fan on, BIG change in ROR curve (or rather polygraph chart in my case lol).

This all tells me theres definitely interference or noise somewhere.

Whats odd to me is that javabuzz says a tc4c has even more issues with grounding.  Considering my hottop with tc4c was used in this exact room, on these exact outlets, and I had perfect curves, is really...really...really...really frustrating  .  However, my thermocouples (from omega engineering) had a grounded junction I believe, not sure if that made a difference.

Any other suggestions?