Saturday, October 20, 2018

LiPo batteries for Nerf Modulus ECS-10 blaster

I have an ECS-10 in my basement, but it's for an Xmas present.  So I don't want to unpack and measure it (yet).

Note: this information applies, in a limited way, to the Stryfe.  The Modulus ECS-10 has additional extra space under the battery tray; the Stryfe (apparently) does not.   Stryfe owners have the option to buy an extended (deeper) battery cover - Modulus owners do not.

Also note this assumes you remove the stock ECS battery tray completely.

See this video for a good look at ECS-10 disassembly and internals.  Also see this video for how to wire a plug.

If you don't understand the battery specs below, see this video.


Voltage: 7.4V (2S) for stock motors; 11.1V (3S) for some upgraded motors.  If in doubt, go 2S.

Current:  Higher C rate is better.  But unless you are truly hardcore, anything over 25C is likely fine.

I have to mention many people get C discharge rate wrong.  The discharge rate is the number times the capacity of the battery.  That is "20C" on a 1500 mAh cell is 20 * 1500 mAh = 30A.  20C is not 20A!

This means that batteries with bigger capacities will tend to have lower C rates, but can still deliver high currents.

Capacity:  Well, bigger is better, duh.  Price is proportional, of course.  No one knows how many shots/mAh you can expect.

Size:  The ECS-10 takes four AA batteries in a tray which is clearly a simple boxSuch a holder is typically 109 x 31 x 15mm minimum

Length is fixed by the blaster.  Say, 110mm limit.

Width: Batteries are max 14.5mm in diameter (call it 15) plus three 1mm walls, so call it 33mm
Maybe another 1-2mm by shaving the edges of the blaster, so 35 max?.

Depth is uncertain, 15mm for a battery, 1mm wall, plus another couple in the blaster.  Say, 18mm? 

Tons of room for plugs and wiring around that area, no need to add any space for those.

Add a little margin, we have 109mm x 33mm x 16mm(-ish).  I've seen someone fit a 34mm wide pack in without difficulty, but their blaster had shaved edges around the battery tray location.  35 miiiight just fit.

Plug:  As you install your own plug into the Modulus, it's dealers choice.  Make sure the balancing plug is compatible with your charger!

Some prefer XT60 or T-plug over JST, arguing the JST is not rated for the current.  However, the limit is really the temperature rating of the connector, which is 85*C and well beyond most hobby applications.

My experience with the WLToys L959 buggy indicates that JSTs do fine and do not get even barely warm, so I'm confident JST is OK.  If concerned, go with T-plug / Deans plug (50A) or XT60 (60A).

Wire:  All cells will come with appropriate wire.  The wires are very short and therefore very low resistance, and highly unlikely to heat up on anything other than a dead short.  Experience with JST-equipped 1500 mAh packs in L959 bear this out.

For the same reasons, I do not believe that upgraded wire in the Nerf is required regardless of what you do, but I've not tested it personally.



The following batteries have been reported to fit the Modulus ECS-10:

2S:
Turnigy 1000 mAh 2S 30C (73 x 34 x 15, 30A peak) [Ref, Link (Hobbyking)]
Turnigy nano-tech 1800 2S 25C (103 x 35 x 13, 45A peak) [Ref, Link (Hobbyking)]
Turnigy Graphene 950 mAh 2S 65C (72 x 26 x 15, 58A peak) [Ref, Ref 2, Link (Hobbyking), Link (Amazon)]
Turnigy 800 mAh 2S 20C (55 x 30 x 18, 16A peak)

Also several Zippy Compact cells, but I don't have model numbers.

3S:
Turnigy Graphene 1300 mAh 3S 45C (75 x 36 x 27, 58A peak) [Ref, Link (Hobbyking)]  <- May be too deep, YMMV.


This would make the peak size about 108 x 34 (or 35) x 18mm.   (I'm leaving out the 3S cell, above, as I don't believe a 27mm deep cell will fit a Modulus.)


For cheaper options, look at NFStrike or Banggood.  Note Banggood can have problems shipping due to upgraded regulations on shipping lithium batteries.


I wanted my packs to do double duty in my L959 and L969 RC vehicles.  Maximum dimensions for the RCs is 84 x 35 x 18mm.  I prefer T plug (a.k.a. Deans plug) because my chargers and L969 are T-plug. 

So my packs have to be 84 x 34 x 18mm, max.

HobbyKing has a selector that will let you search their lipo cells by size - go hereHere are my results using the dimensions above.

My first choices (not yet actually verified to fit my ECS-10s):

Zop 1500 mAh from Banggood.  (Perfect, but they can't currently ship to Canada.)
Turnigy nano-tech 1300 mAh 2S 20C from HobbyKing.

1400 mAh from NFStrike, which is half the cost of the Turnigy.
Other 1400 mAh from NFStrike.

1300 mAh seems to be the limit for brand-name packs.  The Zop and NFStrike cells ratings are probably exaggerated, but probably not enough so I'd notice.

I will likely buy the Turnigy nano-tech 1300 mAh 2S 20C from HobbyKing.  Mostly because it is 34mm and most likely to fit the ECS-10.  Other packs are 35mm and I just don't know if it will fit.

Deals:

-  NFStrike usually has a 10% off coupon.

-  Hobbyking has $50 free ship to Canada, but only from USA warehouse (not Global Warehouse).  That can save you $20-$30 overall, and may make a more expensive battery cheaper than a less expensive battery.






Wednesday, August 22, 2018

First impressions: Seidio Surface Combo vs. Encased Rebel on Galaxy S9+

Big fan of Seidio cases, but not that thrilled with the Surface for the Samsung Galaxy S9 Plus.  I decided to try an Encased Rebel instead.

Construction:  Both two part outer hard shells with one-part inner liners.  If you don't like this, tough - it seems it's the only option these days for the larger phones.  Seidio liner is a little thinner.

Size:  Seidio is a little slimmer.  The holsters are not interchangeable due to the difference in thickness.

Fit:  Seidio case halves slide on easily - so easily, in fact, that the bottom half has detached and started to come off the case five times already.  The Encased has more friction, making it more of a job to take on and off, but no risk it will spontaneously come apart.

(Yes, the Seidio does have a "latch" to keep the case halves together.  It doesn't work well.  I was considering using double-sided tape to keep it from falling apart.)


Appearance:  Seidio is "smooth" and "soft-touch" plastic, and looks smeary.  Encased Rebel is neither and looks cleaner (so far).  Seidio is very slightly spiffier with the silver kickstand.

Grip:  Seidio "soft-touch" plastic is low friction and hard to grip.  This was is the first thing that put me off the Surface case as I always feel like I'm going to drop it.  It's hard to describe but it is definitely low friction.

The Encased is barely better on the back with a slightly textured feel, but includes rib-n-groove sides that greatly increase the grippability of the phone.  Overall a good improvement, IMHO.

Drop protection:  Didn't test.  I expect they're nearly identical.

Holster:  Seidio holster is a touch easier to lock/unlock, but I don't expect it will make a difference long term.  Seidio clip is more aggressive and less likely to accidentally come off, Encased less so and more likely to accidentally detach.  Both are about the same size and feel equally durable.

Buttons:  Seidio definitely has the edge here - buttons are very easy to locate and push.  This is both a pro and a con as I often find myself pushing them unintentionally, but they feel great.

Encased Rebel buttons doesn't have the same feel or ease of location, and require more force to push.  It's more than tolerable, but not as good as the Seidio.  (Less chance of pushing the stupid Bixby button, though.)

Port cutouts:  Both have big fat cutouts, so a draw.

Kickstand:  Surface has non-adjustable magnetic kickstand on the phone.  Rebel uses the holster as the stand.  I hardly use the kickstand so I don't care much either way.


Price:  Encased tends to be a little bit less expensive.  YMMV.


I have both cases now so time will tell if the relatively poor clip of the Encased holster will be enough to make me switch back.  But the GS9+ with the Rebel immediately feels better in my hand than the Surface ever did, so for grip the Rebel is the winner.
In retrospect I should have purchased a Seidio Dilex, but it's clunkier-looking and lacks a kickstand.  (I know I said I don't use the stand, but I had one on my Galaxy S5 and I hate to go backwards.)

Besides, the photos of the Dilex make me think it uses the same plastic finish as the Surface, meaning it's not any more grippable.  As that's kind of the problem, I'm not willing to try out the Dilex.




Sunday, August 12, 2018

Bosch SHP65T56UC dishwasher won't work, but drain pump constantly cycles

Problem:  For no visible reason, and suddenly, a Bosch dishwasher (such as a 500, 600 or 800 series, our model SHP65T56UC) starts with the following:

-  No indicator lights or display
-  Does not respond to buttons
-  No beeps or lights of any kind

and, most importantly:

-  The drain pump turns on continuously, or cycles on and off over and over / periodically, with or without the door open, even though the machine is not on and has no water in the tub.

This may happen right after plunging or otherwise clearing a clogged sink drain.

Power cycling at the breaker does not help.  Pouring water into the DW to try to fool it also does not help, though the pump will happily drain whatever your pour in.


Potential solution:  A small (possibly tiny) amount of water has leaked in to the bottom of the dishwasher, tripping the flood protection.

Basically, the bottom of the dishwasher is the last place to catch a leak.  So there is a small float, made of Styrofoam, built in to the bottom of the unit.  If water leaks from the wash tub into the bottom of the machine, it will shut off and run the drain in an effort to prevent flooding.

I know the description above does not really help very much.  I read similar descriptions on other sites but did not know what I was looking for.  So here is some pictorial help.


The first thing most DIYers may want to do is to remove the kick plate at the front of the machine.  Good, do that. 

But all you will see is an up/down leveling screw and an otherwise impenetrable wall.  You can't get in that way - you have to pull the unit out of the counter.  Bummer.

Fortunately, this is not (usually) too hard.  Installations vary, but many under-counter installs will use simple mounting brackets at the top, like this:


Unscrew any brackets (often 2), then pull out the dishwasher.  (It's actually fairly light, so there is little worry it will hurt your back or scratch your floor.)

On some installations, the top sound insulation (matting) will get caught up on the counter.  You can just ignore this, to an extent, and pull the DW out anyway (I did) or you can go to the bottom and crank the DW all the way down.  If you do this, be sure to level it again after pushing it back in to the counter.

Also, when pulling the DW out, keep in mind that the drain hose and/or electrical cables probably run into an adjoining cabinet.  Make sure there is some slack there so they don't end up getting pulled out of the DW machine itself.  When you shove it back in, be sure the pull the slack hose back through to the opposite cabinet so it doesn't get in the way of the DW itself.


After about  8-12 inches, you will be able to get to the left-hand corner of the DW.  There is no panel there - it is fully open, covered only by the soundproof matting.  Fold the matting up and out of the way:



My handy-dandy flashlight is making the otherworldly glow there.

Get down there and look inside, and you will see this:






You are staring at the white, plastic bottom floor of the DW.  

See that  round area in the middle?  That's the overflow "tray".  It's nothing more than a circular depression in the bottom of the DW. 

Look closely and you will see the the styrofoam "float" resting inside.  Stick your hand in there, you can feel it.  The plasticky box on top is the float switch.



So: if water leaks into the bottom of the DW, it collects under the styrofoam thingy, which rises.  The switch activates, shuts down the DW, and keeps the drain pump cycling until things dry out.

Chances are, when you stuck your hand in there, you felt a little bit of water with your fingertips. That's tripped the flood protection and shut down your DW.
 


Now, if you didn't feel any water, and don't see any, then one of two things has likely happened:

-  The float switch has gotten stuck in the "up" position; and/or
-  The float itself has gotten stuck in the "up" position.

In either case, you will need to replace the offending part(s), or otherwise figure out how to get them functioning properly. 



In my case, there was only a tiny amount of water around the float - just enough to trip the switch.  Cleaning it up with a couple of wads of paper towels was all that was needed to resurrect the DW back to working condition.

I was surprised by this, as I was expecting at least a half-inch of water filling up the DW bottom in order for the flood protection to trip.  But not so.

 In fact, there was so little water in there that I would have to say this was a premature trip.  There was less than a millimeter of water, and not even enough to fully surround the float - barely a Dixie cup full, if I had to guess.  To my mind this is not enough water to justify disabling the DW.

On the other hand, the bottom of the DW should be bone dry all the time.  So I can't really fault Bosch for being conservative here.  I wish they hadn't been, since it cost me time, but I can see their side of it.

I can also see why they felt unable to put some kind of error message on the unit display - they wanted to shut everything possible down to prevent electrical problems due to possible flooding.  But I'm still rightfully annoyed all to hell by that, and I'm the customer, dammit, so I'll be annoyed if I want to.




A lot of people report that this problem occurs when their sink becomes clogged and they end up plunging it to clear it.  I'm not sure exactly what happens that makes the DW leak water, but that is undoubtedly what occurred in our case.  

I've cleared the sink drain before, but this one was unusually difficult, and it required a lot more work than usual - and that is saying something, for our kitchen sink.  It's been a mess ever since we got the new DW, so perhaps the DW is dropping more junk into it than our old model did.

I also - wisely or not - use the "5-second plumber" pressurized gas canisters a lot, since clearing our sinks seems virtually impossible without them.  Somehow, a tiny bit of water leaked into the base of the machine during this last round, even though it wasn't an issue the previous three bouts.

Now, sure, my (excessive?) efforts to clear the sinks could have damaged something - there could be a crack in the internal DW drain hose, or something else I broke.  So if the DW shuts down again, I'll probably call a guy to find and fix the leak.  But there is no need to do so if all the repairman has to do is sop up a tiny bit of water due to a backed-up sink drain.



Sunday, July 29, 2018

Setting up Letsencrypt certificate on Synology DSM 6.0

HTTPS is already the de facto standard.  With Google Chrome now shaming non-HTTPS websites, little guys/gals like me running legacy insecure web servers are now on the wrong side of the line and risk losing business / traffic.

Fortunately, Synology DSM 6.x has a baked-in solution using the free certificate services from Let's Encrypt.  It looks scary at first, but it's very easy once you know all the steps. 

Things to know:
-  It's perfectly free, no cost.
 -  It works with dynamic DNS services.  You do not have to renew the security certificate if your IP changes.
-  It works with virtual hosts.
-  It is maintenance-free, as it auto-renews itself.
-  DON'T read the "Get Started" information at Let's Encrypt.  It will just confuse the hell out of you. 
-  Did I mention it is free?



This article assumed you have Web Station set up and your website(s) are running correctly.  If not, you obviously have other things to fix before you get around to HTTPS support, so go away and fix it.


It also assumes you have DSM 6.x.  If you are like I was, and still running DSM 5.x, you will have to upgrade in order for the security certificate to auto-renew.  Otherwise you will have to manually renew it every 90 days, which sucks.

I've gotten very conservative in my old age, as I've seen waaaay too many "upgrades" cause waaaay more trouble than they were ever worth.  So the prospect of accidentally borking my main file server frankly terrifies me.  However - and in my experience - upgrading from DSM 5.x to DSM 6.x was painless, so woman up and do it.

Note, however, that in 6.x, Web Station moved from within the Control Panel to its own application.  Look for it in the app box at the top left-hand corner of the web interface.


How to enable HTTPS on your Synology Web Station website(s):

1.  Make sure port 443 is open in your firewall.  (Somehow, nobody mentions this - I guess everyone assumes everyone already knows?)

Depending on your setup, the firewall may be in the Synology, or it may be in your router.  It will be in the same place where you have port 80 open for your website(s) to work in the first place.

While you're at it, disable any port forwarding you're not still using.  I had some old ports open from some old PVR applications that I stopped using ages ago.  Typically, all you need open is 80 (for HTTP) and 443 (for HTTPS).

2.  Go to Control Panel / Security / Certificate.  Click "Add", then "Add new certificate".


3.  Choose "Get a certificate from Let's Encrypt". 

4.  Fill in the blanks:
  Domain name = the domain name of your website [i.e. titam.com]
  Email = the contact info for that website  [me@titam.com]
  Subject Alternative Name = the www server name  [www.titam.com]

Note: if you don't put the "www" version of your domain in Subject Alternative Name, the certificate won't cover the web server name, and you'll get verification errors.

 5.  Click "Apply".

6.  The screen will return to "certificate" with the new certificate listed.

7.  Click "Configure".

8.  For the target domain (titam.com) change the "Certificate" from "synology.com" to the new certificate for that domain (also titam.com).

9.  Certificates are tied to the domain name.  So if you have multiple (virtual) hosts on different domains, repeat the above with each one of your vhosts, so they each have their own Letsencrypt certificate.

If you only have one domain / one website, and/or you don't know what "virtual hosts" are, you don't need to repeat anything.

10.  Use SSL Checker to ensure you get everything right. 

Common errors: 

-  Can't resolve:  Your dynamic DNS is borked.  Fix it.

-  Port error:  You didn't open port 443 in your firewall(s).  Go fix it.

-  Untrusted certificate:  You forgot to click "Configure" and change the domain name setting from the default Synology certificate to your new certificate.  Go fix it.

-  Unlisted / Incorrect hostname:  You forgot to list the "www" version of your website in Subject Alternative Name.  Restart the process with "Replace an existing certificate" and fix it.

Done.  Ta-da!

Kudos to Synology for baking this directly into DSM 6.x.  People often whine they are not user-friendly, but if this isn't user-friendly, then what is?  I mean, come on - like 10 clicks and you're done.  Give them some credit.


Wednesday, July 11, 2018

Cardo Freecomm1 volume levels

I still get to ride occasionally.  As I was renting a bike with a GPS, I wanted in-helmet audio.

Reviews on all in-helmet Bluetooth devices are mixed, no doubt because different helmets, bikes, and riding styles all affect how we perceive the volume levels.

I ride exclusively with 32 NRR earplugs in.  I can't understand how people ride without earplugs; seems to me you'd go deaf very soon.

I also have an "older" Uvex helmet I purchased in Germany.  (My HJC was killing me on my rental supermoto, which has much more wind exposure - the wind was pushing the front edge of the helmet into my forehead.)  Uvex has long since stopped making motorcycle helmets, but my helmet is still A-OK.

The Uvex includes a motocross peak (sun shade), which I always have on - even on the highway.  It adds some noise.

My Uvex has no speaker cutouts, so I had to make some.  This would place the speakers right on the shell, but I used the Cardo spacing pads to move them closer to my ears. 

Owing to the way the liner works, the helmet liner covers the speakers.

Finally, I was using the system on a touring bike with a BMW Motorrad GPS unit.  The bike had a windscreen but not a fully effective one.

This all makes for a situation pretty close to worst-case for an in-helmet audio system:

-  High-rated earplugs
-  Relatively noisy helmet
-  Non-optimal speaker placement
-  Speakers covered by helmet liner
-  Significant wind noise at higher speeds

I am happy to report that the system was audible and understandable even under these conditions at highway speeds (~ 65 MPH).

However, be aware that the audio was not great.  Voice prompts were audible, but distorted.  In a few cases they were difficult to understand.  However, they were loud enough to direct attention to the GPS screen so I didn't miss a turn.

At lower speeds, the system was just fine.  I was able to hear and understand the GPS without difficulty.

I imagine this setup would suck for music, as the audio quality will not be adequate.  But one is unlikely to expect earplugs and perfect audio quality to go together.

From what tiny experience I have had with the system without earplugs - which has exclusively been when I am NOT riding - the system appears loud enough to listen to music if you don't wear earplugs.  I would not recommend this since the noise levels are likely to damage your hearing, but to each his/her own.

I am also unsure about holding a conversation, at least at highway speeds.  My Freecomm1 has no two-way audio because I don't need it, so I can't test it.  However, I believe it should be possible to hear a riding companion at slightly lower speeds (< 50 MPH) even with 32 NRR earplugs in.

Hopefully this helps someone who is worried about using the Cardo Freecomm units with earplugs.

Wednesday, May 23, 2018

True 3-way occupancy / vacancy motion sensing wall switches

Simple problem:  Want motion detection (occupancy) at both ends of a hallway or stairwell.

Solution:  an amazingly large steaming hot mess!

-----------------------------

[Update]:  OK, so I've realized that the main reason why there are relatively few residential application for motion switches is: they don't save you money.

You can read the details here.  The original reviewer makes the mistake of confusing a 40W LED bulb (320 mA draw) with a 40W equivalent LED bulb (actually 9W, 72 mA draw), but otherwise the concept is sound.

The reality is that unless you are switching large lighting loads (150W-ish or more) and/or in a commercial environment where lights might get left on 24/7, adding a motion sensor makes no sense.  The "smart, energy-saving switches" will consume MORE energy than if you accidentally left the bathroom lights on an extra 15-20 hours per week, which seems rather unlikely for most situations.

This also means there is no payback period for smart switches.  They end up costing you when you buy them, and costing more every year in the energy they consume.  

There are still tons of reasons to buy motion sensors for the home.  There is no need to struggle carrying loads, or tripping down dark stairs.  Those things may easily be worth $5-$10 per year in energy costs, especially when compared to medical costs owing to an accident.  It's cheap insurance.  Ditto dimmers, which are just as much for comfort rather than energy saving.

But if you really want to save energy - after upgrading to LED bulbs, of course - the best way is to get up and turn off those LED lights whenever convenient, using a good old-fashioned $2 mechanical switch.  Everything will be off - totally off - and consuming zero energy, rather than consuming small amounts continuously to be "smart".

This rather puts California's Title 24 in doubt, at least when it comes to residential applications.  But whatever.

-----------------------------

As one reviewer commented, I have difficulty visualizing a situation where you don't want two motion sensors (at the end of a hallway, opposite sides of an L-shaped room, etc.)  Yet the solution is surprisingly hard to find.

Note:  "Occupancy" means lights turn ON and OFF automatically.  "Vacancy" means lights turn OFF automatically, but have to be manually turned on first.

Lutron:  Lutron advertises various "multi-location" Maestro Sensor switches.  Most models start with "MS", and most do not need a neutral wire.  Awesome, right?

The problem here is that you can only install ONE motion sensor, and the rest have to be "ordinary" 3-way switches.  Not what we want.

One enterprising soul did manage to make two Lutron MS motion switches work together in a "3-way" application, but it does not function 100% normally.  In a normal 3-way, ONE switch must be off to turn off the lights - in his setup, BOTH switches must be off to turn off the lights.

This is not a big deal (perhaps) since both lights are automatic, but may not suit everyone.  If it does, this may be the only no-neutral-wire installation that is possible.

You also must have a ground wire, and the switch is reportedly relatively loud.



Leviton:  Use the IPS15-1LZ (switch) and IPV0R-1LZ (remote sensor) together.  The remote sensor works with the main motion sensing switch to do true 3-way motion sensing.

The problem here is that both of these units need a neutral wire in the wall box in order to function.  Which not everyone will have.

It may be possible to substitute the green / bare ground wire for the required white neutral wire.  I don't recommend this because the ground wire is not supposed to be used in this manner, but it will likely work.  (The Maestro series actually does this on purpose.)

The other available Leviton switches seem to require one of the switches to be a "regular" 3-way mechanical switch.


Enerlites:  The DWOS-3R is the only multi-location switch.  It does not seem to be rated for use with LEDs, and it is unclear if you can actually use two switches controlling the same load.


Eaton:  The OS310U-W-K is not rated for LEDs, and there is no indication you can use two.


HubbellAccording to the instructions, you can use two ATP2000 series or two WS2000 series motion-sensing wall switches in tandem, and they will work properly.

Of course, they have high cost, limited availability, and non-standard wiring.  Other than that, they're perfect!

There are neutral-wire and no-neutral-required versions, and versions with and without a built-in nightlight.  Assuming no nightlight and no neutral, the model number is ATP2000W or WS2000W.


For reference, the "ATP" switches have some kind of auto-adjustment technology that lets them adjust their timeout period.  The "WS" series have an "ordinary", manually-set timeout period.  Both types are 3-way capable.

These switches are rated for "most" LED and CFL bulbs, and it does use a triac instead of a relay, so YMMV.

It is also not clear if the wiring is one-to-one with existing 3-way wiring.  Ordinary 3-way has line-hot to one switch, plus two travelers, while the Hubbel diagram shows line-hot going to both switches with one traveler.  You could re-purpose a traveler to a hot, if you know how.


This has already taken an amazing amount of time for what it is - doing motion in a stairwell - and I haven't even looked at my existing wiring yet.  I will update this if I learn anything more.


Friday, February 23, 2018

How to figure out which program is stealing Windows focus

I recently ran into an issue with losing focus while working.  An incredibly annoying problem for anyone that has experienced it.

Alt-Tab did not show anything useful and Alt-Space didn't either. 

Fortunately, I found the Adminscope Windows Focus Logger, which identified my problem in about 60 seconds. 

It's freeware, and there is no install.  Run the GUI and watch what happens.  Soooooo awesome.....

In my case, Razer Synapse kept stealing the focus. 

I suppose I should have clued in when I saw an "Unacceptable character" message pop up a few times - the font used was subtly different than a standard Windows dialog.  But the message was right over my Word navigation pane, leading me to believe the Word nav pane was somehow stealing focus.

Shutting down Synapse and restarting appears to have fixed it.  I think the problem only happens when you pop up Synapse, record a new macro or something, and then close it.  Somehow, it doesn't fully close.

I have a multimonitor system with multiple desktops (via Dexpot) so perhaps this is contributing.

Monday, January 8, 2018

Wiring the ecobee3 AUX+ and AUX- inputs

Note: this post has nothing to do with the dreaded "C-wire" that is required to run an ecobee.  If you're looking for that, go elsewhere.


One of the more frustrating items for a heating/cooling system is the sheer number of possible ways to set it up.  Sure, there are "standard" ways, but there are also always other ways.

My previous post on a missing G (fan) wire was not my only issue; in fact, it wasn't even the hardest issue.  The actual problem I had - the one I called ecobee about - was that my humidifier wasn't working.

For anyone who may not know, a whole-home furnace-mounteded humidifier is de rigueur  in Canada and the northern parts of the USA.  The cheapest and most common kind is the "evaporative" kind which uses hot furnace air to work - hence, it only runs when the furnace runs.

There are undoubtedly many ways to wire such a system, most of which the ecobee3 will support.  But it sure helps to know what the ecobee3 does.

ecobee3 supports two types: a "one-wire" accessory, and a "two-wire" accessory.  According to ecobee support, the thermostat works thusly:

-  In "one wire" mode, the thermostat supplies +24V to ACC+.

-  In "two wire" mode, the thermostat presents a dry contact across ACC+ and ACC-.

If you put this information together with how your system is currently wired, I bet you can figure out of how to wire and configure the ecobee to run your humidifier (or other accessory). 

If, on the other hand, you don't know what a "dry contact" is, this job is likely not for you.  Best get a pro to help.

Wiring an ecobee3 without the G wire (G-wire, fan wire)

Note: this post has nothing to do with the dreaded "C-wire" that is required to run an ecobee.  If you're looking for that, go elsewhere.


One of the more frustrating items for a heating/cooling system is the sheer number of possible ways to set it up.  Sure, there are "standard" ways, but there are also always other ways.

We had a new furnace installed not long ago.  However, there were not enough wires run to the thermostat to support the additional A/C system.  The cable had only enough wires for heat-only operation.

To make it work, the installers should have run a new cable, but they didn't.  Instead, they disconnected the G (fan) wire, and used it for Y (cold call) instead.  This led to an atypical installation that lacked a G (fan) wire at the thermostat.

Note it still worked.  The thermostat sent the W (heat) and Y (cold) signals, and the furnace controlled its own fan.  No worries, thanks to them thar new-fangled furnace, y'all.

However, when I installed the ecobee3, it said "To use the Y1 wire you must install the G wire".  Since I didn't have one, that was a bit of a poser.


When I described this to to ecobee tech support, they could not wrap their head around the idea that the furnace controlled its own fan.  They actually told me to rewire the entire furnace to include a G wire, otherwise they wouldn't help. 

Note I said "wouldn't" there.  I don't know if they could have because - basically - they just gave up.  The tech support rep said - and I quote - "I have been instructed not to proceed further."

Now, I don't particularly blame ecobee for this, owing to the problems noted at the top, but the workaround is actually built into the ecobee.  There was no reason to tell me to go hang; it was rather disappointing and reflected a lack of understanding of their own product.

Fortunately, in the first 5 minutes of being on hold, I figured it out for myself.  The actual correct workaround is as follows:

-  Run the ecobee setup.
-  Tap "Configure" to set wiring manually.
-  Tap "G" to tell the ecobee the G wire is connected (though it is not).
-  Later on, it will ask if the fan is to be controlled by the ecobee or the HVAC equipment.  Select HVAC.

That's it, it should work as it did before.

Now, OBVIOUSLY, you do not have manual control over the furnace fan at the thermostat with this setup.  But I didn't have that before, did I?  So no change there.

Also, I would not recommend anyone switching existing wires around to get a C wire (or any other wire) by sacrificing the G wire.  My furnace works this was because it was originally set up and configured to operate in the absence of a thermostat fan wire.  A furnace that has been set up otherwise, only to have the fan wire removed, is highly unlikely to function correctly afterwards.

This workaround applies to furnaces that have a manual fan switch installed somewhere else.  This is sometimes called the "summer fan" configuration, and often involves putting a manual switch on the furnace to provide for fan control that is independent of the thermostat.  Not a "normal" situation - unless your installer was too lazy to run new wiring when you got your additional cooling/heating unit(s) installed.

So, if you're unlucky enough to be missing a G wire at your thermostat, you can still make the ecobee3 work.  Hope this saves someone a few hours of their life.