CTEK CTX Battery Sense

[Old Crow]

I installed a CTEK CTX Battery Sense gizmo in my Hawk's battery compartment yesterday to see what I can learn from it.

It's a $50 device you attach directly to a battery's terminals to monitor its state of charge (SoC). The device (the 'sender') collects data for retrieval via your phone (or tablet) running a free companion app (CTEK Battery Sense).

I know many Wander The West members run Bogart or Victron-brand monitors and I may eventually get to that. But at this point my needs are relatively simple (no solar, no fridge, single battery) so I wanted to try this.

I ran into a few small problems, the most significant being the size of the terminals on the wire-ends of the device. They're only 1/4" (M6 metric) so I made up a short length of wire with larger terminals. I got lucky here in that I had wire and terminal lugs leftover from my recent DC-DC charger install. I chose to make an extension rather than cut off and replace the old terminal in case I needed to return the device.

The other minor problem was finding the app on the App store. A search didn't find CTEK at all. And since I was using an iPad (Mini4), I had to change the filter to 'iPhone-only' for the search. (The iOS version of the Battery Sense app is an iPhone app). And even then a search for Battery Sense returned many other battery-related apps. I finally found it after scrolling down 22 lines of other apps. It downloaded and started up with no other issues.

When the app starts the first time, it asks for an 11-character identifier string from the sender unit. I had already zip-tied the sender in place so cut it loose to read the string. (Tip: the ID string is also printed on the outside of the box it came in.)

So on to the display:



Once the sender connected with my iPad, the app did something interesting. It put up a screen showing 80 amp-hours. I don't know if that's a default value or something it calculated. I entered the published amp-hour rating -- 105ah (and will discuss that further below).

On that same screen the app had an optional opportunity to add a photo. I ignored it at that point but later came back and took the photo you see using the app (after it asked for permission to use the camera).



So we see from the detail screen that the app is telling me my battery is 83 percent full. Yet it shows a voltage of 12.62, a voltage I typically think of as close to full. I know, for example, that the resting voltage will be something like that after any of my various battery chargers declare it full.

Offhand, I'm inclined to believe the 83% number. My battery is over seven years old and its days of being a 105 amp-hour battery are long gone.

The other interesting thing is it shows me a graph of the battery's state of charge over time. I'll be interested to see how that graph changes as I use and recharge the battery.

Other notes:

Voltage and temperature are not shown by default. I had to turn them on in the Tools screen. I also had to change Centigrade to Fahrenheit....



The sender communicates with my iPad over Bluetooth. I have the sender zip-tied above the battery and this morning I was happy to see I could get the battery's voltage reading from the driver's seat of the truck. What I DIDN'T expect was to be able to get a reading from inside my house (with the rig in the driveway). It must be approaching the limits of Bluetooth as it won't update with my home's window closed but will if I open the window.

I also like seeing the temperature in the battery compartment.


-Old Crow

 

[Bosque Bill]

That's very cool. Thanks for the post.

I'm guessing it would also work with two batteries in parallel. I'd have to confirm that the combined capacity is within the specs of the unit - I don't recall my batter size off the top of my head.

It would be handy to be able to read the camper battery voltage from inside the truck.

Regarding 83% - my battery tech specs state that 12.8v or higher is 100%, so your 83% for 12.62v is probably close to right. See also below.

The product sheet says the device is only for lead acid batteries - probably would be ok for AGM batteries as they are simply a type of lead acid, I believe.

Here is the product info on the CTEK site:
http://www.ctek.com/products/vehicle/ctx-battery-sense

Product sheet:
http://www.ctek.com/storage/ma/5fda06112093412cb1c6c5f9ae3b166b/b2ed8a8b0fda4d5780ec116630b50222/pdf/0E883A735FE8557769B66CE77E0C6FB8DC704085/CTX_BATTERY_SENSE-Productsheet-low-US-EN.pdf

From the FAQ:
http://www.ctek.com/help?q=ctx%20
CTX BATTERY SENSE: Why doesn’t the percentage show 100% when the voltage meter is higher than 12.8 V which is said to indicate a fully charged battery?
Measuring the voltage for estimating the charge status of a Lead Acid battery can only be done when the battery has been at rest for a certain period -sometimes more than 10 hours. This situation is likely to occur after charging the battery or after a ride with the car for instance. Battery Sence takes this into consderation when calculating the charge status.

 

[ETAV8R]

This looks pretty darn cool. I already use a CTEK D250S for my batteries and solar connection. Wonder if they connect and communicate together. Would be great to be able to pull it up on the phone and see all the given data for my system.

 

[Old Crow]

This looks pretty darn cool. I already use a CTEK D250S for my batteries and solar connection. Wonder if they connect and communicate together. Would be great to be able to pull it up on the phone and see all the given data for my system.

Funny you should mention that. I have the D250SA version of that charger connected to my battery and had a moment of panic about the Battery Sense sender unit and the D250SA working together.

After I checked things out initially I started the truck and watched the Battery Sense app and a separate digital voltage display climb up to charging voltage. All good! But after five minutes or so I saw the voltage falling. What the heck? Voltage fell below--- well below-- 14 volts.

My first thought was the sender unit was interfering with the D250SA. So I shut everything down and removed its wires from the battery. I started the truck again and was dismayed to see the voltage stay below 14 volts. Did I just screw up my expensive charger?

Then as I looked at that voltage figure, it struck me..... I was seeing float voltage! It was 13.7, not the classic 13.6 but that's what it had to be. I re-connected the sender.

It's too early to comment on what I'll see when I start the truck and the D250SA goes to work. I've been running the furnace today to draw down the battery before I do that.

 

[Old Crow]

Update:

I've not started the engine yet to see what happens when the battery starts charging. I want voltage to go lower first. But I have an interesting mystery going on at this point.

Here's the graph of activity since I connected the Battery Sense on November 15 through this morning....

((Note- On the app, I get the graph without the rest of the Detail screen (and a larger image), by turning the phone sideways))




This graph shows the Battery Sense first coming alive on November 15. I see an odd juke down at the vertical line of 16 Nov that seems curious given there was nothing going on. But the real action starts at the 17 Nov line. That's when I turned on the furnace.

Now remember, this graph is coming from draw only. The mystery, then, is why the graph is rising. I started out with a battery at 83% and by the end of the day the app told me the battery was at 93%.

At around 11 pm, I turned the furnace off as I was hesitant to let it run unattended overnight. This morning, I went out to restart the furnace and the app told me SoC was 78%.

.

(I don't have a screen-cap of the 93% at 11PM so the 7:31 one showing 92% will have to do)

I had spoken with CTEK North America tech support late Friday afternoon to ask whether my extension cable was a problem for the sender (no) so I sent him screen caps of the mystery Friday evening and this morning. But of course they're off for the weekend so I'll have to wait for a reply.

-more- (just some notes)

- The Overview (the one with the photo) is where you refresh. I believe refreshes will happen every five minutes automatically if the app is within range of the sender. Pulling down on the Overview screen also does a refresh.

- The voltage and temperature readings on the Detail screen are sometimes just dashes or show an 'updating' message. This happens when you're out of Bluetooth range but also sometimes (for a short time) while in range.

- While in range, the voltage changes every four seconds if it's changing. I noticed this while watching voltage fluctuate after the furnace turned on or off.

- Voltage and temperature readings are only available as instantaneous readings, i.e., you can't retrieve them later. However, if I see something one want to keep, I screen-capture it by simultaneously pressing the sleep/wake/power button on top of the iPad and the Home button at the bottom of the screen. That screen-cap is then available in Photos.

- It's easy to be fooled by a screen-cap of the detail screen showing battery percentage and instantaneous voltage. If the battery is under load, the voltage will be much lower than we'd expect from the State of Charge voltage charts we see on the web or manufacturer documentation. Those SoC charts do typically tell us they are for resting voltage and a standard battery temperature of 77 degrees Fahrenheit.

- The product sheet Bosque Bill posted the link to says data is tracked for three months. The FAQ for the Battery Sense can be downloaded from this page and tells us the sender collects data in five minute increments and its draw is 1 milliamp. Somewhere I saw the sender is good for up to 200 amp-hours but can't find it right now.

Edited to add--

Ah-- there it is. Clicking on Edit on the Detail screen shows the amp-hour figure I entered. Clicking on that brings up the screen showing you can select from 5 to 200 amp hours....



Edited to add--

The FAQ I mentioned above says the green color on the battery icon will turn to yellow when SoC falls below 58%. And from yellow to red when it falls below 35%.

 

[klahanie]

I'll take a stab at this...

Now remember, this graph is coming from draw only.

Not to split hairs but I thought this was monitoring SoC. The graph shows SoC increasing while the furnace was on which I wouldn't have expected. Two things I can think of, first the batt is "working" and it's internal temp prob increased. Second the furnace might have increased the ambient temp.

I don't know how temp affects the voltage reading but wonder if it is playing a part in what you are seeing.

 

[Old Crow]

Thanks, klahanie...

Yeah, I didn't word that well. It is indeed monitoring SoC, not draw. I was trying to say the SoC graph is responding only to current draw from the battery as there has been no charging. I've not started the truck since the day before I turned on the furnace. I don't have solar and the only time I have shorepower is when I run a cord through a tie-down door to plug in a standalone battery charger (and of course have not done that).

I agree there may be some impact from temperature but I can't imagine it's enough to counter the furnace's draw and show a net increase in SoC. It just seems very weird to me to have started today at 78% SoC (at 35 degrees) and now see 89% SoC (but at 55 degrees) and have had the furnace on all day.

 

[klahanie]

Thx OC, for not taking that the wrong way...

So I looked it up and for a given soc, voltage drops with temp. Using this calculator entering 12.425V and 35F calculates soc to 78%. Change the temp to 55F and soc drops to 75%.

So my theory is backwards, warming up should cause the reading to chart downwards.

Can you turn the temp monitoring/compensating off ?

 

[Old Crow]

I was out running around in the rig today and will post what I've learned since the last update soon (probably tomorrow). But I should first reply to klahanie's post....

Thx OC, for not taking that the wrong way...

So I looked it up and for a given soc, voltage drops with temp. Using this calculator entering 12.425V and 35F calculates soc to 78%. Change the temp to 55F and soc drops to 75%.

So my theory is backwards, warming up should cause the reading to chart downwards.

Can you turn the temp monitoring/compensating off ?

Thanks for the info, klahanie. I like your idea of playing with the calculator to explore how changing temperature affects SoC. It's interesting to see a 20-degree change in temperature is worth a few percentage points of SoC.

As to controlling temp monitoring, no, I don't have control over it. The sensor is in the sealed sender box and the app doesn't have any settings for temp control. I can turn off the display of temperature and voltage but that wouldn't turn off temperature sensing or use of temperature in the app's calculations of SoC.

 

[Old Crow]

Update--

Summary:
There's something wrong, perhaps with the Battery Sense, perhaps with my installation. SoC went up to 100% as the furnace continued to draw from the battery through Saturday evening. On Sunday, I decided to ignore the SoC display and charge the battery by driving around. The charging didn't go as expected at first. About five hours to float voltage.

Detail:
In post 7 I posted the status of the battery at 5 pm Saturday. Inexplicably, the app showed the battery 89% full after a day of running the furnace off a battery that showed an SoC of 78% that morning.

At the time it occurred to me that the percentage was still going up and I had a lot of time until my 11 PM cutoff for the day.

And sure enough-- the SoC went to 100 per cent...



So Sunday morning I wasn't sure what to do. I had intended to run the furnace until it drew the battery down to an SoC of 50% before starting the truck to recharge the battery. I was also hoping to validate the FAQ's assertion that the battery-icon color changes from green to yellow at 58% SoC.

The Sunday-morning reading said 96% SoC but a voltage reading of 12.19. Since the battery had been resting all night, that meant the SoC was actually somewhere in the 50% range (depending on what you believe about voltage-to-SoC charts) and it was time to recharge.

I started the truck remotely as I watched the digital voltage display. I was dismayed to see the voltage only rise to about 13.4 volts in five minutes. I had expected to see 14.7 volts within a minute or so (as I had seen before).

I then started driving and occasionally checking voltage via the app, occasionally taking a quick screen-capture. As I saw voltage not rising to expected levels, I thought maybe I should try re-booting the charger but decided to simply wait and see what happens. Once the charger reached 14 volts (after 30 minutes), I had hope it might be OK. It finally made it to 14.7-ish 45 minutes after starting.

I spent the next few hours taking a tour of the eastern shore of the Susquehanna, basically just stopping at all the bike-trail and fishing-access areas along the river. Then I headed up to the state game lands and St. Anthony's Wilderness above Harrisburg, again just looking around.

I finally saw voltage drop toward float-voltage levels about five hours after I first started the truck at home.

This is what the display looked like that evening (a few hours after I turned the truck off)...



I didn't mention it earlier but the Battery Sense app is also incorrect on the time it shows on the graph. Note, for instance, that the screen-cap taken at 1733 on Sunday, November 19th shows the line extending into Monday, November 20th.

The other thing I'll mention is that I do like seeing the D250SA charging at 14.7-ish voltages (they were actually more like 15.0-ish yesterday due to temperature compensation by the charger) while the alternator is doing 13.5-ish....(The photos were taken within minutes of each other. The voltmeter is on the truck's starting battery).

 

[Vic Harder]

Does this device measure current flow at all? If not, how can it tell SOC? As you have seen, simple voltage readings are not enough. Trimetric and Victron use a shunt to measure the flow in and out, and use some calculations to guess at SOC, since they have to assume how efficient the battery is at recharging, given its age, etc.

 

[Old Crow]

Does this device measure current flow at all? If not, how can it tell SOC? As you have seen, simple voltage readings are not enough. Trimetric and Victron use a shunt to measure the flow in and out, and use some calculations to guess at SOC, since they have to assume how efficient the battery is at recharging, given its age, etc.

Thanks for your reply, Vic.

If I understand correctly, the sender would have to connect in series or have some type of current-clamp (to get amps by induction) so, no, it doesn't measure current flow. Also, I don't see any settings related to current flow in the app or mentions of it in the documentation.

I don't have any detail on how CTEK purports to calculate State of Charge. The only clue I've seen is the response to FAQ #19: "Battery Sense measures the battery's voltage and calculates its state of charge (SoC). The calculation takes account of charging and discharging, and also makes assumptions about the battery's characteristics."

All I can say at this point is I'm taking them at their word that they have figured out how to calculate an SoC and am seeing how it works for my application. It doesn't appear to be calculating SoC accurately so far but that could be an issue with my individual sensor, an incompatibility with my battery or some mistake I made during installation (like my cable extensions or the fact that I didn't connect the positive lead directly to the battery (since I didn't have a terminal lug that fit). We'll see.

 

[Old Crow]

Update-

The issues with the Battery Sense appear to be in the app according to my email exchange with tech support. I'm told a new release is due out in early 2018 so perhaps that will resolve some of the problems.

I was also offered a replacement sender and I've received it but I've not installed it yet.

Here's what the app shows about my battery as of this morning:



And if I turn the screen and pinch-in the graph I see what it has recorded since installation in mid-November:



Unfortunately, that graph appears to show another problem. As I looked at this morning's history graph (above), I noticed the double-vee after November 26th. That double-vee happened on an overnight camping trip. It represents the furnace running that evening and then again the next morning. It shows the battery was drawn down to something like 75% SOC.

But that's not how I remember it. So I went back to a screen-cap of the graph on December 4th. That one shows SOC went down to 50%.

 

[PaulT]

Your graph implies that their SOC data source is voltage from your description of the circumstances. While a relatively large current draw is occurring in the absence of a charging source, the voltage will be pulled down below the resting voltage and will rebound somewhat when the load goes away.
I've noticed similar behavior on the factory Zamp charge controller voltage display. Voltage before fridge turned on was 12.7 which dropped as low as 12.4 volts after running a while. Bounced up to 12.5 as fridge shut off then rose to 12.6v after some time had passed.

That's why the better battery monitors count the amp hours consumed or returned by charging using a shunt.
Paul

 

[rando]

I don't think those dips you are seeing are necessarily a problem - just am inherent characteristic of trying to determine SOC from battery voltage alone. Based on what you have shown here, it appears that this device measures the battery voltage, then uses a look up table like this to estimate the SOC:

When your furnace runs, it draws some current, causing the voltage at the battery terminals to drop a little, which in turn causes the CTEK to look up a new lower SOC.

This still seems like a useful device, with the caveat that the SOC reading is only valid under a very limited set of circumstances. No firmware upgrade can fix this.

 

[Vic Harder]

Great info from Paul and rando...

I love being on a forum with smart, kind people! Have you noticed the absence of flames? ;-) Love it.

I am grateful this Christmas to be part of this bigger family! I hope you all have warmth on your toes (fireside chats!) and in your hearts over the next few days!

 

[Old Crow]

Thanks for your insights, fellas. Sorry I'm slow to reply. I had some errands to run right after posting and then of course got involved in family Christmas activities. I hope everyone's enjoying the holidays!

Anyway, one of the reasons I've held out some hope that the CTEK Battery Sense would do a reasonable job of SOC calculation despite the fact that it doesn't use a shunt is the Balmar Smartguage. It appears to be a quality product. I base that on Balmar's reputation in the marine industry and on this in-depth review by Rod Collins (aka 'RC ' of MarineHowTo , aka 'Maine Sail' in some sailing contexts).

Now please understand that I'm not saying the Balmar product would be better than a Victron monitor for our campers. I'm merely saying the Balmar product appears to show it's possible to calculate SOC without a shunt, i.e., without counting coulombs.

 

[rando]

My understanding is that the Balmar Smartgauge is actually using a technique called pulsed impedance spectroscopy - which measures the internal resistance of the battery by applying small loads at various frequencies and measuring the resultant change in voltage. For lead acid battery this is fairly well correlated with SOC:

source: http://www.batteryuniversity.com/images/partone-22f.jpg

This seems to work pretty well under most circumstances, and has the big advantage that it is an absolute measurement (no need to integrate the past usage of the battery to arrive at a result) and accounts for battery aging. However, I think it is bad form that Balmar does not provide a 'theory of operation' for their gauges, so we can't know for sure how it works or what it's limitations are.

In order to make these measurements, Balmar requires some very specific connections between their monitor and the battery, and some very careful measurements of the battery resistance, which is likely why the Balmar is quite expensive. From what we have seen, it looks like the CTEK is just measuring voltage - which explains what you are seeing.

 

[rando]

FYI here is a similar device from Deltran/Battery Tender that is ridiculously cheap:
https://www.autozone.com/test-scan-and-specialty-tools/alternator-and-battery-tester/battery-tender-wireless-battery-status-monitor/240325_0_0

It is wifi, not bluetooth, but has the slight advantage that when your camper is at home you can check your battery levels from anywhere.

 

[Old Crow]

rando-

Thanks for the term 'pulsed impedance spectroscopy'. A search of that term and Balmar Smartguage yields interesting threads on what it may be doing (particularly post #42 by the developer in this thread).

I imagine you're right about the CTEK not calculating SoC in the same manner. It's unfortunate that neither Balmar nor CTEK speak in very specific terms as to how their algorithms work.

I imagine the situation here is the much more expensive Balmar unit is a serious attempt at calculating an SoC while a boat is being used while the CTEK unit is more casual-- perhaps intended as an easy way to check the battery on your stored motorcycle or car without having to take off the cover or dig out the multimeter.

I do get value out of it in that I love being able to get the instantaneous voltage and temperature on my iPad from the truck cab, the driveway, even my bedroom window. I had hoped the notification feature might be helpful but it appears that notification only happens when the battery falls to 35% and it's not adjustable.

It will be interesting to see if anything changes with the new app release. And of course I still have to try the replacement sender.

The Deltran wireless device is also interesting. I'd be trying that out if I could figure out how to get wi-fi at the storage unit where I keep my motor-toys and their battery maintainers.