Low voltage warning, head light cutoff circuit design notes

[FinalImpact]

Over the past couple years several members have ran into trouble from low batteries do to prolonged idle time. Most have dual head lights but not all and if the bike is below 1800 RPM for too long the battery discharges to a point of not being able to continue operation.

My goal here is to discuss some design options to A) warn the rider the battery voltage is an issue and B) potentially drop one of the dual headlights to reduce the load. This last choice comes at a RISK and a good deal of consideration is required before chopping the headlight power. It could be the only safe bet is USER INTERVENTION vs automation.... The other option is too increase idle RPM with a simple solenoid but once again this comes at with RISK!!!

Initially I was looking for an Off The Self (OTS) solution but didn't find anything except for home and RV use which is overkill for our needs. I did find an Integrated Circuit (IC) that does most of the logic detection for us. We simply need to give it some limits to work with. The cost for the IC is close to $5.00 alone. The rest of the cost would be for a housing, resistors, diodes, FETs, and relay or switch assembly. So if not sourced correctly, the cost could easily be over $40 in materials alone. Is that worth it?

Here is one option and I'd guess there are many others. So if someone know of something, feel free to share.

*** ***
http://datasheets.maximintegrated.com/en/ds/MAX8211-MAX8212.pdf

From Manufacture:
Maxim’s MAX8211 and MAX8212 are CMOS micropower voltage detectors that warn microprocessors (μPs) of power failures. Each contains a comparator, a 1.5V bandgap reference, and an open-drain n-channel output driver. Two external resistors are used in conjunction with the internal reference to set the trip voltage to the desired level. A hysteresis output is also included, allowing the user to apply positive feedback for noise-free output switching.
The MAX8211 provides a 7mA current-limited output sink whenever the voltage applied to the threshold pin is less than the 1.5V internal reference. In the MAX8212, a voltage greater than 1.5V at the threshold pin turns the output stage on (no current limit).

*** ***

The basic schematic is very simple and needs only wires, resistors, and an LED to notify the rider there is a problem. Cost could be as little as $10.
The built in hysteresis (trip margin of detection) could use an electrolytic capacitor (RC time constant) to keep it latched for short period of time, but an actual circuit in addition to the basic one here would be best just so the light doesn't flash on/off when hovering at the trip point. Also to drive a relay (head light relay if someone wanted to drop one head light), would require a small signal amplifier like a MOS-FET device to drive the relay.
I just wanted you to see how simple this is with one basic component.

View attachment 49202

This is an open discussion for NEEDS that could be employed to help reduce dead batteries BUT SAFETY MUST COME FIRST! Please try to stay on topic. Thanks!

 

[FB400]

I am going ahead to wire a custom HID headlight harness over the weekend.

This will consist of 2 12volt feeds from the + battery to 2 fuses, 2 relays, separate relay triggers using both H7 and H4 headlight sockets. There will be 2 toggle switches incorporated to break the + trigger wire to the relay.

I hear your message about safety and having 2 headlights off. Both headlights off would be a dire emergency and to prevent the battery from going dead and the Check Engine Light coming on.

Test results prove this. I will put these results together and post to this thread. Randy, as you know, you got PM'd results and best avoidance of CEL and dead battery is both headlights off. Yamaha... geez.. missed the mark on designing this charging system - not enough wattage to support much of any useful mods.

 

[pulsar2]

It is a difficult decision whether to automatically switch one/both the lights off when on low battery. How about this:
1. Provide switches to turn on/off each lights independent of each other.
2. Provide LED indicator to indicate low batt. This has to be a highly power efficient design or else the indicator will just add extra load on the weak battery. This keeps blinking till battery recovers.
3. Provide LED indicator that flashes if one/both lights are manually switched off and the battery has recovered its charge.

If the proper connectors for the headlight assembly if available, it can be built as a plug-n-play device.

 

[FinalImpact]

It is a difficult decision whether to automatically switch one/both the lights off when on low battery. How about this:
1. Provide switches to turn on/off each lights independent of each other.
2. Provide LED indicator to indicate low batt. This has to be a highly power efficient design or else the indicator will just add extra load on the weak battery. This keeps blinking till battery recovers.
3. Provide LED indicator that flashes if one/both lights are manually switched off and the battery has recovered its charge.

If the proper connectors for the headlight assembly if available, it can be built as a plug-n-play device.

We're only talking a few milliamps for an indicator. Of course another option is to simply report the actual battery voltage and do as FB is working on and use toggle switches...

Mini 3-wire Volt Meter (0 - 99.9VDC) ID: 705 - $7.95 : Adafruit Industries, Unique & fun DIY electronics and kits $8 shipped! lol
View attachment 49201

 

[FinalImpact]

A 2 wire version of the same thing.... :thumbup:

https://www.adafruit.com/products/460

The downside, unless your consciously observing the meter its just techno bling and could go unnoticed!


DETAILS:
Dimensions: PCB: 30mm x 20mm
Display: 23mm x 14mm
Combined Depth: 11mm
Weight: 5.45g

Power specifications:
3.2V to 30V DC measurements
0.1V precision
3-4mA draw << 0.004 Amps is very small amount and not a burden to the system.
Green LED display
Use with positive voltages only!

 

[sxty8goats]

My problem is that I'm dumb. I turn the bike off via the kill switch. Occasionally I forget to turn the key to 'Off' after, especially if I'm stopping for a few minutes to have a convo with a friend.

I want a circuit that switches all power off if the kill switch is disengaged.

 

[FinalImpact]

My problem is that I'm dumb. I turn the bike off via the kill switch. Occasionally I forget to turn the key to 'Off' after, especially if I'm stopping for a few minutes to have a convo with a friend.

I want a circuit that switches all power off if the kill switch is disengaged.

That would be invasive and require a big relay to handle the bikes power. Perhaps it would better to find another way so someone doesn't take your bike! An alarm to remind you perhaps?? i.e. kill switch was activated AND the key is still on after 15 seconds ((BLARE OBNOXIOUS HORN)), etc . Just a thought.

Another option is we send a local member to your area and tape your kill switch ON until the habit is broken! :BLAA:

 

[sxty8goats]

Another option is we send a local member to your area and tape your kill switch ON until the habit is broken! :BLAA:

LOL Not sure that would help. I'd knock the tape loose in three or four rides.

 

[pookamatic]

A 2 wire version of the same thing.... :thumbup:

https://www.adafruit.com/products/460

The downside, unless your consciously observing the meter its just techno bling and could go unnoticed!


DETAILS:
Dimensions: PCB: 30mm x 20mm
Display: 23mm x 14mm
Combined Depth: 11mm
Weight: 5.45g

Power specifications:
3.2V to 30V DC measurements
0.1V precision
3-4mA draw << 0.004 Amps is very small amount and not a burden to the system.
Green LED display
Use with positive voltages only!

So besides the obvious range difference (which is a non-issue), what's the deal with a 2-wire vs a 3-wire? What's the third wire for??

I definitely want one. I'll handle the headlight manually with a toggle switch. 99.9% of the time, it's only an issue sitting in traffic. I'll be able to watch it and go down to one light when necessary.

 

[FinalImpact]

So besides the obvious range difference (which is a non-issue), what's the deal with a 2-wire vs a 3-wire? What's the third wire for??

I definitely want one. I'll handle the headlight manually with a toggle switch. 99.9% of the time, it's only an issue sitting in traffic. I'll be able to watch it and go down to one light when necessary.

The three wire has a fixed source (think in terms of seperate power supply). In our case we would be tracking the source (battery voltage) so we only need the two wires (+ & -).

We just need to find an enclosure. Maybe they offer one?? IDK

 

[pookamatic]

They offer a few, but none of them tickled my fancy.

This one is a bit big, but close to what I'd want:

Small Plastic Project Enclosure - Weatherproof with Clear Top ID: 903 - $9.95 : Adafruit Industries, Unique & fun DIY electronics and kits Details

 

[FB400]

subscribed to this one.

As I wrote early today am going ahead with creating a custom harness. This should meet needs at least for now being able to toggle off 1 or both headlights separately.

Thanks for all the input guys

FB

 

[FB400]

Setting out to custom wire a new HID harness
I Spent a few hours in the garage in high heat and humidity (call it mid 90's probably 80-90% humid), wiring what I thought be be a challenging but achievable custom harness for my dual HID 55w projector headlight setup.

I wanted to have separate fused power supply going to a separate relay and also be able to be toggled off on each side.

Did I achieve it? almost. I got the Right side working perfectly and was literally going crazy from the heat and couldn't get the other side (left) to come on. The ballast and bulb test OK when hooking up to my project car battery. Also the relay and a spare relay test fine as well hooked up in test mode. Guessing it is a bad connection or something plugged in wrong the Left side off this harness.

One piece of advice I can give so far is that you cannot use a toggle switch to interrupt the + trigger wire going into the relay (it just doesn't work when cycling on/off/on again). I had thought this would be desirable to do to not have a surge of 12v power going through the toggle switch.

With regard to the testing I did here's how it panned out:

Day 1
Idle: 1300-1400rpm
Temperature: mid 90's farenheit (all days during test, very hot out)

test 1: run bike 20 minutes both 55w HID headlights on. CEL comes on. Run 10 more minutes without stopping. Stop bike. Result: battery too weak to start bike. Console shows code 46 which we know means the charging system isn't putting out enough power

...battery tender plugged in to bike - battery charges overnight

Day 2
test 2: Run bike 30 minutes with 1 ballast unplugged to simulate turning off 12v power supply. Stop bike. Start bike 2x but with VERY weak crank. At least bike starts. Run for 15 more minutes and result is battery is too weak to start the bike.

..again put on battery tender overnight and battery is good again

Day 3
test 3: Run bike ~50 minutes with both ballasts unplugged. Stop. Start. awesome fast cranking. did this 2 times. very good.


So in summary what I describe above is really an acid test. Imagine an insane traffic jam. Most of us would probably turn the bike off after 15 or so minutes if not wanting to split lanes or not able to split due to trucks or other problem like a closed road.

Question for FZ6 electrical guys: Do you guys think putting in a double toggle switch (one on each side) is asking for trouble? (like me being a dumb dumb and forgetting to turn headlights back on)

Please let me know what you guys think :thumbup:

 

[TownsendsFJR1300]

DATEL 2-Wire AC & DC Meters

 

[FinalImpact]

It is a difficult decision whether to automatically switch one/both the lights off when on low battery. How about this:
1. Provide switches to turn on/off each lights independent of each other.
2. Provide LED indicator to indicate low batt. This has to be a highly power efficient design or else the indicator will just add extra load on the weak battery. This keeps blinking till battery recovers.
3. Provide LED indicator that flashes if one/both lights are manually switched off and the battery has recovered its charge.

If the proper connectors for the headlight assembly if available, it can be built as a plug-n-play device.

Hi,

I have been working on a few electronic gadgets for the Fz as you know and one thing that I am handicapped with is the ability to analyse the signals on the bike. I followed a few of your posts and looks like you have access to an Oscilloscope (just guessing). I was wondering if you could help analyzing the waveforms from the speed sensor and the tachometer.

I am working on building an arduino board to interface with all these sensors. This will have interfaces to an app on Android to act as a display.

Regards,
Rajesh

From your PM:

Rajesh,
I brought home a portable scope but on the drive home it dawned on me the info you seek RPM/Tachometer comes from the ECM on the buss along with other data. From this, we have to go straight to the Crank Trigger to see what it's signal looks like. I'm thinking I should be able to capture that so we'll see. I didn't bring a real high impedance probe so results may vary hopefully it doesn't load it down to point killing it! lol

The speed sensor should be easy as its right there to grab. I'll post up later in this thread.

 

[chunkygoat]

I have a number of ideas here. I work in for an industry leader in lighting - I am in R&D on the ballast/LED driver team - so I am well versed in lighting and electronic design. A few things really stand out to me off the bat that may help here:

First, what lamps are being used? HID? Halogen? Etc. Also - what ballast is being used? I would need to see the specs for both the lamp and ballast and perhaps even a schematic depicting this dual headlight mod.

Unknowing of these specific details, I noticed that the ideas present here only would indicate to the end-user when an issue is present, which is a terrific safety feature. However, this does not address the underlying issue itself.

A simple solution would be to replace the battery with one with a larger cranking amp rating. However, I don't see it practical trying to cram a larger battery under the tank, given the space allotted.

I am wondering if replacing with the lamps with LED load of equivalent output power and swapping the ballasts with a driver -- if this might help conserve power. This would enable us to utilize a small micro to PWM the headlight, which might sound insane initially. However, in our industry, we have done a tremendous amount of research and have discovered that the human eye cannot detect any visible difference between 100% light output of a lamp and 70% light output. This equates to substantial drop in power consumption, by applying a 70% duty cycle to our output, expect a drop in current draw, which would address what I would suspect to be our main culprit here (excessive current draw).

Dimming a ballast also equates to energy savings, which is possible as well - but would require a different ballast. Some ballasts come with micros installed to pre program light levels. By programming the ballast to run high only at 70% light output, this too would help achieve maximum energy consumption and lighten the load of the battery with no visible difference in light output.

How it works is the lamps have 2 pins on the inside that are heated with a small voltage (preheat). Once the cathodes are heated to an adequate temperature, it is SLAMMED with a high voltage from the ballast (strike) and creates an arcstream of electrons through the mercury and other vapors inside the lamp. It is in some sense, like a TRIAC. Once the arcstream is created, it will stay "latched" (or emitting photons from the vapor via the arcstream). You can begin to lower voltage to the lamps and the arcstream will not drop out, but begin to reduce the flow of electrons until the electrons ultimately stop flowing and the lamp drops out (similar to breakover V). If we lower our voltage to the lamps after striking, to achieve a 70% light output - which would cause no visible difference - we can achieve a tremendous conservation of current flow to the lamps. We may even be able to take the digital output of our tach and have a micro tell the lamps only to dim to 70% when idling under a specific RPM.


I am not sure if a ballast like this exists for such low DC voltage - I do know they exist, but I have only ever seen these types of ballasts operate with line voltages. I will start snooping around and see what alternative ballast solutions we may have for options.

I would attempt to address current draw on the battery. Maybe if somebody can connect an ammeter in series with the battery (while maybe a volt meter across it) to truly document what is happening at low rpm's versus higher rpm's - this would help us understand if our issue is from loading down the battery beyond its ratings or capabilities.

Reducing the load of the battery by:

• Changing the ballast/programming a 70% light output

• Replacing lamps with eco friendly lamps (ex: 28w eco lamp is equally as bright as a 32w non eco lamp)

• LED loads with PWM'd output via a micro and driver rather than HID or halogen and a ballast

I have much experience in PWM'ing and with high power LED's/switching power supplies and would be very much interested in developing an LED replacement for the stock headlight.

 

[TownsendsFJR1300]

On a stock bike, the system uses standard H4 and H7 bulbs, no ballasts...

There are several dual headlight mods out there.

I personally have the BD43 mod that actually plugs into the factory harness and runs a new line to the right side (stock-no power to the RS, low beam bulb) NOW, sending power to it (now 3 lines).

The right side, now when activated, the other filiment turns off so at NO TIME, are two filiments burning in the bulb at one time.


As for the rest, your talking Greek, and I don't speak Greek...

 

[FinalImpact]

Ultimately the best solution to this problem is a higher output charging system as a I don't see replacing the battery solving the problem, it might hedge it off for a short time, but ultimately you still end up with a depleted battery that requires time and energy to recharge. Its those two variables that kill us.

Unfortunately the design is just cheap. The stator output is effectively 100% all the time and what doesn't get used by the bike in terms of watts of energy is dissipated as heat by the regulator. In this system the output is fixed with the only variable being rotational speed which increases the voltage output but NOT the CURRENT output. From this, the system is not actively dealing with the demand like a conventional system in an automobile.
The more conventional means uses a regulator to vary the current to the field coils (make a stronger magnet) which increases the stators current output based upon demand. In this system, as the load increases, the output can be raised to meet the demand (to point) without having to race the engine. Our bike uses fixed magnets so this isn't a practical.

This is just my take on the thread here, but bottom line is cost effective reliable solutions that are easy to install. Too invasive or expensive and it doesn't happen. If dropping in an LED head lamp would be just as reliable, offer the same or better output - its great solution.

Me personally the Halogen lamp is very reliable - a cheap ebay HID import, I don't want it on my bike as its not as reliable as a tungsten filament until you pay substantially more for it.

 

[Motogiro]

The ballasts you're speaking about are generally called out on HID systems. The ballasts are the actual High voltage device used to produce the arc stream you're referring to in the HID. The word, "Ballast: can be confused because it is often used vocabulary when referring to load resistors or transformers used in fluorescent lighting.

On the HID type ballast the configuration is probably as you described. It would consist of an oscillator driving a toroidal type transformer and programmed as you described for the hot strike and then drop back for the lower current to keep the arc from dropping out. These systems typically us only 35 watts and put out much more efficient light output for the compared wattage used.

The subject here is how to engineer a system that can drop current use in the event of low idle charging in relationship to the standard incandescent H4/H7 lamps.

Are you on the Lutron team?

 

[chunkygoat]

The ballasts you're speaking about are generally called out on HID systems. The ballasts are the actual High voltage device used to produce the arc stream you're referring to in the HID. The word, "Ballast: can be confused because it is often used vocabulary when referring to load resistors or transformers used in fluorescent lighting.

On the HID type ballast the configuration is probably as you described. It would consist of an oscillator driving a toroidal type transformer and programmed as you described for the hot strike and then drop back for the lower current to keep the arc from dropping out. These systems typically us only 35 watts and put out much more efficient light output for the compared wattage used.

The subject here is how to engineer a system that can drop current use in the event of low idle charging in relationship to the standard incandescent H4/H7 lamps.

Are you on the Lutron team?

We would need a driver capable of a constant current output. We would create a sense network to read current and feed this to a micro. At the same time, the micro would be reading the digital output of the tach. Based on the information gathered from the tach, the contant current output could be scaled back by switching between pre-designed circuits utilizing JFETS for selectable constant current outputs depending on the desired light output.

I might start toying with this idea.

And yes sir, I am on the team at Lutron