[theshadow27]

Planning on doing a couple mini/half KBOWE units with H8 connectors before I tap into the main power harness. Circuit look good?

Looks good to me :thumbup:

Get a good ground for each half, as close as possible to the cap, 12awg+ if possible.

 

[dwells]

Looks good to me :thumbup:

Get a good ground for each half, as close as possible to the cap, 12awg+ if possible.

Thanks for the feedback. To clarify, I was planning on using the existing H11 negative connector wiring as the ground, so the finished deal would've looked like this (with the box hiding the capacitor and diode):

Are you suggesting that I add in a vehicle chassis ground as well that would be connected to the negative leg of the capacitor and thus the negative line of the H11 wiring, a la something like this:

 

[Laze]

Did exactly what you have drawn on your last drawing a while ago and everything still works perfect. Happy soldering!
After having fried two windscreen wiper motors it is quite rewarding.

 

[theshadow27]

Are you suggesting that I add in a vehicle chassis ground as well that would be connected to the negative leg of the capacitor and thus the negative line of the H11 wiring, a la something like this:

Yes exactly - do the second version. The reason is to reduce HF/RF impedance to ground; the large capacitor acts as a "short" across the supply at almost any AC frequency, so with a good ground close to the cap all supply wires are at equal potential to the chassis.

If you have metal ballasts (any decent one should be) and you bolt them to the chassis (I would recommend doing so for heat dissipation - there is a good location behind the headlights where the factory GDL module would have gone) then you can share that as the ground point. Otherwise, find somewhere close to the mounting point (keep the wire short) and don't go back to the strut tower where the WMM grounds. Thought in theory it shouldn't matter, I think it's asking for trouble.

Did exactly what you have drawn on your last drawing a while ago and everything still works perfect. Happy soldering!
After having fried two windscreen wiper motors it is quite rewarding.

Good to hear!

 

[dwells]

Yes exactly - do the second version. The reason is to reduce HF/RF impedance to ground; the large capacitor acts as a "short" across the supply at almost any AC frequency, so with a good ground close to the cap all supply wires are at equal potential to the chassis.

If you have metal ballasts (any decent one should be) and you bolt them to the chassis (I would recommend doing so for heat dissipation - there is a good location behind the headlights where the factory GDL module would have gone) then you can share that as the ground point. Otherwise, find somewhere close to the mounting point (keep the wire short) and don't go back to the strut tower where the WMM grounds. Thought in theory it shouldn't matter, I think it's asking for trouble.

Excellent, thanks again for the feedback. I'll have to explore my options for mounting and see what's best. If I find a good spot to bolt it to the vehicle chassis, then maybe I'll use an aluminum enclosure, enclosure/chassis ground it, and then once the closure is bolted to the vehicle chassis it'll be fully chassis grounded.

Really this is more about having an excuse to do a project than anything else; my poor digital Hakko soldering station is far too nice to be as neglected as it is.

 

[mklinker]

Shadow, can you point me towards how you installed the HID kit into your car? Also, what is your current recommendation for a HID kit?

I have a 08 s40

 

[theshadow27]

Shadow, can you point me towards how you installed the HID kit into your car? Also, what is your current recommendation for a HID kit?

I have a 08 s40

Hi there - welcome to the forum.

I installed my ballasts where the factory GDL modules would go. If you take out the wheel well liner and look towards the headlight area, you'll see a roughly rectangular cutout. I bolted the ballasts to the back side (side with wheel) and routed the wires around to the front. When the wheel well liner goes back in they are nearly invisible, and since its bolted to the chassis the heat dissipation is excellent. They are also out of the (very hot) engine bay. From there, I drilled a hole in the removable headlight access cover and installed the kit like normal.

Personally I prefer the 50w kits from HID50.com, my first set ran for 5 years and 3 sets of bulbs before it developed sporadic starting issues (this is it's expected life span), and I replaced it with the same thing. They are built very well, but are expensive (180gbp shipped the last time I checked).

The DDM slim ballasts are a good low-cost alternative, they don't last nearly as long but you could buy a lot of kits for the cost of one HID50.

Remember with any kit you'll need to build the KBOWE described in this thread, DRL or not.

Hope this helps, and good luck!

 

[mklinker]

Hi there - welcome to the forum.

The DDM slim ballasts are a good low-cost alternative, they don't last nearly as long but you could buy a lot of kits for the cost of one HID50.

Remember with any kit you'll need to build the KBOWE described in this thread, DRL or not.

Hope this helps, and good luck!

I am 100% comfortable making the KBOWE, but I also wanted to know how much modification was required of the sealed headlight units. My understanding is you plug the ballast in to the old bulb wires (inside the headlight unit), mount the ballast outside the sealed unit, then route the ballast to bulb wires back into the headlight unit. That tells me I will have to at least drill one hole in each of the sealed units, then seal it back up with silicone with the wires installed.

Secondly, has there been a noticeable difference in the lifespan of the DDM HID kit ($30) and the DDM Ultra HID kit ($80)?

 

[theshadow27]

I am 100% comfortable making the KBOWE, but I also wanted to know how much modification was required of the sealed headlight units. My understanding is you plug the ballast in to the old bulb wires (inside the headlight unit), mount the ballast outside the sealed unit, then route the ballast to bulb wires back into the headlight unit. That tells me I will have to at least drill one hole in each of the sealed units, then seal it back up with silicone with the wires installed.

Yes - one hole per housing as would be required with any aftermarket HID retrofit. The kits come with rubber grommets to seal the holes (one bushing per bulb) pre-molded around the wires for a tight fit. Adding extra silicone to the grommet is always a good idea. If you are like me, and don't want to re-do the silicone when the bulbs burn out ** then you can buy an extra set of "super seal" connectors and install them on the HV leads inside of the housing. This will all make sense once you do the install

Note: HID bulbs don't last forever, though they typically last 3-4 times longer than halogen bulbs (~4-6000h). All automotive HID capsules are designed to run at 35w, so even when you buy a 50w ballast they send you 35w bulbs. The light is brighter, but it cuts the bulb life down significantly. I go through a pair of bulbs every ~18 months as they dim substantially (although still above the level of the original halogen H11s). I like my lights bright If you don't want to deal with this, you can go for 35w ballasts.

Signs of failing bulbs include misfires, dark blue/purple light, black spots on the capsule, and dim light output.

Secondly, has there been a noticeable difference in the lifespan of the DDM HID kit ($30) and the DDM Ultra HID kit ($80)?

I've only ever used the $30 DDM slim kit, so I am not sure about their higher priced versions. It is very similar to the HID50 unit. I did a writeup a while ago:

http://theeshadow.com/files/volvo/hidcomp.html

Hope this helps!

 

[sketchart]

Sorry for bumping this thread, and if this has already been covered, but has anyone used the Canbus unit from The Retrofit Source?

 

[theshadow27]

Moving PM reply here...

I've been spending some time data logging on most of my car starts. I've driven a ton recently (Over 1000 miles in the past week), so the majority of my data is from very recently. Have a look for posterity.

This is only the hard data I've written down on my phone. From experience (starting the car and looking at my headlights), the trend of 'daylight kills headlights' is actually real. Temperature has no effect on the starting reliability. When I last brought up the diode replacement, I saw no functional difference, so I bought new capacitors as well. I basically rebuilt the entire KBOWE twice over, and on the third time, I added the previously discussed heatsink to be sure. It's from a former power supply, so I salvaged the thermal pads and bolt insulators. DMM'd to be sure that there was no electrical continuity between the heatsink and diode cathode, then stuffed everything back into a modified box.

Back in the car, the heatsink gave a pretty easy means of checking whether current was flowing through the diodes (is it hot?). Whenever the headlights failed to start, I'd drive for a few minutes (with no BOW, mind you), and the heatsink would always be ice cold. So no current was going through the KBOWE, yet no BOW would appear! I corroborated this with a DMM. Current flows on startup, then zero current through the diodes upon failure.

To clarify, the 'failure' is never a 'doesn't ignite' mode. The headlights always ignite, but given 3-5 seconds, will then turn off (failure). Usually simultaneously, but occasionally with a half-second interval between them.

When I was about to leave Grand Canyon, I had left a bunch of junk on the dashboard (papers, brochures, and of course the windshield reflecto-cover). I started the car before moving any of it, and the headlights stayed on! Further testing seems to prove this. At 3PM with the car thermometer reading 110 and the interior baking, simply covering the sun load sensor allows them to stay ignited!

I recorded a quick video of this failure mode an hour ago. For reference, 'lovely morning' means 10:30AM and roughly 85°. Car was still warm after about an hour of driving prior to this.

I initially asked if some circuitry magic existed between the sun load sensor and the lights, but now I'm grasping at straws for solutions, so what could this be? I know for a fact that my DRL's are not disabled (for example), yet I also know that my thrice-built circuit works correctly at operating conditions (tested with on-campus scopes and a waveform generator).

The only potential oddity is that your recommended troubleshoot test (voltage from diode anode to ground should be 0V while cathode to ground should be 12V) doesn't work. Logically I also don't see how it can. My EE roommate is a bit of a slacker, but he says that with the diode set that way, the anode and cathode are functionally the same node to a standard DMM.

Measure positive to anode/cathode and negative to ground, get 12V. Measure positive to ground and negative to anode/cathode, get -12V. I've been through like 8 diodes at this point and each time I replaced them or rebuilt the circuit, they were always cathode to positive capacitor terminal.

So given all this stuff, where do you think I should go from here?

Easy stuff first:

The Diode anode (aka input) to ground = 0 test assumes that the impedance of the DMM is low enough to compensate for the reverse (leakage) current in the diodes. A combination of leaky diodes (relatively speaking, microamps vs nanoamps) and a high quality DMM (megaohm input impedance) will result in the test appearing to fail. Try adding a 100 ohm resistor in parallel with the DMM when you do the measurement - this should reverse-bias the diode and allow for an accurate measurement.

In re to your slacker roommate - they are not the same node. A diode is non-linear, behaving different when forward and reverse biased. In this particular test, the anode is connected through a finite impedance to ground (within the DMM), putting it at a lower potential than the cathode-capacitor node and reverse biasing it. A (real, non-ideal) reverse biased diode exhibits a small leakage current, specified in it's data sheet. Small signal diodes have nanoamp leakage, normally negligible, but big switching diodes can sometimes leak microamps or even mA. Since charge on the capacitor is large compared to the currents involved, in the short (compared to RC) timescale the effect is a constant current source with a thevinin equivalent voltage equal to the charge voltage of the capacitor with a characteristic resistance equal to that voltage divided by the leakage current. In your case, the STPS10L25D/G diode has specified a worst-case 800uA leakage (when cold), so with the cap charged to 12V it is ~= 12V constant-voltage-source with a 15kOhm equivalent series resistance (TER). Hooking that to a DMM with an impedance of 10MOhm and you get a voltage divider which should read very nearly 12V. Adding 100 ohms across the DMM will create an equivalent resistance of nearly 100 ohms, which should read around 0.08V if your meter is good enough.

Now the wacky stuff. Thanks for taking the time to collect data, and that's a fascinating accidental discovery - that graph is as clear as day (no pun intended)...

The so called "twilight sensor" and "sunlight sensor" (both in the same plastic housing) input to the CEM, but I hadn't seen anywhere in the literature a mention that it affects the the PWM regulation. I actually remember specifically testing this on my car (2005) back in the day, and it had no effect. Most of my docs are from the initial release, so it's possible that a later software update or design refresh may have changed this. Quickly searching through the latest copy of VIDA I have (2012B) turned up some new sections I haven't seen before - I PDFed them for you:

http://theeshadow.com/files/S40MY2005/CEM_overview.pdf
http://theeshadow.com/files/S40MY2005/CEM_design.pdf
http://theeshadow.com/files/S40MY2005/CEM_function.pdf

Daytime running lights (10/56-57) are activated via an extra "auto-position" in Light switch module (LSM)(3/111). Daytime running lights can also, on certain markets, be activated via position 0 and P (parking lights).

When "auto-position" is selected in Light switch module (LSM) the Central electronic module (CEM) (4/56) uses information from the vehicle's twilight sensor (7/12) for automatic switching between low beams and daytime running lights. In case of missing or incorrect signal from the twilight sensor the low beams are used as usual.

Daytime running lights are never active together with running lights. On vehicles equipped with ABL-lights, there is a diode by the "auto-position". This diode pertains to the ABL-lights' function and not the daytime running lights.

For certain markets, when Light switch module (LSM) is et in "position P", the daytime running lights are deactivated automatically and the parking lights are turned when the automatic transmission's gear selector changes to P- or N-position, or when the parking brake is activated.

This would explain the observed behavior, except that the DRLs they have shown aren't what we have been calling DRLs...

Daytime running lights (from structure week 201044, option) - If the vehicle is equipped with daytime running lights, these are located in the bumper casing below the headlights where the front fog lights otherwise would be installed.

The DRL in the picture refers to something different that I've never seen... And on page 8 of the CEM_function document, they refer to "Low Beam Automatic" (ABL? Automatic Beam, Low?) which, I assume, is what we have been calling DRL. This is the first time I've seen this term... It's not in any of the other documentation.

And in the section on headlamps:

Depending on the position of the knob in the light switch module (LSM) (3/111), the module transmits a request to the central electronic module (CEM) to switch on low beam. Central electronic module (CEM) (4/56) supplies power to the low beam directly via an output. The level of the supply voltage to the low beam is controlled using pulse width modulation of the central electronic module (CEM). This means that the average value of the voltage level to the low beam does not exceed 13.2 V, despite the fact that the supply voltage may be higher.

Nothing about the sunlight sensor here... but clearly your car has a sunlight sensor dependance. Also, it would seem that the BOW is not triggered in DRL (ABL?) mode.

All of this is alien to me, so basic questions first.

What year is your car? US (AME) spec?

Some things to test (you can do this at any time using a flashlight & some paper on the twilight sensor, wait for the ICM display to invert):

(1) Does the LSM being in position '0' vs On make a difference?

(2) Can you reinstall one halogen bulb, and measure the KBOWE output voltage during 'day' and 'night'?

(3) With halogens installed, can you measure the frequency in 'day' and 'night' (DMM ok, across bulb)?

(4) Once started and running at 'night', if there is a transition to 'day', do the ballasts fail?

I have a hunch after looking over the data sheet for the diodes that I used in version 2 and version 3 I noticed that both have mA leakage current, between 10 and 100 times that of yours. It is possible that without this leakage, the CEM can't detect the voltage level at the capacitor and thus can't properly PWM regulate. Total shot in the dark, but maybe try adding 100ohm, 1k, and 10k resistors (start with 1k) between KBOWE input and output to simulate this leakage and see if there is any improvement?

 

[Nick0matic]

Moving PM reply here...

Nothing about the sunlight sensor here... but clearly your car has a sunlight sensor dependance. Also, it would seem that the BOW is not triggered in DRL (ABL?) mode.

All of this is alien to me, so basic questions first.

What year is your car? US (AME) spec?

Some things to test (you can do this at any time using a flashlight & some paper on the twilight sensor, wait for the ICM display to invert):

(1) Does the LSM being in position '0' vs On make a difference?

(2) Can you reinstall one halogen bulb, and measure the KBOWE output voltage during 'day' and 'night'?

(3) With halogens installed, can you measure the frequency in 'day' and 'night' (DMM ok, across bulb)?

(4) Once started and running at 'night', if there is a transition to 'day', do the ballasts fail?

I have a hunch after looking over the data sheet for the diodes that I used in version 2 and version 3 I noticed that both have mA leakage current, between 10 and 100 times that of yours. It is possible that without this leakage, the CEM can't detect the voltage level at the capacitor and thus can't properly PWM regulate. Total shot in the dark, but maybe try adding 100ohm, 1k, and 10k resistors (start with 1k) between KBOWE input and output to simulate this leakage and see if there is any improvement?

To preface, I don't have the halogen projectors anymore. I didn't see any use for them after removal (and relatively successful HID use when I needed them at night), so I got rid of them and the shipping stuff from TRS. Not the smartest decision, in hindsight, but it wasn't at the top of my mind at the time :facepalm:. To answer what I can:

a) Indeed, no BOW. No indication whatsoever that the lights had failed, but they were certainly off. More often than not they stop non-simultaneously. Any significance to that?

b) 2007, US spec, sort-of-facelift version (bigger side mirrors, different roof console, etc)

1) I'll test again, but I don't recall it having any effect. (I'll PM you in a few)

2) Uhh...

3) I now need to buy my own to do any of the DMM measurements.

4) I'll test this tonight. I was bouncing the idea of the day/night transition causing problems off my passenger when I drove to Flagstaff. I was a little too nervous to test it on the open road, however.

c) I don't have any resistors on hand, but are you suggesting I put all of them in series for each input/output?

 

[theshadow27]

I don't have the halogen projectors anymore.

Did you cut off the connectors? If it's actually reproducible with the light sensor you don't need to drive around with them. Hanging out of the side of the housing is ok (just don't let the envelope touch plastic lol).

a) Indeed, no BOW. No indication whatsoever that the lights had failed, but they were certainly off. More often than not they stop non-simultaneously. Any significance to that?

b) 2007, US spec, sort-of-facelift version (bigger side mirrors, different roof console, etc)

Since I have a 2005, I'm assuming that the CEM hardware didn't change with the facelift (according to VIDA, the main part number stays the same 31327215, but the alternate changes from 31254903 -> 31254749 & 31296881)... Usually a functional change breaking backwards compatibility will get a new primary part number. But there is obviously a software difference.

Each side is independently monitored so except for an "on" or "off" I wouldn't expect synchronization.

Been mulling over your lack of BOW. Just thinking out loud here so bear with me...

We know the system can detect both an open-circuit and low-load situation (normal burnt out bulb, relay harness triggers BOW), and the system can detect a short-circuit or overload condition (there is no fuse). There must be a voltage involved in the feedback loop, a differential shunt based system wouldn't be able to meet a spec like 13.0+0.2V.

It would be expensive and complex to have a two-error controller (current+voltage) in something like this, (not saying they didn't, but the odds are against it). So we are likely looking at a voltage-window controller sampling the output during the 'on' cycle of the PWM, essentially using the on-state resistance of the FET as the shunt. Too low an output voltage triggers the overload condition, too high an output voltage triggers the open-circuit condition.

What I'm having trouble with is the regulating 13.0V business. You need an integrating element, such as a capacitor or inductor, to get an average voltage error signal (feedback) from a PWM output suitable for a dedicated PWM regulator - once a warm, an incandescent filament is basically a fixed value resistor, and at 80Hz the parasitics are negligible. I see two possibilities: (a) there is an RC network before the ADC to integrate over several periods, or (b) there is no feedback and the PWM duty cycle is statically mapped to the absolute battery voltage (Vo = Vi * %d) which is already accurately measured by the CEM.

The more I think about it, the more sense (b) makes - lower component count, less finicky calibration, the way everything else in ECU is mapped. An integrating network would have a slow response time, potentially causing damage in a short circuit condition. It would be trivial to implement the map in a software addition after the hardware was designed/built, and I bet it's bit-banged rather than hardware PWM explaining the super low frequency. Some wise-arse Swede trying to extend everyone's bulb life. An 8-bit ADC would fit the bill perfectly, e.g. 25V / 256 = 0.1V which would allow for 13.0+0.2V regulation, and such hardware is ubiquitous in automotive control systems.

Now to correlate this theory with the observations:

-- Connecting the capacitor bank *without* a rectifying (series) diode produced a BOW and failed to ignite my ballast.
>> wild guess: voltage polled during the 'off' cycle to check for short-to-battery condition? It would be overly complex to code for this specific condition (given the map-based PWM duty cycle) but if the ADC is multiplexed between several inputs and polled at a constant rate by an ISR to update a set of "voltage registers" you'd have to account for the off cycle in case the sample was taken during it; eg. back in my embedded days I'd probably write bulb_ok = (v < V_MAX) & ( (out_en) ^ (v < V_MIN) ); if only to confuse the new kids.
<< test using a bench supply hooked in place of the KBOWE, I'll try to do this over the weekend.

-- Connecting the capacitor bank *with* the rectifying diode allows the ballast to work perfectly with PWM (in many cases except yours).
>> If the previous guess is correct, this is consistent. 0v during the 'off' cycle, a loaded voltage during the 'on' cycle.
<< This would be tested the same as above.

-- During 'night' time, your KBOWE works perfectly, but there is consistent failure during 'day' time - based on twilight sensor reading.
>> assumption: the twilight sensor turns on/off your DRL PWM. This is how *I* would have coded it - to provide maximum brightness at night - but it's not how my car (or many of the others who have duplicated my tests) works.
<< hard to definitively confirm without a o'scope, but a DMM with frequency would be a good start

-- When your DRLs are on (assuming this is what's happening) the ballasts ignite but shortly thereafter enter shutdown mode; yet there is *no* BOW.
>> So something is triggering your ballast's protection mode, most likely under voltage. The under-volt monitoring is usually disabled for the first few seconds to account for the extra current draw during startup & warmup. On an brownout shutdown, aftermarket ballasts will typically wait for a full 'off' before retrying, so a fair guess is the voltage dips below 9V but does not drop to 0. The series diode will prevent the 'off' cycle BOW, but something must be keeping the 'on' cycle voltage < V_MAX to avoid triggering the open-circuit error condition. A possible explanation could be that the main system voltage ADC is mis-calibrated in your car. E.g. if the CEM thinks that v_in is 15V when it is actually 14.5V, then v would always be less than V_MAX; which is v_in - (rds_on * i_MIN), the second term being somewhere around 0.2-0.3V. If you get a BOW with nothing plugged in, it's possible that it's miscalbrated slightly, such that the quiescent (standby) current draw from the shutdown ballast is enough to defeat the BOW.
<< be nice to get the pre and post KBOWE voltages with this going on, but readings from a DMM would be of questionable value; it really needs a 2ch o'scope. Voltage calibration can be tested by checking the system voltage readout via a DiCE + ViDA (or BT android OBD-II thing) to see if it matches a DMM. Miscalibration can be indirectly tested by measuring the quiescent current through the ballast, then using a resistance to draw the same current directly across the CEM leads without the KBOWE in place.

Again the above is just my educated guess - gonna have to test to find out for sure. Tempt to hit a junk yard for a CEM to take apart.

1) I'll test again, but I don't recall it having any effect. (I'll PM you in a few)

2) Uhh...

3) I now need to buy my own to do any of the DMM measurements.

4) I'll test this tonight. I was bouncing the idea of the day/night transition causing problems off my passenger when I drove to Flagstaff. I was a little too nervous to test it on the open road, however.

c) I don't have any resistors on hand, but are you suggesting I put all of them in series for each input/output?

Let me know how it goes, I'm glad to mail you whatever you need. No, the resistors would be used one at a time, in separate tests.

 

[Nick0matic]

Alright, so a couple things since I got everything I need - first round of tests intends to cover the HID v. Halogen stuff. I tested with driver side headlight as HID, and passenger side headlight as halogen. Bulb was plugged directly into the H11 socket in the headlight and was resting above the assembly with the back cover temporarily off.

Random observations:

• PWM is off until engine start. Testing V/Hz at LSM "0" with key in position II and engine off shows 12V / 0Hz.
• Interestingly, the day/night effect becomes very inconsistent with the engine off. I had "day" starts where the HID failed and the Halogen remained on, but never vice versa. Numerous occasions where both were on or both were off.
• When "day starting" with the Halogen running through the KBOWE, both lights fail as expected.
• Switching to night from day produces no results. LSM does nothing and both HID and Halogen remain off.
• This is not a cold-start problem. I managed to induce 5 consecutive day-start failures.
• When bypassing the KBOWE entirely, the Halogen consistently starts while the HID fails during "day".
• The measured duty cycle with bypassed Halogens was usually 97-98%, which seemed a little high, but the input voltage was very low (found the undervoltage protection).

Coolest part:
When bypassing the KBOWE, the halogen bulb clearly reacts to the day/night sensor. It's so cool to watch, so I recorded it (See PM!) As well, the bulb has a visible increase in brightness when the LSM is switched from "0" to on. However, this brightness change only occurs when it is "day." At "night," the "0" or DRL setting is identical to "on." In addition, the halogen reacts immediately to changes in the day/night sensor.

Now, for these remaining questions:

Some things to test (you can do this at any time using a flashlight & some paper on the twilight sensor, wait for the ICM display to invert):

(1) Does the LSM being in position '0' vs On make a difference?

(2) Can you reinstall one halogen bulb, and measure the KBOWE output voltage during 'day' and 'night'?

(3) With halogens installed, can you measure the frequency in 'day' and 'night' (DMM ok, across bulb)?

(4) Once started and running at 'night', if there is a transition to 'day', do the ballasts fail?

...Total shot in the dark, but maybe try adding 100ohm, 1k, and 10k resistors (start with 1k) between KBOWE input and output to simulate this leakage and see if there is any improvement?

(1) No. If the light fails, restart is required. Voltages remain constant (and depend on day/night) and frequency is 0Hz

(2) Slight change. Of course, the lights require the car to be started in the "night" condition for the lights to stay on long enough to get a measurement. But once started, switching between "day" and "night" is easy. See PM for data

(3) 0Hz across the board. Seems like the ECM is killing PWM.

(4) No. Strangely, however, if I switch from "night" to "day," the Halogen switches off after a few seconds. Through the KBOWE, of course.

I'll get to the parallel resistor test tomorrow. If you have anything else you want me to try, do suggest it. All the values in the existing spreadsheet are from averages of 3-4 tests each (usually multiple at a time)

 

[theshadow27]

Nice work!

Looking back through the results list it does seem like everyone who has posted a KBOWE success story has a pre-facelift (05-06) year, with the exception of one '08 C30 - but IIRC the C30 "update" was in 2010 so this might be the same as the pre-facelift S40/V50. Don't know how I could have missed this trend, it's pretty obvious looking at it now...

• PWM is off until engine start. Testing V/Hz at LSM "0" with key in position II and engine off shows 12V / 0Hz.
• Interestingly, the day/night effect becomes very inconsistent with the engine off. I had "day" starts where the HID failed and the Halogen remained on, but never vice versa. Numerous occasions where both were on or both were off.

I noticed this too - I suspect it's because the battery voltage is below the minimum for PWM (probably around 13.5) without the contribution from the alternator. For some reason to this day I have issues with the second+ warm start without the engine running - Don't know if this is the CEMs attempt to conserve battery power or something.

• When "day starting" with the Halogen running through the KBOWE, both lights fail as expected.
• Switching to night from day produces no results. LSM does nothing and both HID and Halogen remain off.
• This is not a cold-start problem. I managed to induce 5 consecutive day-start failures.

Errr this is not what *I* expected - I predicted the halogen would work normally through the KBOWE regardless of input voltage / duty cycle. Short of the leakage current theory (which is a long shot) I'm at a loss for an explanation...

• When bypassing the KBOWE entirely, the Halogen consistently starts while the HID fails during "day".
• The measured duty cycle with bypassed Halogens was usually 97-98%, which seemed a little high, but the input voltage was very low (found the undervoltage protection).

Can you clarify on the input voltage being low, despite a high duty cycle?

Coolest part:
When bypassing the KBOWE, the halogen bulb clearly reacts to the day/night sensor. It's so cool to watch, so I recorded it (See PM!) As well, the bulb has a visible increase in brightness when the LSM is switched from "0" to on. However, this brightness change only occurs when it is "day." At "night," the "0" or DRL setting is identical to "on." In addition, the halogen reacts immediately to changes in the day/night sensor.

Now, for these remaining questions:

(1) No. If the light fails, restart is required. Voltages remain constant (and depend on day/night) and frequency is 0Hz

(2) Slight change. Of course, the lights require the car to be started in the "night" condition for the lights to stay on long enough to get a measurement. But once started, switching between "day" and "night" is easy. See PM for data

(3) 0Hz across the board. Seems like the ECM is killing PWM.

(4) No. Strangely, however, if I switch from "night" to "day," the Halogen switches off after a few seconds. Through the KBOWE, of course.

I'll get to the parallel resistor test tomorrow. If you have anything else you want me to try, do suggest it. All the values in the existing spreadsheet are from averages of 3-4 tests each (usually multiple at a time)

Really great data - with visual confirmation through your video! The most interesting portion of that data is the day/night voltage:

Test 3.5: Bypass KBOWE, Halogen Bulb, position "0"

PD State	PWM Freq	Current*	Voltage
Day	82.26 Hz	2.16 A	8.7 V
Night	82.26 Hz	3.00 A	12.16 V

[SUP]* Note - I think your current values were off by a decimal point 10^1 (0.30A -> 3.0A) [/SUP]

The averaged 'day' voltage (8.7 V, probably 9V at the CEM output) is *just* at the ballast undervoltage protection cutoff. This might explain why once started at 'night' you can switch between 'day' and 'night' without shutting down the ballasts - Once it's warm, the parasitic voltage drop over the factory wiring is low enough to keep it on (just barely) but when 'day' cold started the additional warm-up current increases the voltage drop to under the cutoff threshold putting the ballasts into shutdown.

Just to confirm one thing - With the full KBOWE+HID installation and the LSM 'On', the HIDs still fail during the 'day' -- even though LSM 'On' means the halogen is as bright as 'night'? Here's a truth table, can you ensure I understand your measurements correctly:

Eng  LSM   PD   -->   Vcem     OE Bulb   KBOWE+HID
==================================================
0    1/0   x    -->  13 VDC?     full     50/50 
0     R    x    -->   0 VDC      off       off 
1     0    D    -->  9  VPWM     dim      fail
1     1    D    -->  13 VPWM     full     ????
1     R    D    -->   0 VDC      off       off
1     0    N    -->  13 VPWM     full     pass
1     1    N    -->  13 VPWM     full     pass
1     R    N    -->   0 VDC      off       off

I always leave my LSM in "On", and if ???? == pass, then obviously that's the way to go...

Now, for these remaining questions:

(1) No. If the light fails, restart is required. Voltages remain constant (and depend on day/night) and frequency is 0Hz

(2) Slight change. Of course, the lights require the car to be started in the "night" condition for the lights to stay on long enough to get a measurement. But once started, switching between "day" and "night" is easy. See PM for data

(3) 0Hz across the board. Seems like the ECM is killing PWM.

(4) No. Strangely, however, if I switch from "night" to "day," the Halogen switches off after a few seconds. Through the KBOWE, of course.

I'll get to the parallel resistor test tomorrow. If you have anything else you want me to try, do suggest it. All the values in the existing spreadsheet are from averages of 3-4 tests each (usually multiple at a time)

Are you are seeing 0Hz with the DMM connected with black lead to ground (frame, battery, aux lighter shell, etc) and the red lead to positive of the KBOWE input/output or halogen socket - i.e. the same configuration as a voltage measurement? I only ask because the way the voltage would vary is via PWM, so there should be some frequency. It's possible the meter is having trouble detecting the high duty cycle - a surefire way to test this is to switch to the AC volts range.

Looking forward to seeing the resistor tests. Just so we're on the same page:



Still confuzzled by the halogen going out with no BOW. Worst case, you might need to do the DRL removal flash 30679690...

 

[Nick0matic]

Errr this is not what *I* expected - I predicted the halogen would work normally through the KBOWE regardless of input voltage / duty cycle. Short of the leakage current theory (which is a long shot) I'm at a loss for an explanation...

Based on what you said, I expected this to happen. The ECM can't pulse width modulate through the KBOWE for whatever reason, and kills the circuit. Maybe a DiCE can unlock that secret.

Can you clarify on the input voltage being low, despite a high duty cycle?

You saw it in the data. Perhaps "very low" is a misnomer, but 8.7V was what I measured during "day." Both day and night produced duty cycles of 98% on the meter when on a bypassed Halogen.

Really great data - with visual confirmation through your video! The most interesting portion of that data is the day/night voltage:

Test 3.5: Bypass KBOWE, Halogen Bulb, position "0"

PD State	PWM Freq	Current*	Voltage
Day	82.26 Hz	2.16 A	8.7 V
Night	82.26 Hz	3.00 A	12.16 V

[SUP]* Note - I think your current values were off by a decimal point 10^1 (0.30A -> 3.0A) [/SUP]

Believe it or not, those are the values straight from the DMM. I measured current from KBOWE output to ground and input to ground and both were identical, yet were always 0.2 or 0.3. No idea if that's a systematic issue with the DMM, but it's an auto-ranged value, so...

EDIT: For anyone reading this in the future. I never switched from V/Ω/Hz input to current input, so it wasn't actually measuring any current! Whoops!

The averaged 'day' voltage (8.7 V, probably 9V at the CEM output) is *just* at the ballast undervoltage protection cutoff. This might explain why once started at 'night' you can switch between 'day' and 'night' without shutting down the ballasts - Once it's warm, the parasitic voltage drop over the factory wiring is low enough to keep it on (just barely) but when 'day' cold started the additional warm-up current increases the voltage drop to under the cutoff threshold putting the ballasts into shutdown.

Not sure if it's working that way! When running the KBOWE at night, the car doesn't seem to give it any PWM. If I switch from "night" to "day," the voltage decreases very slightly. Though the ballasts won't start during the day, the switch from night to day isn't causing the voltage to drop to <9V. I'll try and watch the voltage/frequency readings during the "day-start" procedure on both HID and halogen, but I have no alligator clips so it'll be a little effort.

Just to confirm one thing - With the full KBOWE+HID installation and the LSM 'On', the HIDs still fail during the 'day' -- even though LSM 'On' means the halogen is as bright as 'night'? Here's a truth table, can you ensure I understand your measurements correctly:

Eng  LSM   PD   -->   Vcem     OE Bulb   KBOWE+HID
==================================================
0    1/0   x    -->  13 VDC?     full     50/50 
0     R    x    -->   0 VDC      off       off 
1     0    D    -->  9  VPWM     dim      fail
1     1    D    -->  13 VPWM     full     ????
1     R    D    -->   0 VDC      off       off
1     0    N    -->  13 VPWM     full     pass
1     1    N    -->  13 VPWM     full     pass
1     R    N    -->   0 VDC      off       off

I always leave my LSM in "On", and if ???? == pass, then obviously that's the way to go...

Yes. Whether HID or halogen, leaving the LSM "On" during a day-start produces no different results than if "0." I'll try and record the values from those 5 seconds of startup time to see what's happening.

Are you are seeing 0Hz with the DMM connected with black lead to ground (frame, battery, aux lighter shell, etc) and the red lead to positive of the KBOWE input/output or halogen socket - i.e. the same configuration as a voltage measurement? I only ask because the way the voltage would vary is via PWM, so there should be some frequency. It's possible the meter is having trouble detecting the high duty cycle - a surefire way to test this is to switch to the AC volts range.

DMM black was always to the same ground bolt as the KBOWE ground. Red lead was always to the KBOWE input cables. The DMM measured frequency in this configuration perfectly for the bypassed Halogen (per data, same during day and night) with 97-98% duty cycle. The bypassed Halogen had the expected frequency, but duty cycle was odd. Since I only have steady-state data, I would like to see what happens during the "day" failure at startup and then compare it to a "night" startup. In either scenario (per the data), the voltage input to the KBOWE legitimately changes between "day" and "night" when unmodulated. I can tell you now that the brightness of halogen did not change with LSM changes or day/night changes when it ran through the KBOWE.

 

[theshadow27]

Facelift info from the stats thread is a little harder to extract, not everyone posts complete answers (especially year)

08 S40:

just the ballast plug and play "Low beam failure" only displayed when my lights where in the auto position due to 2 low of wattage so i just got my Volvo dealer to flash my system to make it an off switch instead of auto

2008 S40, no DRL info:

installed a kit from SharpHID 55watt with KBOWE3's for my lowbeams about 6 months ago. Worked OK for a couple weeks, occasionally one of the bulbs would not ignite, but the problem was intermittent. Figured "Ok, I can live with this" Eventually, I'm on my way home and I got a dash message saying something about "anti skid service required" and "immobilizer".... Needless to say I took the kit out, and I'm back to my stock halogens.

08 S40:

So I got a 35w 8000k of eBay. No bow and drl's still on lights are weaker then my halogens but I'm sure it's cus there from ebay. Had them on for about 6 months now constent low beam failure message and occasional passenger side light goes out. Wipers still good. As of today I'm back to my halogens because I want a good hid kit and done right.

After that, it gets hard to pick out real data from banter. But the first post is especially interesting.

 

[theshadow27]

Got off my ass and duplicated Nick's day/night experiment.


Setup: current/voltage instrumentation at the right headlight power connector (measuring V_load and I_out on the image above).

T#	LSM	twilight	V_load	I_out	V_batt	result
1	On	day	12.80	4.24	14.49	good
2	0	day	0	0	14.57	fail
3	On	night	12.77	4.29	14.57	good
4	0	night	12.78	4.32	14.59	good

Test Notes:

1) From initial start only. Switching back once out does not restart lights. BOW when unplugged EXCEPT if unplugged while "failed" in state 2.
2) After 5-10 seconds (On->0 or night->day), both ballasts self-extingush non-simultaniously. Switching to 'on' and back has no effect. Restart veh. to get lights back. Voltage was observed to drop to plateau at ~12.5 (immediate), then to 9.5 (~1 second), then lights went out, then taper gradually to 5 as the KBOWE discharged. Current dropped to 0 as soon as ballast extinguished. No BOW regardless of LSM position or sensor reading.
3) Can cover/uncover twilight on and there is no change to ripple or voltage. BOW when unplugged, immediate restart when re-plugged.
4) Same as #1 and #3, BOW when unplugged, immediate restart when re-plugged.

Remarks:

Did a benchmark with engine off, key out, high beam stalk activated: V_load = 9.70, V_batt = 11.68, no PWM. Cumulative effect of parasitic resistance (harness and measurement apparatus) and diode FV is 1.98V, probably split 1.5 + 0.5.

There is minor >1% difference in output voltage between the different "good" results, probably negligible based on the DMMs used, but based on V_batt this could just be my battery charging.

Weird that if the lights are allowed to go out in state #2, the CEM seems to "give up" on ever having good headlights, and does not report a BOW even when switched on manually.

I always leave my LSM in 'On', and the lights always work, so I didn't notice this behavior before. It's possible that this behavior (leaving them always 'on') was influenced by some indirect, incomplete observation of the above during initial testing.

I do remember going through the '0' - 'on' tests when I was developing the KBOWE, and though those notes are long gone, I distinctly remember there being no difference. This could be because I was testing at night, or it really could have changed during an unrelated CEM update (there have been a few, latest was the radio failing one). Or maybe I'm just wrong. Either way, based on this data, pre-post facelift seems to be irrelevant.

In summary, the KBOWE only works reliably when the LSM is 'On'.

Edit: Something fishy is going on. Benchmark shows 1.98v of parasitic loss, but 12.75+1.98 = 14.73 which is greater than my battery voltage. I was using a different meter for V_batt, and I couldn't measure duty cycle (being after the KBOWE) but this would imply PWM duty cycle under all "good" results is nearly 100%? Or there is PWM with fixed duty cycle when engine off and I couldn't pick it up on output side of KBOWE.

 

[theshadow27]

So found some interesting stuff digging through VIDA. Specifically, PWM is a traceable variable

Screenshots: http://theeshadow.com/files/volvo/pwm_vida_data.pdf
Compiled spreadsheet: http://theeshadow.com/files/volvo/pwm_vida_data.xlsx

Looks like I shouldn't have been second guessing myself All of the original observations are correct. My 50,000uF (per side) KBOWE just keeps the ripple below background noise!

The duty cycle is highly dependent on battery voltage - not just for the low beams, but also for the brake lights and position (running) lights! The duty cycle for everything was pretty much in lock-step (there might be some variation, but this is an artifact caused by the way the DiCE reads values sequentially independent of the CEM update rate). I couldn't get screenshots of everything, but roughly,

<= 13.5: 100%
14.1: 93%
14.3: 90%
14.4: 82%
14.6: 81%

Super steep curve, 20% per volt or so after 13.5.

To interesting datapoints on BOW behavior:

(1) In the failure mode (day/off), the duty cycle drops to 58-60% before dropping to 0% when the ballasts go out. The CEM knows the outputs are off, and there is no BOW. Changing LSM position and unplugging/replugging ballast has no effect - stays 0%.

(2) With the ballast unplugged (simulated BO), the duty cycle does not change L/R. It reduced slightly (82->81%) but at the same time the battery voltage increased because of the reduced electrical load.

 

[theshadow27]

For the sake of posterity, the above quandary has been resolved thanks to a CEM Teardown, some insight from Nick, some ViDA readouts, and a 'duh' moment...

According to the Infineon BTS443P datasheet, the PROFET on each low beam circuit produces a monitoring current proportional to the current draw through the switch. This is passed through a resistor, producing a voltage which is monitored by one of the ADCs in the CEM's microprocessor. This voltage reading is used for two functions:
(1) A software based 5A over-current threshold, i.e. an "electronic fuse" which when exceeded: (a) stores a "dipped beam, out of range low" DTC and (b) disables the circuit until the vehicle is restarted to protect the solid-state switch and wiring.
(2) A software based 2A minimum-current threshold, if not met displays the "bulb failure" message on the DIM.

From the ViDA readout experiments we know that the control is open-loop (no feedback) based on battery voltage.

The normal condition (PWM-regulated) target voltage is 12.5V (80% duty @ 14.5V VBatt). Ballasts are constant-power switching converters, so at this voltage they draw 4.4A w/55 W or 2.8A w/35W.

The abnormal condition T#2 (day time, LSM=0) target voltage is 8.8V~9V (~60% duty @ 14.5 VBatt), still open loop no feedback. The lower effective voltage produces some unusual behavior depending on the ballast:
(1) If the rectified/smoothed voltage is below the ballast brownout point (typically 9V, but maybe a little less) the ballasts will refuse to fire. Current drops below the open-circuit threshold, and we get the "Bulb Failure Low Beam" on the DIM
(2) If they don't cut out, they will draw more current - nominally 6A w/55w or 3.8A w/35W, but higher during warmup. This will trip the 5A soft fuse, killing power to the circuit until the car restarts and storing a DTC.

Nick found that, given enough time to warm up (about 10 seconds?) his 35W ballasts + KBOWE could be switched to the 60% duty cycle and continue to operate - 3.8A is within the acceptable current draw. This bad for the ballasts and the KBOWE due to the additional current, so the LSM should always stay in position II.