Xoutpost.com

Xoutpost.com (https://xoutpost.com/forums.php)
-   X5 (E53) Forum (https://xoutpost.com/bmw-sav-forums/x5-e53-forum/)
-   -   Seeing PWM Data (https://xoutpost.com/bmw-sav-forums/x5-e53-forum/109612-seeing-pwm-data.html)

jsoto 01-11-2019 09:33 AM

Seeing PWM Data
 
I've never had a need to do this . Probaby a AWynn question

To do a quick down dirty test on a hot wire with PWM, can one just put a clamp meter on it and will I able to see it go the load go up and then to zero, etc.
I know I can do this with a MM....clamp it just a quick simple intrusive method without removing stuff. or is the cycling so quick, the clamp won't see it

upallnight 01-11-2019 10:04 AM

Quote:

Originally Posted by jsoto (Post 1151778)
I've never had a need to do this . Probaby a AWynn question

To do a quick down dirty test on a hot wire with PWM, can one just put a clamp meter on it and will I able to see it go the load go up and then to zero, etc.
I know I can do this with a MM....clamp it just a quick simple intrusive method without removing stuff. or is the cycling so quick, the clamp won't see it

Yep, you need a scope to see the waveform.

andrewwynn 01-11-2019 12:06 PM

Seeing PWM Data
 
My fluke DMM has a duty cycle setting on the voltage setting that will get the results of PWM (it assumes positive so reads inverse for ground controlled pwm but you could hook up backwards).

Eg if pwm is on 80ms off 20ms it will read 80%. It doesn't show the details of the waveform just how long it's above vs below the average.

If you have a DMM that says "duty" on it that's the purpose.

The other quick and dirty test is an LED voltage tester. If the pwm control circuit can put out enough current to drive the led it will get dimmer and brighter to indicate the pwm control is lower and higher duty cycle.

jsoto 01-11-2019 12:30 PM

LOL. UPA. A scope is something that's always been on my list but I won't use it that much...well, that's the same for the other bajillion tools I got I suppose. I got 2 tile saws. Bad habits I suppose.

I would be somewhat interested to see the data on what it looks like with parking lights on and then what the waveform looks like when the flashers are engaged.

And are both values fairly static or there are other variable waveforms one seens on the above 2 scenarios.

Zulu95 01-11-2019 01:54 PM

I have bench scopes but back in 2004 I bought a Vellman Personal Scope 10MS/s just to check out the things you mentioned. They ran around $300 locally but a lot of judicious web searching turned one up for around $90 at the time. I still have it tucked away for quick and dirty testing "in the field" should the need arise.


andrewwynn 01-11-2019 03:21 PM

You're curious what waveform the LCM sends to the corner lights I could probably fire up my scope meter and see what it sends.

I noticed that the two filament bulbs only use the high output filament but at lower voltage. In the back half the red lights use PWM half use both filaments very strange. I have 01. I think after 03 they all use only PWM.

I put lights in for turn signal I had to go back to Incan in front because the led won't come on for running light.

In the back it works but I get fast flash because low power. I did notice when I use hazard flash that bulbs actually blink really fast eg PWM activity going on. I believe LCM uses pwm to clamp output to 12v. Not sure why I don't see the pwm blinks during normal turn signal. Maybe that's allowed to run full voltage.

wpoll 01-11-2019 04:25 PM

Quote:

Originally Posted by andrewwynn (Post 1151806)
You're curious what waveform the LCM sends to the corner lights I could probably fire up my scope meter and see what it sends.

I noticed that the two filament bulbs only use the high output filament but at lower voltage. In the back half the red lights use PWM half use both filaments very strange. I have 01. I think after 03 they all use only PWM.

I put lights in for turn signal I had to go back to Incan in front because the led won't come on for running light.

In the back it works but I get fast flash because low power. I did notice when I use hazard flash that bulbs actually blink really fast eg PWM activity going on. I believe LCM uses pwm to clamp output to 12v. Not sure why I don't see the pwm blinks during normal turn signal. Maybe that's allowed to run full voltage.

I would do the same (fire up my 'scope and check the PWM freq. and duty-cycle etc.) but my signal lights are only that - signal lights, so I can't help much. :(

I do know that the signal lamps are under full PWM control - they glow brightly as signals but glow dimly when a halo fails and they get switched on in a "replacement" role until the halo lamp is replaced. This is on the LCI car - not sure what pre-LCI does.

Interesting comment, Andrew, about the signal having two filaments. The LCI cars have only one (or at least my Euro-spec one does). Same for the tails too, as has beed discussed before - I just didn't realise the front signals were dual filament too (on pre-LCI). I must look more closely at the wiring diagrams for the pre-LCI cars sometime... ;)

andrewwynn 01-11-2019 04:37 PM

The rear run: brake also dual filament but at least one lamp didn't use both filaments. I re-wired my rear brake lights so the unconnected rear brake filament is wired now so I have four brake lights vs two now very bright.

The normal brake light runs dim via pwm and is almost exactly as bright as the run light that only uses the low filament not sure if pwm or full bright. It's slightly different brightness than the pwm high filament on the normal brake light

wpoll 01-11-2019 05:07 PM

Just been looking at the wiring diagrams (via newtis.info) for the front signals on the pre-LCI... jeepers, they are somewhat confusing. There doesn't seem to be a single diagram that shows the entire assembly - it's all spread across several diagrams. And then there's the year variations!

I think part of the problem is terminology - signals, side lights, turn signals, parking lights - they use all these terms, and then we throw in angel-eyes (AEs) and halos to the mix! :confused:

To come back to the OP's question, anytime PWM is involved, the system is designed around incandescent lamps (bulbs), using the slow reaction time of a glowing filament to "average" the PWM pulses into a consistent brightness level. The lower the PWM duty cycle, the lower the average (but not actual) voltage on the filament.

When you put LEDs in the circuit, replacing the incandescent lamp, a couple of things happen. Hot and cold monitoring get screwed up - and they are different things - and can cause warnings about lamp failures on the dash. The fixes included coding out hot and cold monitoring or inserting resistors in parallel with each LED device to provide a sufficient load for the monitoring systems.

These resistors negate part of the advantage of LEDS - lower power consumption. Putting resistors in circuit to fix the hot and cold monitoring brings the power consumption back up to that of incandescent lamps.

Also, on lamps that run continuously, the resistors can get VERY hot, so watch out for that - place them carefully!

Some LEDs are advertised a being "CAN bus" friendly, which is technically incorrect, as they probably just ship with resistors.

The second problem LEDs have in PWM lighting systems is that they switch on and off VERY fast - so they do not provide the pulse averaging that the incandescent lamps do. The resistors mentioned above (if fitted) can act like a pulse averaging circuit and help somewhat, as can capacitors, but the level of dimming on an LED is not the same as on a PWM-modulated incandescent lamp, so it may or may not work as intended. This isn't an issue on a lamp that is either ON or OFF but it may be a issue on a lamp designed to run at different brightness levels in normal operation - e.g. the tails on LCI cars are low brightness as tails (low duty-cycle PWM) and brighten when the brakes are applied (high duty-cycle PWM).

All of which is to say, LEDs can be made to work, and work well, but they aren't ALWAYS a better option and you may need to modify a bunch of things to get them to work correctly, especially in a system designed for (and that relies on) incandescent lamps.

And as I said, getting back to the OPs question, you are likely going to need to add resistors to the LEDs in question to deal with the issues you are seeing... I doubt goofing with the PWM duty-cycle will sort it :dunno:

oldskewel 01-11-2019 05:08 PM

You're talking about one of the mysteries of my 2001 3.0i that has puzzled me ever since I came across it. All my lighting is stock. No LEDs retrofitted.

Why does BMW specify dual filament bulbs (specifically for the front turn signal / parking lights) and then only send two wires to the bulb? I've asked that a few times on here, never with a real response. Any thoughts?

I did not check to confirm whether both filaments light up when installed and lit. Confused, but can live with that.


All times are GMT -4. The time now is 07:57 AM.

vBulletin, Copyright 2024, Jelsoft Enterprises Ltd.
SEO by vBSEO 3.6.0
© 2017 Xoutpost.com. All rights reserved.