Using Oscilliscope to Set Duty Cycle

[Funracer]

With the help a some extremely knowledgeable and patient gurus on this forum I have recently become acquainted with using a digital oscilloscope to check and adjust my duty cycle. There is a huge amount of info here on how to do this with a DVM or Dwell Meter. In my case I tried to accomplish this dozens of times over many weeks (in between the rest of my life) and could not get readings that made any sense. If you are having a similar issue setting yours don’t be afraid to try an oscilloscope.

Let me say before I review the one I bought that, unlike many on this board, I know nearly nothing about these instruments and welcome all corrections and additions to anything I put forward. Also I dealt exclusively with my Lambda equipped 83SC and I am not sure how much of this applies to other years, so proceed with that in mind.

mysocal911 and AndrewCologne both put forward that the frequency valve (FV) itself can interfere with some DVMs trying to read Duty Cycle (DC). I had two new DVMs and neither would read DC. After a series of tests verified my Lamda ECU was working ok, I started looking for a cheapo digital oscilliscope. I paid $42 for this one from Amazon:



Only the size of a pack of cigarettes, this a very basic unit that comes with only the things in the picture and a simple 2 page user manual. You can get scopes with lots more options, but the price goes up accordingly.

Their are a few options to keep an eye out for. The biggest thing to watch out for is that lots of these cheap one are in kit form that you have to put together. Be carful you don’t buy something you have to solder together unless thats what want. This one came pre assembled and ready to use out of the box.

This one has a color screen, some don’t. Among them are three ways to power the device. Some use batteries, some have a clip that you can use with an in-car 12V source like the fuse panel in our cars, but the one I have uses 120 VAC and the included adapter to get 12 VDC.

Some of the digital scopes don’t have a screen. They capture, say, 5 seconds of data and then you view the data on your laptop. The problem with this for me is the ones I saw only worked with Microsoft products. All I have are Ipads, so I opted for one with a built in screen.

Once you receive the scope and attach the (in my case) shortish alligator clips and turn it on you get this:


Now start playing with the black buttons and white knob under the screen until the parameters are set like those in my pics. 5V DC 5ms Auto. Others can tell you more about these adjustment knobs but these are the settings that worked for me.

Attach the clips as you would with a DVM, black to green wire in the test port and red to ground. This will result in an upside down graph as shown here. With the clips attached like this you can turn your 3mm mixture screw slowly (warm engine, 1500 to 2000 rpm, O2 plugged in) until you read Duty of 45% (or whatever you want to set it) at 4L on the scope screen.

I was measuring something else in the pic below but you get the idea. I was so excited to be able to finally set mine accurately that I forgot to take a picture of the actual event.



There are lots of other things you can do with an oscilliscope, but I am just learning how to use it myself. Still not sure exactly why I cannot use a DVM to set my DC, but now I don’t care. Hard to go wrong for less than fifty bucks.

Thanks again to a those that helped me work through this.

Regards

[FrankM_]

I have the same ‘DSO shell’ scope and it works fine in the low frequencies (anything under 75 kHz) but the measurements shown on screen can be seriously off (voltage/ frequency figures). They tend to depend on the timebase you have chosen.

But is is super-handy, small, you can check a lot of signals with this (checked my afm and found discontinuities, isv controls, flywheel sensor outputs,...)


Sent from my iPad using Tapatalk

[dhanl82]

Thanks for posting this information, I looks like a very good tool to use in adjusting the mixture and in verifying that the various CIS enrichment functions are operating as they should.
I ordered one, for the money it is a great deal!

Thanks again.
Dave

[AndrewCologne]

Actually you don't need an oscilloscope for just reading the ECUs/FVs duty cycle.
A simple DMM with frequency and duty cycle read option will give you the needed result as well.
But the oscilloscope in this application gives you a nice plus by showing signal interferences, spikes and ... noise.

BTW: You did connect the oscilloscope the wrong way. The terminals must be connected the inverse way, as the ECU's duty cycle affects the ground signal of the FV. Thats why you see the negative output values on your display.

:-)

[2jmotorsports]

I used the same scope to set my DC after I found the multimeter reading to be a little bit too jumpy. What a great little tool.

It is interesting listening to the engine and the subtle cyclical idling once you get the DC dialed in.

[Funracer]

Originally Posted by AndrewCologne

BTW: You did connect the oscilloscope the wrong way. The terminals must be connected the inverse way, as the ECU's duty cycle affects the ground signal of the FV. Thats why you see the negative output values on your display.

:-)

DMM is set black to green wire at the test port, red to ground and look for 45%.

You are saying to hook up the oscilliscope differently, red to test port and black to ground and set 45%? The difference is not huge but I want to be as accurate as possible.

Thanks

[AndrewCologne]

DMM is set black to green wire at the test port, red to ground and look for 45%.

Same with the oscilloscope – you see negative values in your oscilloscope output, so you connected it wrong.

To test a correct DMM/oscilloscope connection, just lead the wire of the 15c Switch (right chain housing) to ground and you should read a 65% duty cycle (and not 35%).

[tirwin]

I thought some DMMs report differently — i.e., some report ON time and some report OFF.

An oscilloscope has been on my list for many years. I found a meter than has an oscilloscope-like display. It’s handy but the screen resolution is poor. It reminds me of my HP 28S graphing calculator I used in college.

I always wondered if these cheap oscilloscopes were good enough for most of the stuff I’d use it for on the car. Looks like they probably are.

Thanks for sharing.

[Funracer]

Originally Posted by AndrewCologne

To test a correct DMM/oscilloscope connection, just lead the wire of the 15c Switch (right chain housing) to ground and you should read a 65% duty cycle (and not 35%).

Great idea. Thanks

[Funracer]

Originally Posted by tirwin

I thought some DMMs report differently — i.e., some report ON time and some report OFF.

Yes I thought that as well. Due to the minuscule set of instructions and specs that came with mine I am not sure whether its reading the “ON” or “OFF”. Andrews suggestion is the way to verify a proper hook up. From my thread readings I thought we were setting the OFF time to 45%.

So whichever way gives a 65% reading (not 35%) with the 15C switch grounded is the way to go. It’s possible I may have set mine incorrectly to the inverse of 45% which is 55% (100%-45%=55%)

Due to the holiday weekend it might be Monday before I can hook up the scope and reset if necessary. I will update then.

Happy 4th of July!

[AndrewCologne]

Originally Posted by tirwin

I thought some DMMs report differently — i.e., some report ON time and some report OFF.

An oscilloscope has been on my list for many years. I found a meter than has an oscilloscope-like display. It’s handy but the screen resolution is poor. It reminds me of my HP 28S graphing calculator I used in college.

I always wondered if these cheap oscilloscopes were good enough for most of the stuff I’d use it for on the car. Looks like they probably are.

Thanks for sharing.

The DSO150 type as shown above is also available as ready assembled offer and is a nice choice for minor checks. But be aware, if youre expecting more than just do a simple duty cycle test on your car, you should think about a better brand/Model device. Or at least add some bucks and get a Hantek 6002BE, here in Europe for about 60-70€. its a device you connect via USB to your laptop which beside a lot more options/features it is far more comfortable.

[AndrewCologne]

Originally Posted by Funracer

From my thread readings I thought we were setting the OFF time to 45%.

Nope its the "on" time.
Because 45% let set you the CO screw to a higher CO value.

duty cycle is just a regulator/compensator. Means if the mixture is too rich, it reduces the "on" time.
So a 45% base setting while idling with running lambda control and a hot engine gives you a richer mixture at acceleration and WOT, means when the throttle is opened above 33° angle. And also this affects the cold start wich could result too rich.
In my cpu based ECU I offer, beside the common control you can set the cold start and the acceleration duty cycle independently.

If you stricly follow the factory default by using a CO meter and set the CO to 0.4-0.8% with disconnected sensor plug. Finally with sensor plugged and lambda control running you will get a mas on menos duty cycle of about 60% when idling. Cause 0.4-0.8 CO is significantly leaner than Lambda 1 or an AFR 14.7:1

[Funracer]

Originally Posted by AndrewCologne

If you stricly follow the factory default by using a CO meter and set the CO to 0.4-0.8% with disconnected sensor plug. Finally with sensor plugged and lambda control running you will get a mas on menos duty cycle of about 60% when idling. Cause 0.4-0.8 CO is significantly leaner than Lambda 1 or an AFR 14.7:1

I just ran the scope with the 15C switch grounded. With black to test port and red to ground, it reads 65%.

But I am confused a bit by your statement above. With warm car and everything plugged in, should I now set to 45% or 60%?

Thanks

[AndrewCologne]

You can set your duty cycle at a warm idling engine at 45% or even 35% --- as long your cold start runs flawlessly.

The logic is: a 50% duty cycle means no affecting of the base CO setting needed, the base CO setting matches with lambda 1.

If you set the base CO to factory specs then when engine is running with lambda regulation a duty cycle of 60% should be the result when idling. Cause 0-4-0.8% CO results higher than Lambda 1.0 means leaner CO.
The 45% idling duty cycle just means that you obtain a higher enrichment at WOT which you just should check out.

You just should find out your best idle duty cycle. ... I personally would start with the factory specs where an idling duty cycle of 60% would be the result. And than compare to an idling duty cycle of 45%. If you don't feel any difference, keep the factory spec. Cause a duty cycle as close as possible to the factory spec makes your combustion less soot/carbon resulting as well as the spark plugs.

The big problem i.E. when people just unplug the Sensor for deactivating the Lambda control is a very fat resulting combustion at almost all revs with all upcoming side effects seen in a long term. Blow by gas, cylinder washing out effects, gas smelling oil at oil changes etc etc etc. Its no a myth that many good maintained lambda US engines do run far more miles than non lambda ones.

[Funracer]

Originally Posted by AndrewCologne

You can set your duty cycle at a warm idling engine at 45% or even 35% --- as long your cold start runs flawlessly.

This is my exact issue. Car runs perfectly smooth with great idle after its been running for about 5 minutes.

Starts instantly but then idle drops to about 4-500 rpm surges back to 1000 a few times until it warms up. Warmer it gets the smoother the idle. If I jiggle the throttle while it is cold surging it sometimes backfires through the POV.

I will richen it a bit today and see what happens.

Danka (thats all the German I know)

[AndrewCologne]

If your car with 45% at warm engine temp idles ok, but runs at cold start very jerky and maybe lots of soot comes out, then a higher enrichment via CO screw and therefore a resulting duty cycle at warm idle of even below 45% makes it even worse.

But you can try ... that's what I meant -- you should hit the right CO screw setting spot for your car.

Danke ... for your "thanks" :-)

[tirwin]

Originally Posted by Funracer

This is my exact issue. Car runs perfectly smooth with great idle after its been running for about 5 minutes.

Starts instantly but then idle drops to about 4-500 rpm surges back to 1000 a few times until it warms up. Warmer it gets the smoother the idle. If I jiggle the throttle while it is cold surging it sometimes backfires through the POV.

I will richen it a bit today and see what happens.

Danka (thats all the German I know)

Funracer,

This is one of the things that is most frustrating about troubleshooting CIS. The overall system has components that work in isolation of one another. For example, if the AAR were stuck closed on a cold start, then the WUR doesn't know or care. It has no means of compensating for the fact that there isn't enough extra air to go along with the extra fuel it is delivering on cold start.

It sounds like you could be too rich on cold start. All else being equal, if you add extra air, the rpms will rise. If you reduce air, rpms will fall.

If the cold control pressure is in spec, then a low idle would suggest there isn't enough extra air on cold start resulting in an overly rich condition during the cold start period.

Do you have a wideband AFR by any chance? Something like the Innovate LM-1 is really handy in these situations. The duty cycle should be ~65% ('81-'83 US I believe) until the engine warms up. (I think you have a handle on this.) If you're overly rich on cold start it would show up here on the wideband O2. Then it would be a matter of tracking down the cause. Once the engine is warm, the lambda kicks in until the throttle switch is tripped. If you disconnect the CIS narrowband O2 sensor you should see a fixed duty cycle of 50%. (note: there are some difference in duty cycles defaults over model years - I don't have it all in front of me right now so please verify for your model year. It's in the Bentley and the factory manuals).

[AndrewCologne]

The first approach would be just lowering the initial CO to factory specs, means 0.4-0.8% ... then he should check how the cold start changes. Many people do suggest unplugging the lambda sensor and setting the initial CO to 2% or even higher which could really mess up the cold start

[tirwin]

I don’t think most people have an exhaust gas analyzer anymore so that’s why I suggested the wideband approach.

I’ve personally never had a cold start issue with a higher idle AFR.

[AndrewCologne]

Yes, theWideband sensor way is also a very precise option, but If no analyzer is present then he still can easily set the duty cycle to approx. 60% on a warm engine at idle with lambda running, where this corresponds to a basic CO setting of 0.4-0.8%.

I’ve personally never had a cold start issue with a higher idle AFR.

Exactly thats what he should try ... cause the approach above with a 60% duty cycle at idle makes the value of the cold start AFR higher.
In my case I even modified my WUR so that after 2.0 bar when starting it reaches the 3.6 bar quite fast which gives of me a perfect mixture at cold start with no jerky driving etc.
Also its possible to lower the cold start idle rpm with a "plug with hole" mod of the AAV where mine just starts with 1.100 rpm – the neighbors love it as no more "comments" regarding pollution and "noise" to the environment are addressed to me. Don't know how this is seen in the US but here in Germany many "cars are evil"-people changed to crazy ones.