Wet Manifold and Typhoon Assistance
Turbo Mini Forums

I know some people have some experience with the Typhoon and after a few months of struggling I need help. My installation is nothing fancy, but required a few accommodations to be installed in a MK I Traveller shell.

It is a rebuilt MPi engine using the Specialist Components wet manifold throttle body. It runs, but doesn't appear to be running well. We had a number of issues, but I have worked through most of them. It starts and is drivable at this point, but I cannot really tell what is happening under load while driving.

I initially was trying to use an Innovate LM2 for AFR, but I switched to using an SC provided wideband sensor which is actually a 14point7 controller. This at least allows for the values to be viewed in the same interface. It seems the sensor configuration is not right as the lambda value is bouncing all around.

Here is a video of the software interface with what it is doing at idle:
https://youtu.be/xciHe724APU

I am thinking the Lambda Sensor Input Scaling table is incorrect.
It only has two values
0.00 = 0.50
4.45 = 1.50

It almost seems like the wide band acts like a narrow band, maybe when it is in closed loop it does that, going high and low.

I have checked for an inlet leak, but have not found one. The wiring is little trickier as we are using the SC provided MPi harness with modifications for use in a MK I. I was thinking noise or ground was an issue possibly.

There is a lot more to the story, but that is a start. Before I can sort out other issues I need this to work at idle or at least to be consistent.

The Typhoon also has a "quick fuel" mode where you can copy the fuel value from the % correction into the table, but I do not seem to be able to get it to work. Maybe because the sensor output is not stabilized, the quick fuel mode is not working.
Terry

Yep if I remeber correctly you need to extend the table, it was a key stroke to edit the table values.

I run a supercharger and I don't care the TB is on the wrong side.
VEMS + 12 PSI + Liquid Intercooler = Small Bore FUN!

Thank you, do you remember what values you added? This is a 4.9 sensor, I would think I could look it up in the specs that have been posted for the 4.9. One would also think the vendor might provide those values, but that is another story too...

I would ask sc...

I run a supercharger and I don't care the TB is on the wrong side.
VEMS + 12 PSI + Liquid Intercooler = Small Bore FUN!

Asking SC is another part of the story. They were very helpful at the outset, but less so lately. They even connected remotely back in November and made some adjustments, but that was when the narrow band was installed. I thought purchasing the wide band through them might encourage a response, but that has not happened. It has been suggested that I need to call them...as numerous emails are not responded to. I may have overdone it at some point with the emails...

Edited by tmsmini on 24th Apr, 2016.

Just checked again and while on the Lambda Input page live, I see the voltage fluctuating as well which I think means that even if I had more or different reference points, the sensor output still appears to be fluctuating greatly.

If am reading correctly from the 14point7 documentation, these are the values that should be in place:
Green Linear Output
Interfacing device; ECU/Gauge/datalogger/etc...
0[v] @ 0.68 [Lambda] Linear to 5[v] @ 1.36 [Lambda]
, equivalent to 10-20 [AFR] for gasoline fuel

Now I am wondering if the temp sensor scaling is correct as well. I am using the Rover temp sensor, but the stock SC lit uses something else.

Edited by tmsmini on 24th Apr, 2016.

Unless they have a WBO2 controller with a custom calibration different from what is used on all other 14point7 products, the calibration you initially mentioned is way off. The one in your last post is the one that is used in the normal 14point7 products. And looking at the picture you had sent me, it does look like the standard Spartan 2 which I doubt would have a custom calibration.

Using the wrong calibration would make any small AFR variation look much bigger and trying to use it for closed loop would not give good results not to mention that it would not give you the AFR you want.

And from that, I too would be suspicious of any calibration they provide. Fortunately, a temperature sensor is a lot easier to check with room temperature and boiling water.

http://www.jbperf.com/

maybe take it off closed loop so its not chasing its tail too?

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

Why did you swap from the LM-2?
Apart from wanting a common interface, what was it reading before the swap to an SC supplied 14point7 controller.

Assuming any form of closed loop control is currently off (which is the best starting point until you have it running basically OK) were they both doing essentially the same?

That youtube vid could be something as basic as a misfire which the 14point7 controller is reading correctly (if it was a misfire it could also explain the poor running).

To check the integrity of the sensor, wiring, controller etc. just unscrew the sensor and with everything powered up (engine not running) just blow some butane gas gently over the sensor and watch the gauge in your interface - it should run smoothly from lean to rich as you pass the gas over it and go back to lean (free air) once you stop blowing the gas - for the gas use an UN-lit plumbers blowtorch or cigarette lighter.

If it moves smoothly when you artificially run lean to rich, chances are what you are seeing is correct and it's just showing the underlying problem.

Calibration, as per Jean, the later figures look right, not the first ones you were given.
I've used 14point7 stuff quite a lot (albeit the digital stuff rather than analogue) and I've not seen erratic behaviour like that video when the wiring and sensor connections are sound and the engine is running basically correctly.

Is/are the readings doing the same under load or is this just idle at the moment?

Schrödinger's cat - so which one am I ???

I am afraid I have been spinning my wheels so much that I have lost any logic. The whole process was going to be logical and methodical, but I lost that along the way. Probably when I took the suggestion to "just buy" the SC system and get the car on the road.

I have one of Jean's MicroSquirt based systems and if I had spent the time on it that I have on the SC system I would probably have a working running system with idle control.

I have tried both closed and open loop and even though it does seem to chase its tail, it seems to run better on closed loop, even with the scaling.

My goal at this point is to get it running well enough to get it back to the rolling road. I did have it initially setup at the same rolling road. Unfortunately there were engine and installation issues that I did not expect with a new engine that happened at some point. These probably got in the way of the operator getting a good setup. I believe the issues have been corrected.

I have had a range of issues with the LM-2. The tach reading is worthless with the induction clamp. Leaving the system in place with the extra cabling required was not practical as I could not leave the sensors un-powered. The reading have been inconsistent with it even on other cars. Some of that could be down to using the tailpipe clamp, but even with the sensors in the LCB, there were odd values. I did upgrade to a 4.9 sensor and cabling for the Innovate, but I have not used them yet.

The idle is actually very smooth and even driving it not bad. What I am basing my "not running well" statements on is the AFR readings.

I need to do another run on the road under load to give me a fresh baseline.

I do have two ports in the exhaust to make use of two sensors. Maybe the sensor ports are too close to the head. They are in the location used by the Maniflow LCB for the MPi. Is that too close for a wideband?

Thanks for the comments so far.

If the AFR readings are not matched by equally poor running, I would suggest pulling the sensor and trying the butane gas to see if the erratic behaviour is a real engine reading or just a wiring issue or a setup or incompatibility issue with all the hardware.
The readings should be nice and smooth up and down with plain butane.

Just one extra point, you say upgrade to 4.9 sensor. Is the 14point7 controller you got from SC the correct one for a 4.9 (ie, when did you do the "upgrade" ?). The 14point7 controllers are 4.2 or 4.9 specific, not interchangeable (well the ones I use certainly aren't interchangeable).

If it's a N/A setup, the sensor is highly unlikely to be too close but, from your description is it/they one just one of the manifold legs (ie just inners or just outers) or the common bit (I'm not familiar with the MPI setup).

Schrödinger's cat - so which one am I ???

Good point about the 4.9. I've had a 4.2 controller sent to me by mistake and with a 4.9 sensor connected, the reading was all over the place.

As Rod says, the 14point7 units are sensor specific and you can only use the one the controller was programmed for. The code is different and there is a very small hardware difference also.

By the way, this is something that is common in many different controllers, not just the 14point7 ones.

http://www.jbperf.com/

Good question about the 14point7 controller. I need to check and verify. As the connector is different for the 4.9 sensors I did not think there could be a mismatch. I just bought the 14point7 controller and sensor from SC a few weeks ago.

There is a firmware update for the Innovate LM-2 that makes it 4.9 compatible. I think my Innovate LM-2 may have been defective from the beginning. At the end of today's testing, this is what it is showing when on the lambda setting:
https://youtu.be/G4XaF8HpZEg
Maybe I need to retry the firmware update.
I hooked up the Innovate and the SC systems at the same time. The Innovate was in the tail pipe and the SC 14point7 in the center branch. Both readings appear to have issues:
https://youtu.be/NqC83z21T2U

I think the fueling is probably better than I thought, but both wideband systems have issues.
I think the Innovate is defective and I either have a wiring issue or a configuration setting wrong for the 14point7 with the SC Typhoon.

I meant to check with butane before I got started today while the engine was cool. Something to check tomorrow after work.

Edited by tmsmini on 25th Apr, 2016.

If anyone wants the SC Typhoon SX-Tune software help documents in a printable format, it is attached.

I am old and like to be able to read while I am trying to make an adjustment instead of juggling multiple windows on one computer.

Attachments:

3.02mb

Schrödinger's cat - so which one am I ???

Interesting question about the 4.9 sensors. The one from SC is a D2P, but at least has a sticker on it.
http://d2pautoparts.com/all-car-parts/exha...nd-sensors.html

The one from Innovate has no markings at all which I found unusual as it claimed to be a Bosch sensor. I can't believe that Bosch does not have either a lot number or some other QC identifier. I asked and was told that is the way it came.

The fist 4.2 kit I had from sc was faulty. They never believed me, they said they have never had a faulty one and they test all of them before they leave.... The only way I could get them to return it was by stating that I would post it back, and pay for them to test it and return postage if it wasn't faulty. The upshot I sent it back, I never heard anything and a new one magicked itself In the post....

I run a supercharger and I don't care the TB is on the wrong side.
VEMS + 12 PSI + Liquid Intercooler = Small Bore FUN!

Thanks for that Graham. I hesitated to post due to issues with previous threads on the forum. I had asked about an adapter to make a wideband work with their harness that I am using. Eventually I found on their website store there was a drop down where the MPi style connector could be selected.

I was a little surprised when it came with a 14point7 Spartan controller. I did send them another email earlier today.
Again, thank you for the responses. The car is very close to being finished..if you ever really finish.
Terry

Re. the markings on the 4.9 sensors, back at home I think I still have Graham T's failed 4.9 sensors. I'm still out of the country at the moment and can't remember if I posted them back before I left....

Different Graham (but we are both using a very similar MS2 setup so we share a lot of notes) but he had a couple fail in quick succession on Innovate controllers (LC-2 or LM-2, can't remember which) so he posted them to me to test on my 14point7 controllers. Mainly because the 14point7 controllers I use read the cell working temperature so you can see if the heater in the sensor has failed (one of the Innovates was suggesting that was the fault as I remember).

Anyway, back to the point, when we compared the 4.9 sensors we had between the two of us there was at least two, possibly three, different markings/stampings on the bodies of the sensors, none of which gave any real confidence they were "genuine".

IIRC (and as I'm out of the country I can't be 100% sure whether it was 3 or 4) they were from 3 or 4 different suppliers.
A supposed genuine Innovate supplier in the UK, a supposed genuine Innovate supplier in Eastern Europe (both via eBay but claiming to be authorised), an eBay supplier who didn't claim to be Innovate authorised but said they were genuine Bosch and finally (I think) one from Alan.

Part of the numbering difference I think was different cable lengths, the rest is un-explained.

I can dig out some more detail when I'm back in the UK (or if Graham T sees this he may be able to add some more detail) but the fact I was testing (nearly new) failed ones is why I made the comment earlier that the 4.9 sensors seem more fragile.

Also, one that Innovate reported faulty and wouldn't heat up, heated up fine on my 14point7 controller - to exactly the right temperature and maintained it - but still refused to read, so a "failed" sensor can be interpreted different ways by different controllers.

Schrödinger's cat - so which one am I ???

I sent an email to Innovate and already received a reply. It is out of warranty, but they will evaluate it. I think it has had issues all along and they have just been getting worse. Initially I thought it was due to the use of the tailpipe clamp, as the values were not consistent. Even using the ports in the LCB manifold, the readings were inconsistent. And certainly there was a difference between inner and outer, but that is a different story...

Assuming the responses from 4.2 and 4.9 are linear which means they only need two values and the software interpolates from that, the attachment is the values in a simple graph.

Attachments:

43.17kb

The response from the sensor is not linear but the output from the controller is. So you indeed only need 2 values to fully define the output.

And the 14point7 controllers use the same output calibration regardless of the sensor used (as long as you use the correct one). Note that there is an exception for their lean burn product which has its own specific calibration.

http://www.jbperf.com/

Thanks Jean.
The more I think about it, it seems like there should be additional configuration settings when switching to a wideband in the SX-Tune software. There is the input scaling setting and the closed loop setting, but where is a setting for which sensor is in use if closed loop is off?

Both 4.9 sensors failed the butane test.
That leaves me with either a configuration error or a wiring error.
I think it is unlikely that both the sensor from Innovate and the sensor from SC are defective.

I am still not sure about the sensor from Innovate as it has no markings on it at all. At least the D2P has a part number on it- TFO1681964.

With minimal support from SC, this could get difficult. The sensors are getting hot so I know the heater relay is working. I will have to see if I have any info on the wiring harness they provided. As the cars starts, runs and drives, it seems like the issue is in the wiring for the sensor only.

I did notice that the software gauge in SX-Tune was stuck at 1.36, the value I have substituted for the 5 volt output. I am guessing this means the controller is outputting 5 volts.

So while I am waiting for the LM-2 to be returned and maybe a response from SC, I plugged the sensor ports and went for a 30 minute drive at speeds up to 50 mph. The car actually drove quite well.

Initially we had backfire through the inlet and dying at stop signs. This was after it came back from the dyno. I have worked through a few things including a bent valve on a new engine that I did not expect. So perhaps I have been stupidly cautious when I was seeing inconsistent readings from the gauges.

So while I am waiting I was thinking I could test the Spartan 2 and the 4.9 sensors off the car.

How would I use a multimeter to measure the output? The heater circuit is straight forward, but the 14point7 directions say to wire the 5 volt ground wire to the same ground as the gauge.