Hellen NB1 RusEFI

Your chance to introduce yourself and your vehicle
Post Reply
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Hellen NB1 RusEFI

Post by OSprey42 »

Installed a Hellen NB1 ECU today on a RHD 1999 NB1. The Installation was incredibly easy as it fits directly into the OEM location and 3/4 OEM mounts which is still enough to secure it. All that needed to be done was to install a vacuum line on the back of the manifold and a Bosch LSU 4.9 wideband O2 sensor directly to the ECU options port. I really like how the wideband goes to the options port so that no modifications to the OEM harness need to be made. Still learning how to tune it but after playing with the VE table it drives moderately well although there is still room to refine the tune a lot. Startup is currently pretty terrible and needs some persuasion from the throttle but still learning how to dial that in.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hellen NB1 RusEFI

Post by AndreyB »

Welcome to the forum!

Your positive feedback is marketing material ready to go! :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
BeerMoneyMotorsports
Posts: 151
Joined: Thu Jan 14, 2021 10:16 pm

Re: Hellen NB1 RusEFI

Post by BeerMoneyMotorsports »

Thanks for posting! share some pictures and videos!
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

Attached are some photos of the install. Namely the ECU connected as well as the vacuum and O2 sensor lines being poked through the firewall. The supplied base map is good enough to get the car up and running however I found that with the 4 tooth trigger wheel the car struggled to start. It was reccomended to install a 36-1 trigger wheel (I used Mazda part # ZM0111408 which was reccomended by Beer Money Motorsports who produces the Hellen NB1 ECU). After configuring this the car started as well as stock after a bit of cranking tuning. See my first tune with the 36T trigger wheel here https://rusefi.com/online/view.php?msq=706.
The next stage after getting the car to crank was idle and tuning the fuel maps. I decided to leave spark alone at this stage.

After working on the VE tables a bit and trying to work on the idle settings I got to this tune: https://rusefi.com/online/view.php?msq=708
Getting the car to idle well has been difficult and I have not been able to make much progress on improving it to date from the base tune supplied with the ECU. The OL idle is easy to configure but the CL is difficult with the PID CL idle settings, idle VE and spark tables and CL spark timing table. I'm going to try tuning a bare bones PID idle then add features on from there at this stage. The main other issue I have with the tune is that the car bogs when blipping the throttle from idle and I'm not sure if it's due to a terrible VE table or idle settings leaning it out too much. I'm quite new to tuning at this stage so I'll try and figure it out as I learn more about tuning. I mostly used the autotune for the VE after using a base target lambda and VE from another NB but the values are apparently terrible as I was unaware that autotuner requires settings like the fuel trim, acceleration enrichment and DFCO to be disabled to get good values. The wideband O2 sensor also has an issue where it occasionally just goes to 0 lambda in softare and only comes back after power cycling the ECU. Not quite sure if this is a wiring issue, ECU hardware issue or software bug.

Follow me as I eventually work out how to make this thing run!
Attachments
20211031_182609.jpg
20211031_182609.jpg (3.83 MiB) Viewed 9612 times
20211031_144247.jpg
20211031_144247.jpg (3.85 MiB) Viewed 9612 times
20211031_144237.jpg
20211031_144237.jpg (3.44 MiB) Viewed 9612 times
20211031_144231.jpg
20211031_144231.jpg (3.68 MiB) Viewed 9612 times
Nightsknight45
Posts: 41
Joined: Thu Oct 28, 2021 2:31 am
Location: Tampa FL

Re: Hellen NB1 RusEFI

Post by Nightsknight45 »

OSprey42 wrote:
Fri Nov 05, 2021 7:49 am
The main other issue I have with the tune is that the car bogs when blipping the throttle from idle and I'm not sure if it's due to a terrible VE table or idle settings leaning it out too much.




My car does this as well! Do let me know if you find a solution within your tune to remedy this.
BeerMoneyMotorsports
Posts: 151
Joined: Thu Jan 14, 2021 10:16 pm

Re: Hellen NB1 RusEFI

Post by BeerMoneyMotorsports »

Hey Oliver. Turn the data rate to max in Tuner Studio and take a log of the bog. lets see what it is doing. could be ve or accel enrich.
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Hellen NB1 RusEFI

Post by stefanst »

Blipping the throttle from idle is one of the hardest things to tune in accel enrichment, especially when using MAP on a small engine.
Fortunately, it's not really an important area- just roll on the throttle when you get going and you will never experience it during regular driving. Fix your VE table first, then tune accel enrichment for all normal driving conditions. If it's still there, you can fix the blip-problem with the "TPS/TPS" enrichment.
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

Nightsknight45 wrote:
Fri Nov 05, 2021 3:13 pm
OSprey42 wrote:
Fri Nov 05, 2021 7:49 am
The main other issue I have with the tune is that the car bogs when blipping the throttle from idle and I'm not sure if it's due to a terrible VE table or idle settings leaning it out too much.

My car does this as well! Do let me know if you find a solution within your tune to remedy this.
I nicked your lambda and VE as a base so it could be an issue with that. In my most recent tune from today I found that it improved a bit by increasing the fueling for idle RPM above idle but I need to play around with accel enrichment. I also increased the VE% in the idle VE table at the higher RPM so the lambda richens out earlier as RPM starts increasing. I was focusing more on idle in my tune from today and only spent a few minutes on this but it's not as bad as it was. For me it's now more of a hesitation than a massive bog. I've attached a log to this post because

This is my most recent tune https://rusefi.com/online/view.php?msq=709. The log is attached to this post. At the time of writing the tune isn't showing up in the menu of tunes I can assigne the log to.

I spend a lot of time on idle in this tune. Started by disabling the idle spark table, leaving idle VE on so it runs lean to save fuel (which I want) and disabling CL idle timing. Also took all the cells in the spark map used for idle with and without AC on and made them all 13deg. First I set the idle to OL and got that working well with shown values in tune. Then I turned on the full OL+CL and messed around with that. Iterm min/max value is pretty arbitrary as it never reached either. Min was set to limit maximum speed which PID controller can drop RPM and max was set so that the IAC would never reach that value. For the PID tuning approach I set I,D to 0 and messed around with P until it was a decent mix of overshooting target RPM and stability (too high P results in instability). Then I tuned I until the engine could acheive target RPM. There was a bit of time in varying PI gains to get a stable controller that wasn't too slow to respond and didn't fluctuate around target RPM too much. From there I also found that I needed to reduce OL idle % a bit to give the PID controller a better point to start from. Getting this right seems to be a key in getting a fast PID tune otherwise the CL controller wastes time adding or subtracting from the OL IAC % before it changes it enough do do anything. Lastly I added a bit of derivative to smooth out some slight idle fluctuations but note that too much D makes it really slow and sometimes won't allow RPM to reach target if it's too high. Then I messed around with all PID values a bit to try and refine it. I did most of my tuning with the AC on because its summer soon in Australia. Importantly the RPM deadzone needs to be lowered for the PID controller to be easier to tune. I chose 0 so that any slow fluctuations could be tuned using I gain but the lower the deadzone the better the PID controller needs to be to get a good idle. My current controller isn't perfect but it's "daily drivable".

Next I'll look more into tuning the VE tables and the bog a bit more. Based on reccomendations, I may do the VE manually or use autotuner but with the CL fuel trim, DFCO, and idle sections blocked out as well as driving a lot more smoothly.

Also still seem to have the issue where the lambda sensor will go to 0 during a drive in TS on the laptop and the only fix seems to be power cycling and it comes back after 20+ seconds. Not sure if this is a wiring issue still or if it's a porblem with the ECU SW/HW.

The adventure continues!
Attachments
2021-11-06_17.20.33.zip
(75.39 KiB) Downloaded 153 times
Last edited by OSprey42 on Sat Nov 06, 2021 7:51 am, edited 2 times in total.
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

stefanst wrote:
Fri Nov 05, 2021 8:12 pm
Blipping the throttle from idle is one of the hardest things to tune in accel enrichment, especially when using MAP on a small engine.
Fortunately, it's not really an important area- just roll on the throttle when you get going and you will never experience it during regular driving. Fix your VE table first, then tune accel enrichment for all normal driving conditions. If it's still there, you can fix the blip-problem with the "TPS/TPS" enrichment.
Thanks, I'll take that advice! Now that the car starts and idles well enough I'll focus on VE first
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

I was having some issues with the wideband giving a highly unstable Nernst voltage as shown in this log https://rusefi.com/online/view.php?log=401. This was causing the lambda sensor to periodically go to 0 in TS until the ECU was fully power cycled. It was also making it hard to tune as an unstable Nernst voltage means that the Lambda values are likely to be misrepresentitive.

The issue was that I had unknowingly been given a fake Bosch LSU 4.9 WBO2 sensor. After purchasing an official one from a reputable and authorised Bosch dealer (should cost about $100 AUD give or take), this seemed to clear up the problem. Looking at this log the Nernst voltage is now a stable 0.45V just about. https://rusefi.com/online/view.php?log=417
BeerMoneyMotorsports
Posts: 151
Joined: Thu Jan 14, 2021 10:16 pm

Re: Hellen NB1 RusEFI

Post by BeerMoneyMotorsports »

Oliver, what happened to the rusefi sticker?
image.png
image.png (1.38 MiB) Viewed 9431 times
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

Today's Tune: https://rusefi.com/online/view.php?msq=726
Today I discovered that the base tune I was running up until this point had really high fluctuations on the alternator voltage between 12-16V which was making the headlights flicker so I went into the alternator settings and gave that a bit of a tune up. These were my final values:
Screenshot 2021-11-13 120631.jpg
Screenshot 2021-11-13 120631.jpg (75.33 KiB) Viewed 9363 times
P had to be dropped from 40 to 16 as the excessive P gain was the cause for the excessive voltage fluctuations. I had to increase offset from 0 to 20 as that seemed to give the controller enough of a kick to get the voltage up to ~14V. To help the controller reach a steady 14V I had to also be increased massively from 0.1 to 8. This gave it enough authority to pull the voltage to 14V. Changing the current load quickly by switching AC and headlights on I found that there was still overshoot up to ~15V so to remedy this I added a bit of D gain to damp it out.

These values aren't perfect but now the voltage manages to get up to 14V and stay there. Changing the load or RPM has a small effect on the voltage and it will usually drop up to 0.5V but after a few seconds it manages to go back up to 14V. Still some work to be done in tuning this but it's a lot better than it was.

In this tune I also fixed up the VE table a bit more as well as a few small tweaks to idle and ignition timing around idle. I also added in an IAT multiplier for the fuelling as the car started running lean from some hot under bonnet temps. It's getting more drivable but everything still needs tweaking.
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

Put a 70mm throttle body and a ported upper VICS intake manifold onto the car (smoothed out inside and port matched to throttle body). This meant that I needed to re-tune idle and VE again which is what I did in this tune. Using autotune and some smoothing the "hot spot" of maximum fuel delivery in the VE interestingly moved up from 3k to 4k RPM which could be a result of the intake changing the power curve of the car slightly (or it could just be bad tuning, would need a dyno to test). Taking it for a drive the VE table seems to stick within reasonable values. Idle is drivable but pretty sluggish and needs tuning. It has a tendency to drop below target RPM then slowly rise to target. I'll work on this more later, it's a combination of getting the idle screw to raise the base RPM as well as idle tuning. Butt dyno says the car has a bit more WOT torque but who knows, that could also be the cooler air letting the car run without retarded timing. Main things I need to work on with the tune now are:
- Better idle, faster to reach idle setpoint and less idle droop.
- Reduce bog on throttle blip.
- Reduce RPM droop on startup, need to open the idle screw a bit more as well as tune idle PID.
- Make alternater voltage a bit smoother (it's pretty good but could be better).
If I can at least get the first three a bit better tuned I'd say that the tune is pretty close to an OEM feel. I'd almost feel confident to take it to a track day as is but cars are too expensive to send it.

Tune and log here: https://rusefi.com/online/view.php?msq=742
OSprey42
contributor
contributor
Posts: 26
Joined: Thu Sep 16, 2021 8:18 am

Re: Hellen NB1 RusEFI

Post by OSprey42 »

Today I refined the idle tuning a bit more and in the process learnt a bit more about how it works.
First I turned off the CL idle and idle timing and dove the car around until it was warm then I set the idle screw until the RPM was slightly below my RPM target (for my 875RPM target I tuned it to 800 RPM ish). I set my open loop base position to 0 otherwise the IAC will be active during normal driving.

Next I turned back on the CL idle settings and noticed that the idle would always drop below the target then come back up. After messing with the PID values for a while to tune this dip out I realised the importance of the CL idle Offset. Basically when the car goes into idle, the PID controller goes to this value. Having it too low will result in the idle droop I was experiencing, having it too high will result in the car not reaching target. You need to tune this to be just right so that when the gar goes into idle mode it is pretty much at the right IAC valve percentage. The best way to get this offset value I found was to just wing it and look at the datalogs of the RPM and IAC position. For my car I found 20 to be a good offset but it will vary for your car because it is dependent on the idle screw position. Screw it in more to close it off and the offset will need to be raised, opposite if you open it. If the idle screw and offset combined are too high, the car may not be able to lower it's RPM down to the target so it's important that the idle screw OL RPM is slightly below target when the car is warm because the screw OL RPM depends on how warm the car is as I found out by chasing it around adjusting it when the car was cold.

Similarly, I also found the AC adder % is also important to get right. Too low and the RPM will bog then rise to the AC target RPM (I set it 200RPM above normal idle), too high and the RPM will overshoot then settle. You've got to get it just right which was 15% ish for my car.

Starup is also impacted by the PID offset. I found the car RPM dips on startup because the IAC goes from the cranking open% to 0 then the PID controller kicks in and the IAC % rises to the PID offset value. It might be good to look into a feature for changing how the cranking to idle IAC% works in RusEFI to taper from the cranking IAC% to the PID offset% not cranking%, 0, offset %. If you look at the RPM and IAC% in the log in this post's tune you'll see what I mean:
Screenshot 2021-11-25 224505.jpg
Screenshot 2021-11-25 224505.jpg (65.4 KiB) Viewed 9072 times
By tuning these parameters and the idle screw well, this minimises the work the PID controller needs to do to get the car to idle correctly which also means that the PID controller doesn't need to be as perfectly tuned to get a decent idle. I think my PID values in this tune are alright but still need refining.

Talking to Matthew Kennedy, you apparently want the OL IAC % to be above 0 as the IAC duty cycle% = OL% + CL%. I will later experiment with swapping the OL% with the offset% so the OL% becomes +ve and the offset becomes 0. This may help to remedy the startup droop.

Today's tune: https://rusefi.com/online/view.php?msq=745
10AE0739
Posts: 15
Joined: Mon Aug 15, 2022 7:02 pm

Re: Hellen NB1 RusEFI

Post by 10AE0739 »

Excellent information. I think this thread will help me with my CL idle tuning. I love that there's at least 3 10AE running the Hellen NB1 RusEFI.
Post Reply