ECMlink vs Standalone ECU
This is a topic that comes up a lot. Which is "better", the stock ECU and ECMlink, or a standalone ECU from whatever brand is the flavor of the month. In my opinion, it's a bit of a trick question. What's better for one person is not necessarily better for the next guy. What you run for an ECU is a personal choice, and there are a LOT of factors that can affect that decision. Your average internet discussion about this topic rarely covers the points that actually matter. I'm going to attempt to cover this thoroughly here, perhaps painfully so, in case it helps someone sort through his or her own thoughts on this matter. First, a bit about my own history with ECU technology, so you can attempt to determine if I'm biased, and to which side. :o)
I got into DSMs in 1999, finally buying one, a 1995 GSX (still under warranty!), in January 2000. Right away I started "tuning" using potentiometers on the baro wire, Apexi AFC, MAF hacking, 1G cam angle sensors, etc. At the time a pocket logger, which ran on a Palm Pilot, was high tech for 2Gs. 1Gs did at least have a laptop application available from TMO. Both gens made use of modified ECU chips also from TMO. None of this was ideal, but it worked, and people went fast this way. AEM's V1 standalone ECU came out sometime in the early 2000s, which I mostly ignored at the time. ECMlink however came out in 2002 and that was a very interesting option to me. A nice program that worked within the stock ECU and did 90% of what we thought we needed at the time, still passed state emissions testing every year, and was very affordable and easy to use. I loved it immediately. I tuned my own car and many local cars on it back in those early years.
Skip ahead to 2005. I sold whatever DSMs I had at the time and my tow vehicle and bought a new EVO8 RS in May of 2005. I started using the software for the stock EVO ECU and immediately did not like it. That doesn't mean it sucks, or there's anything wrong with it. But 13 years later you still can't pay me enough to use it, I avoid it at all cost. The next best option at the time was to wire up an adapter harness to run a 2g ECU and my beloved DSMlink program. With some help from local DSMers and friends Matt Burns and Bill Kehn who had already done this on Bill's car just prior (the first to ever do so), I was tuning the car on the 2g ECU by about June. So much for just doing a few simple mods at first! For the most part, it worked great. There were some minor quirks that were largely worked out later on with version 3 of ECMlink, but my EVO was long gone by then. At any rate, the car was running mid 11s either on the stock turbo and a 50 shot or on an FP Green. Either way I was happy with that, I had no intentions of putting a roll cage in the car. I made thousands of drag passes in that thing.
At some point I got a hair up my ass to try a standalone ECU. I bought an AEM V1, still a very popular option at the time (despite a fair number of unhappy customers who already hated that ECU for a long time). Really it was reading the manual for it that got me excited about it, so much new shit to learn. I couldn't wait. I spent a lot of time learning that ECU. After 6 months of daily driving it and constant tinkering, I had it running at least as good as the stock ECU (making power is always the easy part). I was able to learn a lot more about EFI in general (and I already thought I knew a fair bit by that point, lol), and even got into some pretty advanced stuff regarding injector phasing, for example. Things I learned here would serve me well for the next decade plus.
In 2007 I bought my RWD Talon (hard to believe how long I've had this thing). It came setup for another AEM V1 ECU. By now there were other options, but it was there, and I was already familiar with it. It worked great for me and I made good use of many of its features. The car ran into the 8s pretty much immediately, at the DSM Shootout of that first 2008 racing season with the car.
A few years later, in 2012, I got the idea to swap a stock ECU into the car. There was a lot of internet talk at the time about how the EVO ECU was superior to the DSM ECUs because it had a faster processor. I felt that was largely horseshit as far as power is concerned, and thought it would be fun to go just as quick on the stock ECU and ECMlink as I had gone on the AEM ECU, by this time already a relic of course. I like bizarre challenges, which should be apparent looking at the rest of my car, and I knew the stock ECU would not slow the car down. And of course, it didn't. The car continued on into the 7s on that stock 1990 ECU with ECMlink V3. This is the quickest ECMlink has ever gone, and nearly as quick as the DSM ECU has ever gone (7.81 in Buschur's old 2g tube chassis car). I have a few things holding the car back from going quicker than 7.91 at 180, but the ECU is not one of them. This ECU is still in the car, though the setup has grown somewhat. More on that in a bit.
I started tuning cars for other people back in the very early 2000s, even before ECMlink came out, but it was just a side hobby. By the time I opened Six Sigma Tuning in 2015 though, I had already remote tuned hundreds of cars on ECMlink in dozens of countries, and also on AEM, Holley, Big Stuff 3, Fast xfi, and other standalone ECUs. But, it was still just a hobby. In just a few years at the shop however I got my hands on ALL kinds of ECUs, dealt with all kinds of successes and failures, good and bad support from ECU manufacturers, and learned a whole shitload more above and beyond what I already thought I knew. I have a hard time calling this work, I still enjoy it.
In late 2016 I knew I preferred the Holley ECU for various reasons, and I came up with a custom harness and laser cut CAS trigger to me allow me to run one. I got it set up and running sometime that off season just to be sure it worked, then went back to the stock ECU. In 2017 Nick Stack got his truck running on Holley, which I set up for him, and we got some great experience tuning that system on a 4g63 (two seasons worth now). Once I finish a few other things I'll finally switch my car over to that Holley ECU permanently. There are some pretty advanced things I want to experiment with eventually.
I don't really make any money from any ECUs in particular. I install and tune them all, but I rarely sell parts. I have no dog in this fight, since I use all manner of ECUs myself and tune them regularly. But, I hate to see people get pushed into a fancy ECU they don't need, which doesn't fix their problems, probably causes new ones, and then the added cost and complexity kills the whole project. I hate to see it, and it happens all the time. I especially hate to see it when the car in question would have been better off just keeping the stock ECU. It's worth taking a very close, honest look at your needs and situation.
ECMlink vs Standalone ECUs:
There are a number of common arguments comparing the two options. I think looking at those, and later at the pros and cons of each, all in one place, would be helpful to a lot of people.
1) Standalones have safeties, ECMlink does not.
This is of course true. If safeties are really important to you, there is no simple way around this. Some safeties can be built into the setup however. You can still cut ignition with various pressure switches, switch boost levels in a variety of ways, make use of external safeties like AFR safety enabled WB kits, switch maps in ECMlink a number of ways, setup solid safeties with good meth injection kits, and of course, just build a robust setup that doesn't suffer from unusual issues every other pass. More often than not I see safeties shutting down the engine for people on perfectly good passes, due to sensor glitches and other issues. Personally, I don't want to lose a round for a false alarm. I rarely have failures that would trigger or be prevented by any ECU safeties. In the past I had a lot of bizarre issues, and I've learned to build the setup to avoid those situations in the first place. But, this is a personal choice. Go with whatever you are comfortable with here.
2) ECMlink has limited inputs.
Also true. The 1G ECU has one spare input and the 2G ECU has 2 spare inputs. If you run the WBO2 to the front o2 input (there are pros and cons to this) you can use those extra inputs for things like fuel pressure, oil pressure, etc. I do also log my transbrake button and brake pedal switches on the clutch switch input (can also trigger 2 step and antilag) and on the IdleSWpin input (this an also switch maps), which can be used to some advantage on auto trans cars. These also work great as start of run markers in logs. I find it most useful to log Fuel Pressure on my one spare analog input, since ECMlink will automatically do the math with MAP pressure to display base fuel pressure in the logs, which is incredibly useful in diagnosing a vast array of possible fuel system problems at a glance. Keep in mind the usual engine parameters like ECT, IAT, MAP, etc are all logged here already for ECU operation.
Personally, I have no problem running an additional data logger. An AEM AQ1 in my case. I get 8 more analog channels, dozens of things logged over AEMnet as well, 3 switched inputs, in addition to the built in sensors (3 axis accelerometer, battery voltage). Also logging brake pedal and transbrake on the switched inputs, in addition to RPM, I can very easily match up the two logs. This logger is pretty affordable at ~350 bucks, and between this and my (or others') spare input on the ECU I've logged the following things:
- Fuel pressure
- Comp outlet pressure
- IC pressure drop
- Comp outlet temp
- Pressure and Temp between compound turbos
- Shaft speed on both compound turbos
- Ice water temp
- Crankcase pressure
- Oil pressure
- 4 channel per-cylinder Wideband O2
- Various EGT probes
- Turbine drive pressure for both compound turbos
- Exhaust system pressure
- Meth injection pump pressures and solenoid duty cycles
- Meth injection safety status
- Meth injection flow sensors
- Air filter pressure drop
- Ambient or compressor inlet temps/pressures
- Turbo oil feed pressure
- Oil temperatures
- Trans fluid temp
- Wheel speed
- Driveshaft RPM, HP, and TQ (still working on getting this logging correctly)
- Transmission cooler loop pressure
- The list goes on.
So, you'll have to forgive me if I don't feel like I'm missing any data.
3) Ok, so you can log more inputs but you can't "do" anything with it.
This is true. I can't trigger outputs etc with data logged in a standalone logger, while with a standalone ECU, I could.
4) ECMlink (or the stock ECU) only has two outputs and they're only for nitrous.
Yes there are only two outputs. No, they are not only for nitrous. I've used these two outputs to control the following things:
- Dry and Wet nitrous
- Single fogger and direct port nitrous
- Additional upstream injectors
- Staged injectors per cylinder
- Pre-compressor methanol nozzles
- Meth injection kits
- Boost controllers
- Shift lights
- Intercooler pumps
- Staged fuel pumps
- Additional pressure to top of wastegates
- Fast idle air bypass solenoids (this is why I asked for ECT to be added to triggers)
- Radiator fans (same as above)
- Preventing WBO2 from heating until exhaust is warm
- Switchover to secondary fuel system (i.e pump gas to straight methanol)
You can get pretty creative with these two outputs. I pick the two things I really want the ECU to have control over, and often additional outputs can be replaced with simple sensor/switch/relay systems.
5) ECMlink can't control additional or staged injectors!
Completely false. I've been doing this forever, and I was far from the first. I use a simple injector driver controlled by the ECU. It's still sequential injection since the secondary injectors fire with the primaries. The nitrous outputs handle the triggering very well, the Secondary Global fuel value seamlessly adjusts fueling. If I didn't know I had 8 injectors with the second set staged to come on under boost I would never know it was switching over. There's no reason you couldn't run HUGE secondaries, or even 12 injectors.
6) But certainly there's no way ECMlink can control a DUAL FUEL system!
No, that's not true either. I've done it. One system with small injectors, an electric pump, regulator, and pump gas. The other system with large injectors, mechanical cam driven fuel pump, regulator, and M1 methanol. It would switch over to the methanol fuel system under boost. In this case it never shut off the pump gas system, but through clever injector sizing, fuel pressures, etc, it ran around M80, or 80% methanol 20% 93 octane. This was a non-intercooled car with lots up upstream injection controlled by ECMlink as well. This is an interesting option in places where E85 is just not available, but Methanol is (common in other countries). The benefits of methanol at WOT, but it starts up, idles, cruises, and shuts down on gas. No additional maintenance. This does require some math skills, since there's is one single global fuel multiplier for the switchover. But, in 30 seconds with a calculator, I can account for different fuel stoichs, injector sizes, and fuel pressures in that single value. It works quite well.
7) But I want to run Methanol, or flex fuel, or whatever...
ECMlink can do that too. I've run gas, ethanol, methanol (intercooled and non-intercooled), E85, dual fuel, and flex fuel, and it all works great. ECMlink's flex fuel setup is particularly good, though one limitation that bothers many people is that it can't switch boost. Personally, I tune a lot of cars to 35 psi on pump gas anyway. Just switching to E85 there can be worth 100+ HP. Totally worth it. Boost can always be adjusted manually by a number of different methods.
8) ECMlink is great for 10 second street cars, but not for race cars.
My car has run 7.91 at 180 on ECMlink and a 1g ECU. I like to think it's a race car. Various fuels, air to air intercooled, air to water intercooled, non intercooled, single turbo, compound turbo, nitrous, the setup doesn't matter. I've run them all. And I've remote tuned a number of street and street/strip cars into the 8 second ET range.
9) Ok, but you would make MOAR power if you switched to X brand ECU!
I really don't think I would, but I haven't tested it yet. When I switched from AEM V1 to the stock ECU in 2012 I was running a 16g turbo at the time, in the 10.30s ET range. There was no discernable change in ET there. At up to 1000 hp I make 10 DJ HP per lb/min airflow, which is usually the best I achieve on standalone ECUs as well. It should be possible to achieve higher BSACs than this on methanol, but I've yet to push it past a BSAC of 10. This is through an auto trans, for what it's worth. When I switch to the Holley ECU eventually, I'll make a note of any power change once it's tuned. But the power level is determined much more by the setup than the ECU. I have far far bigger limiters right now on my ET than the ECU, or even my power level for that matter.
10) Ok, you can do all those things with the stock ECU, but you have to add a datalogger, an injector driver box, etc.
This is true, it requires a few devices to make this possible. But most people that are so quick to point this out also end up running multiple boxes with their standalone ECUs. CAN WB modules, CAN EGT modules, external traction controls, CAN IO expansion boxes, etc. Some ECUs however (like the Holley Dominator) do it all in one box, and I do see the appeal in that.
11) But OEM 30 year old hardware is junk, newer ECUs are "better"
I think the fact that these ECUs are still running so well 30 years on is a testament to how well they were engineering by the OEM. Personally, I tend to trust OEM engineering over most aftermarket stuff. I've seen a high rate of premature failure in aftermarket ECUs. Most failures of the OEM ECU are simply old leaking capacitors (easy fix, most were probably done when they were socketed) and damage that results from shorted ISCs and other wiring issues. The old ECUs are certainly less powerful than newer processors, but does it really add up to more power? Not from what I've seen. But, it can't be a bad thing. The extra processing power makes all those new features possible.
I personally feel that a vast majority of issues people perceive as being the ECU's fault are actually the fault of the corresponding 30 year old wire harnesses. Those have certainly taken a beating over time, and from users hacking them up for various reasons. Good wiring is critical to the success of ANY ECU. So many people plug their new aftermarket plug and play ECUs into these aging engine harnesses and continue to have nagging issues that are very hard to track down. I think most people's money would be better spent on a new harness before a new ECU. I would do both when upgrading to a new ECU.
12) ECMlink is ~750 bucks out the door, and X brand plug and play ECU is only a couple hundred more...
This is sometimes the case. But again, I encourage people to consider the real costs. A new ECU is not going to fix your gobshyte wiring, but now problems might be harder to troubleshoot, unless you have really good support for that ECU in your specific application. Do you have a local tuner or shop to help with this, if needed? Something to think about. ECMlink might be slightly cheaper, do 95% of what you need, and has a huge support group of users and continued updates from ECMlink after 15+ years. I won't say this is an easy decision, you really have to look objectively at your own particular situation.
13) But this new aftermarket ECU uses "VE based tuning!"
That's a good thing. But, ECMlink has been using VE based tuning since the speed density option came out with V3 in 2008. And it's VE tuning done correctly, in my opinion, with mass airflow provided (can be very useful).
14) With this standalone ECU I can run a number of cam and crank triggers...
Oh man. If I had the space to list every issue I've seen with cam/crank triggers on aftermarket ECUs... I think most ECU makers try to cover every possible cam/crank trigger out there, and sadly end up doing a poor job with the majority of them. I rather like that Holley only supports a handful of triggers and you have to pick one off their list, but they work perfectly, so I don't mind. To take that a step further, the OEM ECU was designed to work specifically with the stock cam/crank sensors and the wiring in between. Noise rejection is generally superb, and issues are exceedingly rare. Once you experience the frustration of mystery crank/cam trigger errors and the long list of damaged engines used up in troubleshooting, and the string of false fixes that seemed to work at first but didn't ultimately pan out, just to find out it was a firmware error from the manufacturer in the end, you'll really appreciate the millions of engineering dollars the OEM put into this. This can really put a major damper in a lot of people's projects, stalling progress and killing budgets.
I even question the need to run more than the stock 2 tooth crank trigger. Higher resolution can't hurt here, but what are you really gaining from it? Slightly tighter spark control perhaps. But I've yet to see an issue using the Kiggly 2 tooth crank sensor (this is a worthwhile upgrade) at 10,000 rpm. I think many people see the OEM trending toward more and more crank teeth, and incorrectly associate that with some kind of a power advantage. It's really done for the sake of emissions (misfire detection), component reduction (firing stroke can be found without a cam sensor), etc. I'm not aware of any standalone ECU using its fancy ~60 tooth wheel for these purposes. I don't mind switching to a 12 tooth wheel when a good opportunity arises, but I'm not expecting the crazy improvements the internet likes to talk about.
15) No internal logging, you need to have the laptop in the car!
True. This is a definite. But, I've found that I don't mind having the laptop in the car. More than once before a round I remembered some change I meant to make to some setting but forgot, and it was easy to do in the lanes with the laptop handy. I do also like internal logging. I really can go either way on this topic.
16) But I really just want to dive into EFI tuning and a standalone ECU is sooo tempting.
I won't argue there, that's how I got into my first standalone. :o) But be sure this is what you want. I see so many projects stall just because the owner had to have the latest and greatest ECU, just to find out that it didn't fix any of his old problems, caused a few new quirks that needed sorting out, had no good local tuner support, and gave up on the whole car. I hate to see it, but it happens all the time. I think this was a significant factor in how I ended up with my own race car!
Pros and Cons...
So, all that said, what happens if we try to boil this down to a simple pros and cons list. Here is why my list would look like. There is of course some overlap between for example ECMlink cons and Standalone pros, but it can still be helpful to deal with the redundancy in order to look at it from both perspectives.
- OEM hardware, which after replacing caps really is quite robust and works brilliantly with the OEM sensors (particularly cam and crank).
- Simple program that anyone can learn to use
- The ability to tune your own car with minimal generic EFI knowledge
- Huge support group for your particular car with ~16 years of history on known fixes, troubleshooting procedures, etc
- Certainly good enough for high 7s in the quarter and well over 1000 whp, with no known limit yet
- Great list of features including VE based speed density tuning, flex fuel, configurable outputs for nitrous/etc
- TONS of self tuners out there willing to share methods, guides, videos, etc. And some good remote tuners as well.
- It's a stock ECU, so it generally will run the engine. You don't have to put a ton of work into a base map to tell the ECU how many cylinders the engine has, what kind of cam and crank sensors it uses, etc. The important, basic stuff, all just works. Likewise, it's hard to get it so out of shape it won't run.
- Brilliantly, every ECMlink log includes all settings files with it. This is great for checking settings from old logs, no hassles trying to match the log up with whatever settings file it might have been using. It also makes remote tuning much, much easier.
- Still needs attention given to old wiring harnesses
- Only 2 outputs
- Only 1 or 2 "extra" inputs
- Old school hardware is beginning to limit the addtion of new features
- Additional logging, traction control, etc requires external boxes
- No "internal" logging.
- No "advanced" features, like custom tables
- Most generic tuning shops have never seen it before and may not want to mess with it
- Latest and greatest features
- Possibility for more functionality included in a single box
- Often come with or require a new wire harness, which solves many issues on its own
- Internal logging
- Additional inputs and outputs ranging from just a few to a shitload
- Additional features that can make use of some of those IO
- Possible custom tables for advanced users
- Faster processing, which may or not be a benefit for you
- Access to all kinds of EFI shit
- Swapping to other parts, like DBW throttle bodies, becomes possible
- More options and configurability for staged injection
- Cost can range from ~1000 to cubic dollars
- Wire harness or adapter harness often required
- Software can be complicated, or intimidating, for many users
- Access to so many parts of the ECU can be overwhelming, and issues with idle and driveability are common
- Generally narrower user knowledge base for self tuners, and many manufacturers will only support dealers, if there's any good support at all
- For Hardware, you're at the mercy of the manufacturer. Some ECUs are robust, some are shit.
- Some software is shit too. Hard to navigate, buggy, etc.
- Many ECUs still require a stack of external boxes to do what you want, adding cost and complexity.
- For many people, getting the ECU tuned professionally is the only option. With frequent setup changes this can get expensive. Many tunes done on the dyno simply don't work out on the street/track, which leads to a lot of back and forth, and frustration.
- I've seen the cost and complexity and seemingly infinite number of possible issues derail or completely kill a lot of cool projects.
You'll have to weigh all these, adding your own as necessary, against your current situation and see what makes the most sense in your case. My typical advice is to first decide if you can tune either ECU on your own, or if you need to have to professionally tuned. If you're going to have it tuned, who is going to do that? How reputable are the local tuners you'll be relying on? What systems do they specialize in?
When having someone else tune your car, you're at the mercy of that tuner being able to get you the most from your chosen system. Picking an ECU with a million advanced features won't help much if your tuner doesn't make effective use of them. While most good tuners can tune any ECU, it's hard for every tuner to be 100% up to speed on 100% of the ECUs out there. I know I'm not. Often a good tuner that's very well versed in a simpler system can get you more bang for your buck than a generic tuner that does a mediocre job on a superior ECU. Time slips don't care about name brands of course.
When tuning the car yourself, you're mainly concerned with end user support from the manufacturer. Will they support self tuners, dealers, or no one at all. There is a huge range here, mostly between piss poor and "ok." If you're the hacker/tweaker type and can generally figure things out on your own through dilligence and clever methods, support becomes less important. In using one ECU for a number of years you have the possibility of becoming quite good at that one system. It may be worth looking into online or in person tuning courses for an ECU you're interested in. Some have all kinds of options, some have none. Many ECU specific courses can be taken for less than the cost of one professional dyno tune. But, this isn't going to work well for everone of course.
Most internet discussion focuses too much on specific features. Endless debates about this ECU's traction control vs that ECU's traction control, when in reality, you may never even make use of traction control. It's easy to get lost in the bull shit. But, hopefully this look at the subject from a different angle is helpful to some.