Eliminating need for voltage correction using stabilized 12V

[Dennis_Zx7r]

Just a thought at the moment.
Behaviour of the injectors (deadtime), the ignition system (dwell) and other electrical systems (non-mechanically regulated fuel pumps) is all dependent on the voltage it is supplied with. Thus there are several corrections for voltage implemented in the software, needing additional data / calibration.

This is a typical BattV log on my engine (LiFePo4 battery). As you can see, besides the short after start period, BattV varies between 13.8V and 14.3V, depending on RPM (unfiltered, LF100). I think that's ok, considering the regulator was designed over 20 years ago.


However, this is where I wonder why we use the raw battery voltage to power the MS, and all of the engines systems and then fight the symptoms instead of eliminating the source of the problem.
With the graph above, I don't think it would be too difficult to generate a stable, regulated 12V voltage powering (only) the MS, the fuel pump relay with the ignition system and the injectors. I think voltage dropout would only be a problem during the first minutes in warmup, where some (much smaller) voltage correction could be used. As for the systems I want to power, peak current would be below 15A. Stuff like lights, fan, horn etc. would remain unregulated.

Are there practical limitations why this isn't done or couldn't be done this way? Is it just cheaper to compensate for this via the ECU, or do modern alternator regulators simply provide a steadier voltage?
Efficiency, fuel economy, cost/benefit? Did somebody here already try this? I'm curious.

[whittlebeast]

I use one on these on the race car. I save the weight of the alternator and the power required to spn it's guts up to 35000 rpm.

https://raceenergyperformance.com/colle ... cts/tdc-30

[dontz125]

I think part of it is since that you DO have to compensate for it during cranking and initial starting, the benefit of NOT compensating through simpler coding etc is already lost. Why then add the expense of an additional voltage regulator? Note too that you will be wasting power and running your secondary regulator hotter because it is ALWAYS dropping the voltage below the generated level, instead of using whatever is produced by the alternator.

[jacky4566]

whittlebeast already answered the question.

You can use a voltage regulator but to get 20-30 amps your looking at some fairly beefy circuitry. $$$ As MS is a budget focused system, I am quite happy to use voltage correction. Most aftermarket injectors already give you this information so its just plug and play.

[billr]

Oh my, 35,000 rpm! I gotta ask... what type engine is that?

[whittlebeast]

Honda B18C Type R autocross car. You will commonly dig out of a corner at say 4000 RPM and get real close to the 9000 rev limit. Now look at the size of the crank pulley and the typical alternator pulley and figure out that the alternator is spinning at about 3.5 to 4 times the crank RPM.

Things get crazy fast.

Andy

[WestfieldMX5]

It might be time to upgrade the regulator to something more recent. The voltage on my 2000 miata is rock steady within 0.2V under any condition (except starting of course).

[Dennis_Zx7r]

I use one on these on the race car.

That's a nice way of doing this, I think. With the (non-MS'd) race bikes I'm familiar with, you usually throw out the alternator and... that's it
I'll consider using something like that when running without an alternator.

I think part of it is since that you DO have to compensate for it during cranking and initial starting, the benefit of NOT compensating through simpler coding etc is already lost. Why then add the expense of an additional voltage regulator? Note too that you will be wasting power and running your secondary regulator hotter because it is ALWAYS dropping the voltage below the generated level, instead of using whatever is produced by the alternator.

With the buck converter I originally considered, some rough estimations would suffice, as this period would only be quite short and therefore not that important, and the code for this is already there.
With the booster whittlebeast mentioned, that wouldn't be necessary at all, even during cranking. Battery efficiency isn't much of an issue for me on the racetrack, and I'd estimate the power loss to be something below 20W in my case.

[...]$$$ As MS is a budget focused system, I am quite happy to use voltage correction. Most aftermarket injectors already give you this information so its just plug and play.

The voltage correction in MS unfortunately only seems to work in relatively coarse 0.1V steps, in addition to the measurement with 0.1V accuracy. I don't doubt that this does the job good enough in real life, but I think dealing with the issue itself may have some benefits under certain circumstances. And of course it would simplify things when you have to gather the data yourself.

It might be time to upgrade the regulator to something more recent. The voltage on my 2000 miata is rock steady within 0.2V under any condition (except starting of course).

Hm, unfiltered? Having a look at the regulator is on my ToDo list anyway though, as I highly doubt it's still state-of-the-art, not that it ever was
I also plan on modifying it to limit the voltage to about 14.1V to better suit the LiFePo4, and that in itself will probably narrow the spectrum.

---
Thanks for participating. What I take from this so far is that it would be possible to do something like this, but most likely not worth the effort with a good alternator.

[billr]

Ah, got it. I misread, thought you were talking about engine rpm, not the alt. Yeah, even my BBC alt gets up towards 13-15K rpm, fortunately only for brief periods.

[DaveEFI]

If you regulate at 12v, you'd possibly have to alter quite a few other components like say the pump. As these are all optimised for 13.8v.
Modern cars tend to have much larger alternators than once was common. And this isn't fully explained by a very much higher electrical load, as much of that is intermittent - like electrically operated tailgates, seats, and so on. But even then, the voltage at idle with a full load may well drop slightly.
However, the performance of the engine isn't so important at idle.

[Marek]

I think what is not explained by a simple graph is that the electrical environment is not remotely stable. Every time the injectors open and close, they go from consuming no current to consuming "maximum" current in an instant and whilst the overall graph may look steady, what is in reality happening is current usage is in a never ending stream of large pulses.

Consider the analogy as explained with water:- your graph is a steady and flat canal with, on average, water flowing steadily and slowly downhill. The reality is that all of the electrical components are like power jet-washers being switched on and off rapidly in sequence with water quickly and violently being turned on and off.

A large capacity battery and alternator are effectively trying to smooth out those electrical pulses and provide a steadier environment for the electrical components to operate in. The better the regulation you want, the more storage (and over)capacity the system will need, such that the normal operating environment represents a just smaller ripple in a bigger pond.

For a conventional petrol engine, the cost/benefit clearly isn't there, as the oem manufacturers will have done this years ago if it made sense.

With the advent of hybrid motor design, this is no doubt being considered as part of the overall design as every bit of energy saved/optimised is reflected in lower running costs and/or emissions. It manifests itself by the operation of an electrical motor when that is more efficient than operating a petrol motor, especially when using what would have been wasted energy from the petrol engine.

kind regards
Marek

[motthomas]

I'm not sure what you are trying to achieve by regulating battery voltage down to 12V... That can only be a bad thing as you will end up regulating the voltage twice. Once with the built in alternator regulator and then with whatever regulator you end up adding in. That would make for a pretty inefficient charging system.
If your concern is that you are operating in a region where the injectors and coils have not been characterised properly then you should characterise them at a voltage that is more consistent with what you do actually achieve when running the engine. 14V for your setup looks like a good point to do that. Your system already has a voltage regulator built in to do a specific job so you should use it. If voltage is not stable enough for you then maybe consider upgrading the regulator. Modern units can provide very steady output.

Even the dedicated voltage regulator that whittlebeast showed is designed to boost battery voltage up to a minimum of 13.5V. 12V is only a reference voltage anyway as all components in the vehicle will be designed to run at typically higher voltages, demonstrated by the fact that the vast majority of alternator regulators are designed to regulate to c.14V. If an individual component wants a very specific voltage then it will have its own voltage regulator to do the job.

[billr]

I think the suggestion was to regulate to a suitable voltage, not exactly 12V. There are certainly ways to do it, like a switch-mode PS; but I, too, feel the extra cost and complexity is hardly justified. If even justified at all!

As a side note: I recently had to do some work on my son's Prius. It seems the "12V" system there is powered by the HV traction battery, regulated down to 14.5V The 12V battery is used only for the engine-cranking current and as a back-up for the 12V system?

[whittlebeast]

Just for fun, here is my race car with the voltage regulator running the show. This data was when we were roughing in the VE tables on the dyno. Notice that you can not even see the voltage dip wile cranking at 150 RPM.

http://www.nbs-stl.com/CrxTypeR/EarlyDynoSweeps.msl

Andy