Quantify power savings of 3077 vs 4110

[electr0n]

Since I'm shopping around for another controller I'm just wondering how much power savings would I get if I went with a 3077 over a 4110 setup. I know it's probably hard to say but would 10% to 20% real world savings be realistic or is it likely to be much less? Assuming I'd be riding the same with either controller, stop and go urban and generally under 40 kph. 44.4v 40amps, direct drive, either Crystalyte 407 or HS3540.

I do like the option of running 66.6v and 88.8v which the 4110 gives me, but most of the time that's more speed than I need and I realize this is not an efficient way to ride. If the real world energy savings could be in the range of 10 to 20% I think it would be worth giving up the higher voltage capability for the increase in efficiency but if it's going to be more like 5% or less I think I'd rather stick with the 4110.

Anyone have practical experience here, that's worth more than theory but all opinions welcome.

 

[Alan B]

What you are asking is actually quite difficult to do with any precision, and it is unlikely that the difference will be anything like 10-20%.

assuming other factors are equal (which is not likely, so this is flawed from the start)

typical forward DC loss
2.7 vs 3.8 milliohms per FET
assume 2 FETs in parallel for 40 amps (12 FET controller)
each FET carries 20 amps
power loss is I squared times R
1.1W vs 1.5W per FET DC loss at max current
total DC loss is four times this (two pairs of FETs conducting)
so 4.4W vs 6W

There are many other factors that will overshadow this tiny difference, so this comparison is not all that useful.

 

[Drunkskunk]

Mostly you're looking at a 0.1% to 0.2% overall efficiency improvement, if you're riding hard against the bleeding edge of performance. you're looking at a 0.001ohm difference between the two. (0.0028 VS 0.0037 Rds)

Mostly the difference shows up in power available at the edge of performance. if you're trying to pull out another few feet per minute in speed, or get your power use down by another watt, or maybe run just that tiny bit more power through the controller before frying the FETs, then it's an obvious choice. But if you're not going for high performance, it really makes not one bit of real world difference.

 

[electr0n]

Appreciate the insights. I figured the savings would be negligible but just had to ask before making a decision. I'd much rather retain a wider voltage range than save a couple watts. I suppose I'd be better off focusing my attention on the bike itself to make any real efficiency gains. Things like higher pressure / thinner tires, better bearings and most of all rider position and clothing. Although I will never wear spandex

 

[Drunkskunk]

Yeah, there is a lot more to be gained from changing how your bike cuts through the wind than from the controller.

As for the thinner tire idea, well... http://www.schwalbetires.com/tech_info/rolling_resistance
Turns out, the opposite is true. Wider tires are actually better. Also, a thin tire isn't going to be much advantage aerodynamically on a wider Ebike, with all it's batteries and components. Add in the larger contact patch of a wider tire for better traction and braking, and thin tires don't make much sense.

 

[veloman]

I run 77v on my HT and it is just as efficient as 48v, assuming you ride exactly the same speed. To answer your question, the difference is probably more like 1-2%. Afterall, if you lost 10% efficiency just through the controller, it would be so hot that it would fry very quickly.

Higher voltage at the same speed is just not going to lose you any noticeable efficiency, if any. I never ride WOT, but for those that do, then of course it's a totally different story since speed would change a lot.

 

[Sunder]

As for the thinner tire idea, well... http://www.schwalbetires.com/tech_info/rolling_resistance
Turns out, the opposite is true. Wider tires are actually better. Also, a thin tire isn't going to be much advantage aerodynamically on a wider Ebike, with all it's batteries and components. Add in the larger contact patch of a wider tire for better traction and braking, and thin tires don't make much sense.

I'm not sold on that idea, for several reasons:

1) Pretty much every article, including the one you quoted, say "at the same pressure". But if you're running your 23s at the same pressure you're running your 37s, you're doing something wrong somewhere.

2) Anything up to about 25mm tends to be racing tyres, or tyres optimised for lowest rolling resistance. Anything from 28 through to about 37 tend to be optimised for puncture resistance as a commuting tyre. This lowers the rolling resistance as the tyre is less compliant. Anything 37 upwards tend to be optimised for grip, which means even higher rolling resistance.

So saying "all else equal" in a lab is one thing. In real life it's another. The other thing to keep in mind, is that at 40km/h, on a flat, rolling resistance is less than 20% of the resistive force on your bike, Is it all that important?

If it is all that important to you, from the research I've done, there is only minimal difference in rolling resistance and aerodynamic performance between 23mm and 28mm. Some manufacturers still make a racing tyre (e.g Continental GP4000S) in 28mm, and from reviews, 110psi is still safe for that width, not much lower than the 120psi you might run on a 23mm.

Me personally? I've gone 37mm commute tyres at 55psi for my cracked and potholed ride to work. I rarely make it to top speed except on very short stretches. My "backup" wheelset, which I am planning to use on a very good quality road to my friend's house (motorway shoulder) is 28mm Continental Four Seasons (commute tyre) at 95psi front, 100psi rear. If I can notice a substantial difference between top speed, pickup, handling, etc, I'll repost here. But it could be a couple weeks before I do that.

 

[spinningmagnets]

I wouldn't have the first idea about how to start to calculate the efficiency difference, but...I've read that if your batteries nominal voltage is 65V or lower, the 3077's will run cooler, and heat is the enemy. If your system is running barely warm using the 4110's at 100V, then at the very least...the controller is in no danger.

 

[Ykick]

I wouldn't have the first idea about how to start to calculate the efficiency difference, but...I've read that if your batteries nominal voltage is 65V or lower, the 3077's will run cooler, and heat is the enemy. If your system is running barely warm using the 4110's at 100V, then at the very least...the controller is in no danger.

I now have a seat of the pants understanding of the difference.

I put together a small 48V front gear motor bike and was running it on a 9 FET with 4110's and programmed for about 15A. Consistently required about 0.5Ah to ride 1 mi or 24Wh/mi. For giggles, I wired up 6 FET with 3077's same 15A and consumption improved to about 0.3Ah per mi or 15Wh/mi.

The end result is like having an extra 1/3 battery pack.

 

[fellow]

Why use 3077 when there is a better TI CSD19506KCS 80V version with Rds_on=2mOhm? My expirience with 4110 is that it works reliably but gets too hot for my taste at 55V (48V nominal), 14A, 6fet configuration and climbing.