How do you predict what voltage sag will be with nanotechs?

John in CR

100 TW
Joined
May 19, 2008
Messages
14,954
Location
Paradise
I have a nice stash of 65/130C Turnigy nanotechs, and I'm deciding what voltage and ah to run. People have said they sag very little, but I want to reasonably estimate sag. eg What's a good guess what the sag would be near the top of charge with 10ah at 30s pushing 250A? Am I correct in thinking the sag would be cut in half with 20ah instead?

I have my insufficient brake issue worked out, and I'm finally setting up for a well documented top speed run with HubMonster. The last time I held the throttle at WOT from a stop for more than 8-10 seconds I hit 107mph, though possibly wind aided. The CA reported a 103V minimum voltage with a pack voltage at the end of the run of 119V. :shock: I'm expecting a whole lot less sag with the nanotechs. 8)

John
 
Have you measured the IR of the pack? I'm pretty sure that's how you would determine sag using V = R * I. Or am I underestimating the problem?

Edited for proper formula
 
cal, you're correct. that's exactly how you would calculate it. and john: yes. double the Ah and your sag will be halved.
unfortunately i don't know if you can estimate the voltage sag just from knowing the battery make. but if you measure the internal resistance of the pack you can calculate the sag for a given load.
 
Hopefully someone will know a "typical" value for a Nanotech cell internal resistance.

Then Vsag = IR

Where I is your peak battery current and R is total pack internal resistance (series resistances just multiply, parallel ones add reciprocals [see Google])
 
Punx0r said:
Hopefully someone will know a "typical" value for a Nanotech cell internal resistance.
)
Maybe, but be careful there are a lot of confusing figures floating about.
John needs a IR reading from a high current DC measurement.
Many of the posted figures are from auto measurements by chargers or even IR meters that use an AC measurement protocol.
This may be helpful...http://endless-sphere.com/forums/viewtopic.php?f=14&t=48739
 
I would suggest testing one pack using the IR measurement method detailed in that link and then you can extrapolate your results to the larger pack. Tricky thing will be getting a load big enough, you're probably going to want something a bit more serious than a few lightbulbs to replicate the loads you're likely to see during your speed run.

IR measurement can be a little tricky because there are a number of factors that can affect it. Weirdly more heat can reduce IR - you can see this in the form of a 'bounce back' in the graph on a single cell high current discharge, as the cell heats up it actually increases its power delivery briefly before falling off.

You're going to want as much battery as you can carry (within the realms of sanity) - for the purposes of top speed your system weight will have relatively little effect, almost entirely down to aerodynamics. The only case where less weight might make a difference in your top speed is if you have a short distance in which to accelerate.

Perhaps you can make a fairing of sorts from batteries? :D
 
Need to find the videos but, I remember 10Ah7s nanos went down from 4.1 to 3.7 for few seconds at MI-2 (2 turbine helicopter) start. By papers, start current is 1200A at static rotor 20C ambient down to 600A at ignition point. Is it 0.0003 Ohms per 10Ah cell :D?
 
parabellum said:
Need to find the videos but, I remember 10Ah7s nanos went down from 4.1 to 3.7 for few seconds at MI-2 (2 turbine helicopter) start. By papers, start current is 1200A at static rotor 20C ambient down to 600A at ignition point. Is it 0.0003 Ohms per 10Ah cell :D?

That seems low to me by 1 decimal point, since I've seen A123 AMP20 cells at around .003 ohm. Could the IR of 65/130C nanotechs really be 1/20th for the same ah?

<edit> That .003 ohm for an AMP20 seems high now that I think about it. That would mean at 200A a 24s20ah pack of them would create almost 3kw of heat...that can't be right, can it?

<edit 2>I never really paid attention to it, but I vaguely remember my CA telling me that the 20ah 111V nominal pack of lower c-rate lipo that I was running before had an IR of 19mOhm.
 
Ah, I didn't realise the CA could measure pack IR. Are the measurements accurate, or just provide a relative measurement so a user can see how their pack ages?

IIRC a "good" standard lipo cell has been stated in this forum to be about 5mOhms (2.5mOhms for two identical cells in parallel). I'd expect the Nanotechs to be better.

Theoretically we should be able to calculate it from the cell spec :D 65C discharge should correspond to nominal cell voltage (3.7V). So for a 5Ah cell we have 4.2 - 3.7V = 0.5V sag at 65 x 5A = 325A.

0.5/325 = 0.0015ohms (1.5 mOhms)

Assuming we believe the manufacturer's spec ;)
 
My power system is apart right now, but I'm sure I've seen pack IR when flipping through the CA3 menu. It has current and voltage data, so as long as it's calibrated and does some kind of averaging it would be the best source I can think of for pack IR. Once I get my bike back together I'll share the data. Then when I do the have my never used Analogger w/GPS capturing more data than I'll know what to do with. I'll look into the IR info on the CA, which may recalc after each reset, and if so I'll be sure to check IR with each cycle to confirm or disprove that IR changes and improves over the first handful of cycles.

wb9k was kind enough to confirm A123's AMP20 cells have a typical IR of around 0.3mOhm, so your previous number seems pretty good. This will be the first time I use a homogenous pack, all 65-130C nano's, so I've got my fingers crossed for 0.0003 ohm or better for each 10ah of 1s for quite low sag. 10s of my old pack was some well used 20C peak rated 18650's, which I'm sure sagged significantly, though capacity and perfect balance have held up well.

I never realized that voltage should sag to nominal voltage at rate C-rate current. I understood nominal voltage to be the halfway point of capacity at very low discharge rate.

Good learning days. Thanks guys.

John
 
John in CR said:
I never realized that voltage should sag to nominal voltage at rate C-rate current. I understood nominal voltage to be the halfway point of capacity at very low discharge rate.

Good learning days. Thanks guys.

John

Yeah that is correct.

According to this:

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=19153

One 5 ah cell is 1.2 mohm, so 10 ah cell is 0.6 mohm, or 0.3 mohm for 20 Ah cell? Same sa a123? hm...
 
Yeah, you'd think to be 65/130C that they'd need to have a lower IR than an A123 amp20 for the same capacity. If it wasn't for the surface charge pushing voltage so high, I'd be tempted to reconfigure my swingarm to fit a nice stack of 37 or so AMP20's since I have enough. :twisted: Less than 3V of sag at 250A sounds good to me with 20ah either route, and about 5V of sag with 10ah is a big improvement over the 16V I've been seeing. :mrgreen:
 
John in CR said:
Yeah, you'd think to be 65/130C that they'd need to be lower than an A123 amp20 for the same capacity. If it wasn't for the surface charge pushing voltage so high, I'd be tempted to reconfigure by swingarm to fit a nice stack of 37 or so AMP20's since I have enough. :twisted: Less than 3V of sag at 250A sounds good to me with 20ah either route, and about 5V of sag with 10ah is a big improvement over the 16V I've been seeing. :mrgreen:
John, remember, I did not measured current, 1200A is from specs papers of the heli and I used original 2x8AWG (cumming on those 2x 5Ah batteries, paralleled) wires soldered to a monster starting cables ~2m with a big (600A continuous) plug on the end. All this may have increased conductor resistance and lowered current flow, co IR of the cell should be higher then calculated.
But in any way, those batteries are awesome. :D We could do about 3 starts per charge, about 10sec per start (all under 20 sec by reglaments), then the cells where under 3.7V resting and where charged again. 30-45sec duty cycle is impressive stuff, all in 4 kg package in comparison to ~400kg airfield starting stations.
 
I decided to work on the swingarm to better fit batteries and move them closer to the pivot. That combined with a bit of weight trimming on HubMonster will let me get significantly more battery in my swingarm battery bay with the same suspension performance. My compression and rigid mounting method for the pack has proved to be safer for lipo than the many packs plus foam padding in a sprung battery bay that so many guys use, so I'm sticking with what started as just an experiment.

Since I get so complacent once I have a bike running, I'll just go all out with changes that improve everything. On the 107mph run I saw something on the road ahead and let off the gas before reaching a stabilized speed after only going about 3/8 of a mile. Last time I road the bike the 2.5 miles to the highway, so it wasn't a full charge and only to 4.15V anyway and a warm motor. I'll go from 30s of 21ah of a hodgpodge mix to 31s of 65-130 nanotech. I'll fully charge the pack to 4.2V/cell, and start with a cold motor. I'm even installing the 20.5" OD tire instead of the old 19.25" tire I've been using. Aero is getting addressed a bit with little change in the bike's look, including more aero clothing and a slightly lower saddle. Extra air will go in the tires, and bump the current limits up about 10%, but go to WOT more slowly to avoid heating the windings unnecessarily. I'll also scout the run right before to avoid surprises that could cause me to let off the throttle prematurely after only 600-700m. I want to get at least 1km of WOT. Significantly narrower bars will be part of the aero treatment, as will practicing an optimum tuck using a 70% speed limit setting into a good headwind. All that is sure to offset any beneficial wind I had last time, and hopefully improve on my 107mph with good video and documentation of the run.

Then I'll be done with the top speed thing and go down to 22s. The high voltage is too inefficient, because I spend 90% of my riding at less than half speed. 22s will get me over 70mph which is more than fast enough for me even on the highway, though I won't get to embarrass sportscars anymore.

Thanks for everyone's input on sag, and low sag should have a nice effect. Sorry about all the off topic info, but if I don't put it public view, then my follow-through is less likely. This way is more to challenge myself than anything.

John
 
John in CR said:
I never realized that voltage should sag to nominal voltage at rate C-rate current. I understood nominal voltage to be the halfway point of capacity at very low discharge rate.

I stand corrected! It must have been coincidence that the IR worked out close to the manufacturer's spec...
 
littleskull99 said:
John, do you have a build thread going for your bike? Thanks. Matt.
+

I appreciate the interest. I've been using this bike for over nearly 4 years. It's been primarily a test platform thru a number of performance upgrades and changes. I've shared some pics along the way, but it ain't pretty though what it does is beautiful. Below is close to how it's been for almost a year, but the pic is missing the front moto wheel I run now. The hodge-podge of batteries (over 20ah of 30s) are in the swingarm and bolted to the frame with a healthy dose of duct tape. It may not look it but the rig is fairly water resistant, and all batteries are rigidly secured.
SuperV sml.JPG

Now that I'm not just winging it and have a better picture in my head of a slick end result, only an easy swingarm mod puts 10ah of 30s or 31s in the swingarm, I'll try to get more pics of the process. I'm still flip-flopping on the quantity and placement of more batts, and whether to do something with the controllers. Those will all dictate wiring cleanup. Inside the main tube is out, because it has a solid plug of fiberglass cloth with epoxy vacuum infused to prevent a failure at the shock mount already cracked when I bought the frame.

Very soon my HubMonster driven SuperV will be proven the fastest and quickest hubmotored bicycle frame based ebike in the world, so it needs to look the part. That will be with 270lb me aboard, so imagine performance with a normal weight person.

IR is super important for batteries, the lower the better. The same is even more true for motors. Low phase-to-phase resistance is the route to high efficiency and power, with our BLDC motors. Copper losses = current squared X phase-to-phase resistance is a sobering fact, especially when you consider hubmotors are limited only by heat. Don't forget resistance in copper goes up about 0.4%/°C at our operating temperatures, so guys running their motors at the high end of hubmotor thermal limits are creating up as much as 50% extra copper losses compared to the motor at ambient. The cooler the better. Ignore heat at your peril, because that simply forces you to buy and carry extra batteries to make heat with. High current is meaningless without low resistance. At 0.016 ohm phase-to-phase resistance, supposedly high performance hubbies in HubMonster's weight range have 5-7 times higher resistance.

Combine high quality laminating steel with a low slot and pole count, and you keep iron losses low as proven by a no-load current of just 2.7A (186W at 69v) at 1260rpm, and 3.4A (365W at 107v) at 1965rpm. Imagine an extremely well built RC outrunner with a Kv of 18.3rpm/v that weighs 15.5kg, but it's sealed in a steel shell with a thick aluminum cover on just the wire side...heavy but the big majority of the weight is the stator. There's no empty space inside unlike the caverns in other hubbies (air is an insulator not good in our heat limited motors), HubMonster is all motor baby, and that lack of air volume means better heat transfer both sealed and ventilated. When ventilated, the lack of open wasted space forces air sucked thru HubMonster to interact more directly with the exposed parts, for better heat transfer directly to the outside world.

John
 
Back
Top