24s or 25s on a Lyen controller?

[Drunkskunk]

I'm designing the frame for my next bike. I've been using 5s packs, and have 5s chargers and bulk chargers that can do 10S. but peak voltage on 25s will be 105 volts.

Since I'll need new batteries anyway this year, would using 6s packs for a max of 24s be a better idea?

I've got a Lyen controller and a direct plug in CA. I've read they can take 24s, but how about 25?

 

[Philistine]

I reckon you should PM Lyen for a definitive answer. I really want to know the answer to this question as well, because I too run 20S, but would like to knock it up to 25S at some stage with my stock Lyen 24 Fet - but when I bought it I remember Lyen asking me if I was ever going to run it at 100V, as if that was relevant to a mod he might make when I ordered (I could be wrong about that, but I am sure he asked me that question). So I say PM him and post back with his answer.

 

[neptronix]

Voltage limit specification for irfb4110 is 100v. Does that answer your question?

 

[MadRhino]

I recall the 4110 mosfets are rated 100v.
I run 24s with them, and feel that I am playing with their limit already.
I tried the 4115, but they have higher resistance and I couldn't feed as much current.

 

[neptronix]

Yup. yir working them at their limit. a few have blown here already due to that.
The trick with 4115's is to run a 18/24/36FET, unfortunately.

 

[MadRhino]

My 18 fet 4115 doesn't like feeding more than 4000W. I believe its only advantage is for feeding high voltage, but my goal is not top speed. I am fast enough in the mountain with 100v, so I'm after torque and the 4110 controllers can feed a lot of Amps.

 

[Philistine]

I recall the 4110 mosfets are rated 100v.
I run 24s with them, and feel that I am playing with their limit already.
I tried the 4115, but they have higher resistance and I couldn't feed as much current.

Ahh that would explain why Lyen was asking me the question about 100v, as to whether he would send me his 4115 "highway" 24 fet or his 4110 "muscular" 24 Fet. The penny drops. I had seen his 4115 24 Fet later and wondered why it wouldn't simply be better (if it could take higher voltage), I hadn't realised about there being a current trade off as well.

 

[zombiess]

My 18 fet 4115 doesn't like feeding more than 4000W. I believe its only advantage is for feeding high voltage, but my goal is not top speed. I am fast enough in the mountain with 100v, so I'm after torque and the 4110 controllers can feed a lot of Amps.

My 18 FET 4115 laughs at 4kW. I've fed it over 12kW on multiple occasions now but now stick to about 10kW to keep the reliability factor higher. The only reason I can get away with such high power is I've done some pretty time intensive mods (I'm working to find a faster method of construction) to my controller. You'll have to search the tech section for threads I started to see them. Deals with changing the thermal insulator for a better one. Loads of room for improvement if you have the time to do the labor.

 

[dogman dan]

There is also the resistor on the 12v part to consider. For above 100v, Lyens put a different resistor on my 12 fet controller.

Need an excuse to treat yourself to a new Lyens controller? You got it.

Older CA limited to 100v too, btw. Nother good excuse to spend.

 

[MadRhino]

My 18 FET 4115 laughs at 4kW. I've fed it over 12kW on multiple occasions...

Yep, I've seen your thread about changing the heat sink and other mods that you've done on this controller. That is a success well deserved, but way beyond my electronics abilities. The controllers that I use are all upgraded to feed more power, yet not to cool better. I don't have the time nor the electronic knowledge to work on controllers, they've been done for me by kind members of ES. Steveo have repaired and upgraded 6 of them (thanks again if you read this), and they sure deliver a lot of power. My only problem is heat now, with controllers and motors as well. Riding the mountain trails with large diameter wheels, I have no top speed problem with 24s and one need serious balls to ride my bikes WOT on a bumpy singletrack with big trees each side, but I have to be very careful with overheating.

 

[itchynackers]

I ran my lyen 4110 12 fet on 25s for a long time (one summer, maybe 800 miles) with no ill effects. I generally only charged to 4.1v/cell, and occasionally to 4.2v/cell. I now run that controller at 24s, and my newer 18fet at 25s. I ran the 18fet at 9000w peak for a bit, but backed it down to only 7000w peak since I didn't feel a real seat-of-the pants difference.

 

[ohzee]

yea fwiw I have a lyen 12fet controller with no mods and I have done about 200 miles at 24s 97v hot off the charger.

I figure if I bog down I will pop something so try to take it easy.. Still I like The dogman's suggestion to order a new controller.

 

[auraslip]

So the 18 fet rated at 150v max can only handle 4kw continuous? That sucks. What if the traces are beefed up? I think I could handle that mod.

I've been putting 10kw peak and 5kw continuous into my 100v 18 fet controller. Unmodded.

I guess the 24 fet controller is where I need to go if I want to upgrade for more voltage AND more power?

 

[MadRhino]

Beefing the traces can make it feed a lot of power, but that doesn't make it handle the heat of it. The 4115 fets produce much more heat than the 4110, because of their higher resistance. The kind of mod that Zombiess did, is what need to be done with the 4115 controllers, to make them shed the heat better. So far, the best way to feed big Amps above 100v, unless you are willing to live with the price and weight of much bigger controllers.

 

[auraslip]

So basically just clamp a bunch of heat sinks to the fets? Sounds EASY. HAHAHAHA. NOT.

 

[neptronix]

From what i read, the issues with the longer controllers is that the current sharing gradually gets worse the, more FETs you add.

Also, the higher voltage the FET, the lower the current it can handle.
4kW continuous might be at a lower voltage figure. 4115's can handle *up to* 150V, so say you want to be conservative and run 135v on 45a constant? that's 6kW and you got some room for big bursts of power above that.

 

[MadRhino]

4kW continuous might be at a lower voltage figure. 4115's can handle *up to* 150V, so say you want to be conservative and run 135v on 45a constant? that's 6kW and you got some room for big bursts of power above that.

I ran an 18 fet 4115 at 100v, and it was really hot when feeding more than 4000w. I didn't try running it at higher voltage, but presumably yes, it would have been able to stand more power. The problem is that I would have to run at low efficiency speed in the mountain with more than 100v, and it is the motor that would overheat.

After frying a few controllers, I started to monitor heat and found how far I can push the various motor-controller combos that I use. Now I can play with the limits to pull the best reliable performance out of each, and to select them according to the terrain and conditions. I always run 100v, but swap the wheel and controller to perform in the conditions of the planned course and weather. The advantage is that I can have each motor wheel ready with appropriate tire for a slow muddy day or rocky trail, fast dry pack or bumpy jumps. 2 motor weight, 2 motor windings, 6 controllers, let me pull the best of 100v in any conditions. I can swap the combos on my bike in 5 minutes, put a fresh battery pack and ride the same trail to compare.

Since the Cro-motor won't fit with this setup that is made for Clytes X5 and H series, I'm building a new bike for this one, and hope that a different winding of it will be available during the season. I don't know yet what I'll do with the two X54 that I ordered, but unlikely to ride this 30 lbs boat anchor in the mountain anyway. I might want to run higher than 100v on the road with those, and the mod to feed big current with the 4115 fets would be very interesting. In any case, each bike has to have at least 2 different motors to swap, and I like that all of them use the same battery packs now.

 

[zombiess]

Reason my controller lives is because I help equalize the feed by using the tabs where possible which also adds heat sinking. Lots of room to improve thermal transfer on a controller if you take a little time. I'm doing another 18 fet in a similar way and will post the results. Switching to a bergquist q pad insulator to test them.
Never run a fet at its Max spec if possible. 25s is too much. Even 24s is pushing it but seems to work well probably due to voltage sag.

 

[cell_man]

You should not run any controller at close to the voltage rating of the fets. It is bad practice and the only reason you might get away with it is because the motor is probably quite an easy load, a different motor or some differences to the supply battery supply cabling is all it could take to fry the controller. Safe levels are a max, ie a hot off the charger voltage of no more than about 90% of the fet ratings and a running voltage of about 75%, maybe a touch higher.

I sell IRFB4110 equipped controllers with 100V caps, they are 72V rated with a max of 90V. You are welcome to put 100V into them, but I will advise against it. There is no professional controller makers that will advise controllers are operated at the voltage rating of the fets. There are spikes and you need to leave some safety margin or the controller is more likely to fail. IMO and the opinion of people who know a whole lot more about controllers than me, 20S lipo is a safe upper voltage level for a IRFB4110 controller, 24S is too high.

 

[zombiess]

And cell_man backs me up. Bad practice to run a controller at it's max voltage rating. We are lucky that International Rectifier specs their FETs a little on the safe side and most of the IRFB4110s are good until at least 5V, but all it takes is a good spike and bye bye FET, hello soldering iron.

Those of us running 24S on IRFB4110 equipped controllers (me included) should be under charging LiPo to 4.15V max. I believe most of us are just lucky and have these survive because we get enough voltage sag when starting off that the surface charge drops off pretty fast and gives us a higher margin of safety. I will never run my IRFB4115 controllers over 132V and for right now have settled on 126V if I charge to 4.20V. I usually stick to 4.16V per cell hot off the charger. It's amazing how robust these simple Xie Change controllers are for being so cheap and the punishment we give them.

 

[MadRhino]

Well, I charge at 4.15v per cell, and haven't used any higher Kv than the X5304. So far, my most fragile controllers are the 18 4115 and one of the 24 4110. Those 2 just can't survive a full week. I've had the 18 4115 repaired again, but left it on the shelf since. All the other controllers that I have, most are 18 fet 4110 and also 2 small 12 fet 4110, have survived a lot of hard mountain riding last summer.

24s may be risky with the 4110 fets, but it gives the best performance that I could pull of my bikes yet. 24s is also making for the faster speed that I can ride in the mountain trails, 30s is too fast for the terrain that I have here, would make me climb at partial throttle too often and it is the motor that wouldn't survive. 20s is not fast enough, I don't want to downgrade. You see, 24s is my ideal setup.

Do you think that cooling an 4115 controller can make it able to survive the same power as the 4110 at 24s?

Are there any type of mosfets that are perfectly happy at 24s ?

 

[cell_man]

I'm not claiming to have all the answers, just stating I think it's advisable to run controllers at a safer voltage level. Wouldn't it be possible to use a slightly faster wind to achieve the same speeds on a lower voltage? There shouldn't be any reason why a slightly faster wound motor being driven at the same power but on a lower voltage would have a significantly different performance.

 

[auraslip]

I thought that partial throttle was hard on the controller and not on the motor. It lowers the voltage the motor sees via PWM and so lowers the speed of it's efficient range. The ebike.ca simulator confirms this, but perhaps I'm missing some condition like hill climbing.

There shouldn't be any reason why a slightly faster wound motor being driven at the same power but on a lower voltage would have a significantly different performance.

I'm guessing more amps is the issue. It seems like we are starting to hit the amperage limit of what 18 fet and even 24 fet controllers can pull. So people, like me, think "oh we can use a 4115 controller with higher voltage to get around the amperage limits." But of course that doesn't work. Another issue is availability of fast winds. I'm not gonna buy an x5 when for a little more I can get some of the new motorcycle class motors. But everyone is waiting for a faster wind version of those motors. Personally, I'd like a faster wind of the magic pie in a smaller wheel size. But even then, we're back to the issue of popping controllers.

Seems like some controller maker needs to copy zombiess cooling design.

 

[MadRhino]

Yep, climbing mountain trails where you can't run the motor in its efficiency range, and the use of large diameter wheels that are needed to ride rough terrain, is making a fast wind motor overheat quickly AND is pulling very hard on controllers.

The 5304 is my fastest winding, and I can't ride it in all trail conditions. I have to use the 5306 when the trails are bad and I have to slow down a lot in some sections. It is not bad, since I have the slower windings laced in different rims and tires, that are more appropriate for wet and slippery trails. I use the H motors when the trails are nice and I want to glide over obstacles, even try some little jumps, because they are lighter and let me tune my suspension much better.

I have found a good way to make my motors and controllers survive: To swap them according to the trail conditions. I'm still playing with the very limits of the 4110 fets and risk controller failure anytime, but that is how I achieve the best performance so far. I would like to use controllers that are more reliable with 24s, but they would have to let me feed the same current, without adding much weight and size.

 

[auraslip]

I think you are the resident DH guy on the forum here madrhino. I'd love to see some of your bikes, and videos of you tearing up the trails. You'd probably laugh at what I consider "off road!"

All I want is a 60mph 'highway' bike. I suppose it could be done on 24s.