The E-cumbent - A project by Matt Shumaker

[recumpence]

Finally, PICS!

I received my E-logger, Power Jazz, programming card for the Power Jazz, and I had an Eagle Tree GPS I never used from another project.

I put the Power Jazz next to my HV110 for size comparison.

The E-loggter will record volts, amps, watts, 3 separate temperatures, all GPS information (3 dimentional), RPM, and a host of other things if I really wanted to get crazy. :wink: Man, the E-logger has an 11 page manual to read. Plus I have alot of other info to familiarize myself with for these new components.

I will be installing these tomorrow afternoon.

Matt

 

[Miles]

8) Looking forward to this....

 

[jnuts]

Awesome work!

 

[recumpence]

8) Looking forward to this....

You and me both.

So, I know someone with a really slick road bike with a beautiful belt drive who NEEDS one of these E-loggers.

Hmm, I wonder who that could be? :wink:

Matt

 

[deecanio]

Hi Matt,

thats some really slick looking kit there, best of luck, im hoping you can lead us to the promised land
more pics please


Cheers


D

 

[recumpence]

I will post as much information and pictures as I can.

I am already very familiar with the E-logger (I have two of them in RC models). They are fantastic. The controller looks interesting. It is not nearly as programable as the Castle Creations controller. But, I have been told by a number of RC plane racers that they have smoother startup software and more tractable throttle "Feel". We shall see. It also has a slw ramp up feature that can be set for 6 second ramp, 8 second, 10, or 12. 6 second ramp might be a good thing to program in. I will try it out.

The GPS is another interesting item. I may love that thing or hate it. I know I can only display 4 items at a time on my digital readout. So, I will probably choose not to display any GPS data, but rely on pulling all GPS data off the logger straight to my computer after each run for analysis.

Anyway, I do not want to rush this. I want to take my time so I do not miss anything. I also need to replace a bearing in my drive unit today.

Matt

 

[Miles]

So, I know someone with a really slick road bike with a beautiful belt drive who NEEDS one of these E-loggers.

I have a Cycle Analyst on my bike, at the moment - I don't need to know the poor performance of my present batteries in any greater detail Just about to order some new batteries - when they come, I'll put the eLogger on and do some proper testing :wink:

 

[deecanio]

Hi Miles,

what did you decide on the batteries?


Cheers,

D

 

[Miles]

Hi D,

http://endless-sphere.com/forums/viewtopic.php?f=3&t=5339&p=81383

I was just about to send you a PM :wink:

 

[recumpence]

This simple controller upgrade has become a complete bike rewire and reconfigure job.

I took the time to make a decent harness for the entire electrical system complete with disconnects for all components. I had to lengthen leads and do ALOT of soldering along with wrapping the harness in spiral wrap and generally cleaning everything up.

I will post pics tomorrow.

Oh, it is getting harder and harder to get all components mounted under the seat. :wink:

Matt

 

[recumpence]

OK, good news and bad news (more bad than good).

My Power Jazz took a dump on me as soon as I started it up. I installed it and went outside with the bike. I powered it up and heard the arming tones. Then I gave it a touch of throttle. The motor just whistled for a half second, then I got a heavy 8 to 10 inch flame like a rocket out the side of the ESC. I had to reach down near the ESC to unplug it. I singed the hair on my arm doing so. It actually burned some wiring and a bit of the fiberglass under the seat. :x So, I am sending it back. I will get a refund as soon as they pronounce it dead.

Hmm, the good news (shadowed by more bad news) is that I got the Eagle Tree data logger hooked up. I hooked up another HV110 controller from my 100+ mph RC car. That controller was heavily beat from pulling 130 amps through it on my high speed runs and, as you can guess, I fried that ESC. Hmm, not too good. However, that being said, I got good data from my Eagle Tree;

#1 My bike requires 350 to 400 watts to cruise at 20mph. Not the best, but not terrible. With the tire off the ground, it requires 300 watts to sped the rear wheel at 30mph and 400 watts to spin it at 42mph. That is good data to know.

#2 The reason I fried my second HV110 (the beat up one) is I pulled waaayy in excess of 5,000 watts through it a number of times. Heck, I could have pulled 7,000 through it if I tightened the clutch and went for it. :wink: So, needless to say, the data logger is fantastic! Now I just need to find an ESC that can drive this motor to its potential.

I contacted Castle Creations about their industrial ESCs. However, they are not available yet, and they are super expensive. Plus you have to buy the software separately. The next option is their Hyra series. The downside of those is they are liquid cooled because they are designed for boat use. The high side is, they use the same software as all their other recreational controllers (fantastic software) and their biggest Hydra esc is a 240 amp monster designed for 50 volts, 190 amps continuous, 240 amps burst. It has dual runs of 10 guage wire on the input and output to handle the load. I sent them an email to find out if this controller would work well for my requirements. The fact that it is designed for liquid cooling may not be a big deal because the cooling is just a few aluminum tubes along the side of each layer of MOSFETs. I think it will air cool just fine especially considering it will only be run at 20 amps or so most of thw time with the occasional high current burst.

So, one question I have for you electronic guys is;

should I add capacitors near the inputs of the new ESC? I cannot get the cells closer than about 12 to 14 inches of wire away. I can increase the wire guage to maybe 8 guage if need be. Or would added caps near the inputs be good enough?

Let me know your thoughts.

Matt

 

[eP]

should I add capacitors near the inputs of the new ESC?

Yes, you should.

I cannot get the cells closer than about 12 to 14 inches of wire away. I can increase the wire guage to maybe 8 guage if need be. Or would added caps near the inputs be good enough?

You should increase the wire gauge even to 5 gauge if it is possible as far as you need 100+ amps currents.
Look at these recommendations: http://www.powerstream.com/Wire_Size.htm

But still crucial will be your final (motor to wheel) ratio and related battery voltage. (the higher ratio - higher battery voltage you need for the same kv and max speed)

Would you like to post more your idle currents for lower speeds and different (both)motors and related gears ratios ?

Best regards

 

[dirty_d]

you just cant stop breaking stuff can you Matt, lol? thats why you need to get that current limiter set up, the slipper clutch doesnt seem to be cutting it. the ultimate thing we need here is a custom controller for bikes. when these motors are in rc cars or turning a propeller they accelerate to full speed really quick, so the controller doesnt see that high current at low rpm for very long.im going to make my own brushless controller when i make my next bike using one of these motors, i have no idea when that will be, but there are a lot of people here on the forums that know enough about electronics to make one and sell them here. maybe make it an open source thing, provide the schematic and PCB layout, or even mail you the etched PCB and you can order the comonents online. ill definitely do that when i get around to making it.

 

[recumpence]

I cannot give you that data because I no longer have a working ESC to test with.

What is odd, though, is the voltage drop. It never drops below 44 volts under the heaviest loads. But, I know from my car audio SPL days that a small drop in voltage is extremel hard on MOSFETs. That being said, I do not mind going with larger guage wire if need be. I can open up the packs and run larger wire. However, connectors are an issue as well.

What would be good connectors that will handle 150 amps without problems?

Matt

 

[eP]

What would be good connectors that will handle 150 amps without problems?

Maybe something like this one:
http://torberry.co.uk/shop/index.php?main_page=product_info&products_id=46

 

[recumpence]

I just posted this on my thread about my ESC issue. I thought it would be good to post it here as well;



I just got off the phone with Bernie at Castle Creations. He spent 45 minutes on the phone with me. What a fantastic guy. I have dealt with him in the past. Anyway, he mentioned the Hydra HV is perfect for my needs. He also mentioned my issue (and, yes, this has been told to me by Ep and a few others, but it was good that Bernie confirmed it for me). He told me it is not my current draw, but my wiring that is the problem. He said to get the ESC close to the pack (this is not possible with my larger Power Jazz, but is with the CC controller) and run dual 10 guage wire with doubled up connectors to the ESC. He also told me where to get the caps to make my own cap bank before the ESC. Untimately, he said just getting the ESC within 6 or 8 inches at most away from the pack would be my ultimate cure. However, the HV110 is barely adequate for my requirements even with these changes. So, I will order a Hydra HV240 today. I found one source that I can get it from for $400.

So, here's the conclusion I have come up with; New Hydra HV240. Dual 10 guage wire from pack to ESC. Doubled up connectors. Less than 8 inches of wire from cells to ESC and a bank of caps within two inches of the ESC board.

Bernie said current ripple is my enemy here more than current, thought the ripple is caused by current.

Anyway, a few guys have mentioned this to me on this forum and I appreciate their help. It was good to have a long talk with Bernie to explain it all to me. He mentioned I am a good study for them because they want to be in this market badly.

Matt

 

[jnuts]

Dual 10 guage wire from pack to ESC.

Why not switch to 5 or 6 gauge wire?

 

[johnrobholmes]

Dual 10awg is easier to work with vs 5awg.

You could use dual plugs, like a deans plug for negative and a deans for positive. Otherwise the 6.5mm plugs that Castle has are the Bees Knees. That is what I use on my rig.

Once I get my next rig together I will contact you so we can bounce information back and forth. It will be a mountain bike, but the componetry will be very similar.

 

[recumpence]

Dual 10awg is easier to work with vs 5awg.

You could use dual plugs, like a deans plug for negative and a deans for positive. Otherwise the 6.5mm plugs that Castle has are the Bees Knees. That is what I use on my rig.

Once I get my next rig together I will contact you so we can bounce information back and forth. It will be a mountain bike, but the componetry will be very similar.

Exactly. If I go up to big wire, everything gest pretty crazy and hard to work with.

I will be running dual Deans connectors everywhere. I am evey planning on running them to my motor.

At any rate, my Hydra HV240 is on its way.

I bet I can see over 7,000 watts from this thing.

Matt

 

[jnuts]

Any thoughts about throwing in a large circuit breaker before the ESC (200 amp or so, depending)? That way you could provide protection from smoking the controller(s) while you test out your design.

 

[dirty_d]

can anyone explain the need for short wires from the battery to prevent the destruction of the esc? i though the only thing that could blow the fets having anything to do with the battery wires would be the inductance, is there anything else? why would longer wires cause more ripple?

 

[lawsonuw]

The Inductance of the battery wires IS why short wires are needed. First up, nearly all PWM controller for motors draw current from the battery only in short pulses. Inductors by there nature resist rapid changes in current. Add these two together and long wires from the battery could cause the voltage at the controller to go to zero when the FETs turn on, AND this same inductance can cause the voltage at the controller to double (or worse) when the FETs turn off. Voltage drop out is a great way to heat up FETs because it will often remove the gate voltage for a short while making the FET partially turn off. While doubling the voltage is a great way to fry FETs as the extra voltage can force a FET into avalanche break down, aka the FET turns ON due to the over voltage till something stops current from flowing through the FET. Adding large capacitors to the controller input is the usual fix. This works because capacitors resist rapid voltage changes and supply current to the controller during the time it takes for the battery and battery wires to start suppling current.

I remember that the OSMC has a voltage suppression component on it's inputs and outputs. A "TVS diode" digikey p# 1.5KE51CADICT-ND. Looks like it's rated to start clamping the supply at ~48v and never let the supply exceed ~70v. The FET's the OSMC uses are only rated for 55v! The OSMC also fits a 1500uF 50V input capacitor.

Marty

 

[ZapPat]

The Inductance of the battery wires IS why short wires are needed. First up, nearly all PWM controller for motors draw current from the battery only in short pulses. Inductors by there nature resist rapid changes in current. Add these two together and long wires from the battery could cause the voltage at the controller to go to zero when the FETs turn on, AND this same inductance can cause the voltage at the controller to double (or worse) when the FETs turn off. Voltage drop out is a great way to heat up FETs because it will often remove the gate voltage for a short while making the FET partially turn off. While doubling the voltage is a great way to fry FETs as the extra voltage can force a FET into avalanche break down, aka the FET turns ON due to the over voltage till something stops current from flowing through the FET. Adding large capacitors to the controller input is the usual fix. This works because capacitors resist rapid voltage changes and supply current to the controller during the time it takes for the battery and battery wires to start suppling current.

I remember that the OSMC has a voltage suppression component on it's inputs and outputs. A "TVS diode" digikey p# 1.5KE51CADICT-ND. Looks like it's rated to start clamping the supply at ~48v and never let the supply exceed ~70v. The FET's the OSMC uses are only rated for 55v! The OSMC also fits a 1500uF 50V input capacitor.

Marty

Very good explanation, Marty!

One problem with these RC controllers is that they are made to be very close to the battery, so thus they can get away with using less bypass capacitors on their main DC power bus.

Think of capacitance being the opposite of inductance. So the more inductance you add into your circuit (between the battery and controller), the more bypass capacitance you also need in (or very close to) your controller. Also make sure these capacitors are good quality ones. I use the Chemi-com KZE series as they have very low ESR for the price, but you may also want to put a good film type capacitors too into the mix too (this is for high frequency bypassing if needed). As others have mentioned, always put these as close as possible to your controller's DC input - specially so for the high frequency film type caps if you use these too.

One thing also to remember is that inductance is not really wire length, but really wire loop area. This means that it's better to run your DC bus wires (+ and -) right beside one another, and twist them also if possible.

 

[scottclarke]

Matt,

Very sorry to hear about the Jazz frying - I do hope mine doesn't suffer the same fate when it finally arrives! I'm curious - did you test it with no load on the wheel and it was fine then went pop when loaded?

As regards connectors - search for powerpole on ebay - genuine andersons are used on commercial gear (read forllifts) with no issues up to and above the rated current. I am about to order some 120A connectors from a seller on ebay for use on my ride and expect them to perform very well, based on experience.....

As far as wiring goes - for my up and coming 7 parallel pack the 8 gauge wire I use has over a thousand cores - split into 7 bundles would round each other down the wire - standard stuff. Terminating the 8 gauge is hard work unless you have the right crimp tool and connectors (which are big and bulky) and I prefer a soldered connection anyway so what I have done is heatshrink double over each of the seven cores then a bit at the split. I have done this for plus and minus with the ends terminated - for now - with 30A chocolate block connectors (until the 30 small andersons arrive). This is perfect for my setup but I can see no reason you can't use a similar method to terminate large gauge wire....

My 8 gauge has been good on 120A - just gets warm if you sustain 120A for more than a few seconds but the insulation appears to be good for 150C so no worries at all really..... I got it from maplin - its the second from largest cable they sell..... Of course - doubling it up is the real way to go for zero losses - and I may just do that if the Jazz works as it should!

BTW - what gauge are the Jazz input leads?

Scott

 

[recumpence]

The Jazz has 10 guage inputs.

My Jazz ran without load. But, when I sat on the bike, it didn't ever try moving the slightest bit. It just blew up. I think I got a bad one.

There are two reasons I am going with doubled up wire and doubled up Dean's connectors;

#1 My charger and everything else I use has them. So, I will not have to change anything.

#2 The Hydra ESC uses doubled up wires already. So, it is natural to go that route.

I have a car audio background. I am very familiar with using wire all the way up to 00 gauge. But, I want to stick with what is easy and compatible with what I have. Of course, I do not mind going crazy with custom buss bars and what-not. I just don't want to if I don't have to.

Now the next question is;

Who knows of good 50 volt low ESR caps I can run on my ESC input?

Matt