The Cycle Satiator, universal charger for the enthusiasts

[dogman dan]

I just think it's badass that finally we have an on board charger for bikes.

Carrying around kingpans, I'm just fretting all the time they will get bounced to death.

 

[circuit]

Carrying around kingpans, I'm just fretting all the time they will get bounced to death.

Talking about bouncing, what has been done to make this one capable of surviving road bumps? It would be very interesting to see it's insides..

 

[justin_le]

Talking about bouncing, what has been done to make this one capable of surviving road bumps?

We subject it to both SAE J1455 (4g's) and UL 458 (2g) vibration testing and passed both, but I think that the real test is to mount it to the deck of a skateboard that gets ridden for a weekend at a maker faire. If you want to shake the hell out of something till it fails that seems to be a surefire approach!

It would be very interesting to see it's insides..

Much obliged:


Most of the larger parts have a silicone adhesive for vibration dampening and the larger transformers are all sandwiched between the PCB below and a thermal pad on the the enclosure lid. A number of the failures we've seen first hand on the "high power" brand chargers came from the transformer leads cracking at the PCB trace, since the only thing supporting the entire transformer mass was the copper wire to the board.

 

[circuit]

Thanks! That looks like a decent design.

 

[worldpax]

While I am not nearly qualified to be a beta tester. I will be a customer when the final product comes out.

 

[justin_le]

Hey all, I've been hoping to send a "your beta chargers have shipped" post for quite a while now but we've been hung up on a few firmware tweaks that have taken a lot longer than expected to iron out. So I thank you for the patience of those who have ordered and are wondering what is up. At this point the details are mostly cosmetic, so if it's not sorted by the end of today we'll still go ahead and ship out tomorrow with a small note about the known issues in the V0.807 firmware build, and hopefully by the time they arrive there will be a new firmware to install anyways.

 

[liveforphysics]

Are those Litz wound soft magnetic cores by chance? No wonder you've got extremely high efficiency, that is baller stuff my friend!

It looks like I'm seeing a high voltage PFC front-end feeding a 3 phase transformer assembly that gets rectified and paralleled and fed through the final very nice looking Litz smoothing output inductor and DC filter caps on the far right side of the photo?

If so, that's clever to break the phases up so you deal with something like 1/3rd (maybe less?) of the ripple current/voltage to smooth on your DC output. I like it!

Did you see Tesla just opened all it's Patents today my friend? The world is awakening.

http://www.teslamotors.com/blog/all-our-patent-are-belong-you

 

[neptronix]

I really wanted to like this, but the low voltage is a deal-breaker unfortunately. I already have a hyperion 1420i that is capable of doing this ( it's just not nearly as nice to carry around, lol.. )

I understand the position where legality requires that >60v means that you have to spend crazy money, making it uneconomical as that cost has to be factored into the price, which i find to be high already.

Though something that could be adjustable up to say 100v would be filling a niche in the ebike charger world.

Maybe this is worthy of a kickstarter campaign to bring a very high voltage capable version to life.
Judging on how many backers there would be, you can gauge how many people are willing to pay the premium for something capable of charging 15-24S lipo setups that are so common.

Just a thought.

Otherwise it's cheap crappy china chargers and parallel-series swaps for us high voltage guys... which is less than desirable.

 

[cal3thousand]

I really wanted to like this, but the low voltage is a deal-breaker unfortunately. I already have a hyperion 1420i that is capable of doing this ( it's just not nearly as nice to carry around, lol.. )

I understand the position where legality requires that >60v means that you have to spend crazy money, making it uneconomical as that cost has to be factored into the price, which i find to be high already.

Though something that could be adjustable up to say 100v would be filling a niche in the ebike charger world.

Maybe this is worthy of a kickstarter campaign to bring a very high voltage capable version to life.
Judging on how many backers there would be, you can gauge how many people are willing to pay the premium for something capable of charging 15-24S lipo setups that are so common.

Just a thought.

Otherwise it's cheap crappy china chargers and parallel-series swaps for us high voltage guys... which is less than desirable.

Why stick with parallel-series swaps? If your pack can be broken into parallel-series configurations for charging, it could clearly be broken into 2 halves and charged with 2 of these things.

I'm planning on charging 24S2P with this thing, but using parallel-series swap at first. When, and if, I get a second unit, each one will just charge it's own 12S section.

Also, allowing me to charge a 13S setup without getting inside and swapping resistors or turning pots is huge since I will be playing around with those newish 52-cell bottle battery boxes at 13S4P

 

[liveforphysics]

His charger is isolated, so it only requires firmware and enable you to set it's 0-60v range on top of whatever constant voltage source you need if you want to charge higher voltage packs with it.

You can charge any pack that doesn't have an empty to full voltage difference greater than 60v.

Want to charge 30S of RC LiPo? Put a 70V constant voltage supply under this charger, and with only a firmware change you're set.

 

[neptronix]

Why stick with parallel-series swaps? If your pack can be broken into parallel-series configurations for charging, it could clearly be broken into 2 halves and charged with 2 of these things.

I'm planning on charging 24S2P with this thing, but using parallel-series swap at first. When, and if, I get a second unit, each one will just charge it's own 12S section.

Also, allowing me to charge a 13S setup without getting inside and swapping resistors or turning pots is huge since I will be playing around with those newish 52-cell bottle battery boxes at 13S4P

Parallel-series swap is just as inconvenient as breaking the pack apart and charging each part separately. Same problem. What's ideal is plug, play, and forget.

Now if this charger could do the full voltage, i would be willing to pay the premium just for the convenience and form factor alone, and accept the lower charge rate.

But i'm a cheapass. If you can afford to buy two of these and use them that way, go for it.

 

[neptronix]

His charger is isolated, so it only requires firmware and enable you to set it's 0-60v range on top of whatever constant voltage source you need if you want to charge higher voltage packs with it.

You can charge any pack that doesn't have an empty to full voltage difference greater than 60v.

Want to charge 30S of RC LiPo? Put a 70V constant voltage supply under this charger, and with only a firmware change you're set.

Are you 100% for sure?
Because series-ing this baby up to a surplus/pulled 48v server power supply would be a golden solution. I mean, those are what.. $10.. $50 at most..

That would be a determining factor on whether this cheapass buys one or not.

 

[parabellum]

How about unit that has everyting ready to output 100V (for reliability of course) but oficially 60 V max? Then unofficial 3rd party software apears (unexpectedly), capable of crazy things and voiding warranty of course.

 

[liveforphysics]

His charger is isolated, so it only requires firmware and enable you to set it's 0-60v range on top of whatever constant voltage source you need if you want to charge higher voltage packs with it.

You can charge any pack that doesn't have an empty to full voltage difference greater than 60v.

Want to charge 30S of RC LiPo? Put a 70V constant voltage supply under this charger, and with only a firmware change you're set.

Are you 100% for sure?
Because series-ing this baby up to a surplus/pulled 48v server power supply would be a golden solution. I mean, those are what.. $10.. $50 at most..

That would be a determining factor on whether this cheapass buys one or not.

The only thing in my life I'm 100% certain of, is that all that currently composes my model of reality (and each individual consciousnesses model of reality) is entirely and completely wrong, paradoxically including the certainty that it's all wrong.

From my best guess, unless there is some technical hardware limitation that I'm not currently accounting for, it seems it would only require a firmware change. The most simple option being entering in a given CV voltage you have in series with it, and all it's same cool functions of measuring energy and power and things would be displayed correctly. An improved option that I have no doubt someone will hack theirs to do would be to leverage some unused A2D pin on the uController to be a kelvin measurement of the voltage at the pack, so it compensates for all charge lead resistive drop along with whatever other voltage supplies you may have in series with it (they wouldn't even need to be regulated supplies that way, the charger could simply control current based on the kelvin voltage input at the battery connection, and compensate as an unregulated supplies voltage sagged under load or whatever).

If you always plugged things in the correct order of operations (always start the CV supply first before connecting to battery!), and then never had the CV supply trip-off for some reason during charging, you wouldn't require additional charger protections.

However, in a world where people do trip over cords and unplug them, or blow fuses and breakers and things, the charger must have protections to prevent from being exposed to full pack voltage. My guess is this is the reason why Justin has seemed hesitant to create the firmware. I know for certain it is possible to make adequate protection to protect the charger in the event of CV supply failure or other issues, and it will likely cost under $10 bucks in components. Likely the cheapest way would be a few well selected TVS diodes, a full charge current capable diode, and perhaps a pair of fuse holders for ultra-fast-blow DC-rated fuses, one on each side of the TVS diode stack, so whatever event may have caused the TVS to clamp can be protected from an event on either side for total protection.

 

[justin_le]

While, it's looking like a 72V model may be more feasible (and available sooner) than expected . We've been going over it in some detail the past week and it might not be that bad. Because the satiator doesn't actually output any voltage and turn on the output switch until it detects a battery plugged in, then there isn't ever a live voltage present on the exposed pins which simplifies the connector spec. It does mean some planned firmware features like generic power-supply mode might have to be disabled, but that's a small price to pay. We'll be verifying the more stringent hi-pot isolation tests and spacing requirements but it could be reasonable with just minimal layout modifications.

How about unit that has everyting ready to output 100V (for reliability of course) but oficially 60 V max? Then unofficial 3rd party software apears (unexpectedly), capable of crazy things and voiding warranty of course.

There is always that option too. However, I think we can do this legit, and legit means that it could be used in OEM performance ebikes which would definitely help get the numbers up. Of the high voltage users here, is a max output of 94-95V (ie 22S regular lithium or 26S LiFePO4) sufficient for 90% of you, or would the lack of 24S support (100V max) be a bummer? Keeping in mind that a higher max potential voltage does generally mean a lower max current when used at lower voltages. The top 75-80% of the voltage range will generally be 360W power limited. So at 90V it would be 4A, 80V would be 4.5A, and 72V would be the full 5.0A.

 

[teslanv]

24S is a sweet spot for battery building. Multiples of 4S or 6S are ideal, and 24S works for both. And with 4115 FET controllers quickly becoming popular, I'd bet there would be strong demand for 100V capability.

 

[liveforphysics]

If it also would help others out there, I would prefer 28s simply so it could be a compatible lightweight onboard or offboard charger for all Zero's or Brammo batteries, as well as retaining the awesome lab power supply function (which is so cool and worth the device cost alone!), even if the lab power supply function is limited in programming to 0-60vdc.

 

[parabellum]

I use mostly 24s and sometimes 32s. Mostly 4s and 6s steps as teslanv writes

 

[jkbrigman]

Justin: +1 for 24S LiPo.

I just bought two 48V Meanwell LED supplies with the plan to charge a 24S pack split into two 12S blocks. But a 24S-capable Satiator and suddenly life gets really simple.

 

[Ypedal]

i have lots of 36 and 48v bikes so this unit is going to get a workout when it arrives... but as others i have a few 24S lipo bikes... if i was to series connect this with a big 48v meanwell capable of well over 8 amps, set to a fixed voltage, could/would the satiator work as a current limiting device to prevent the MW from melting down ( i know from experience they do not handle full power well at all... )

I realize the satiator will only show voltage and ah from it's own source, still good enough for me just to know Ah back into my pack.

example : set open circuit voltage on the MW to 49v ... set the Satiator to 50v, let'er rip..

bad idea ?.. possible .. but with risk ?

 

[parabellum]

Would be cool, if Satiator could switch off something else on charge end or shortly before or in fault event, like relay. This way, some badass/badboy charger (Capacitive reactance type for example) can pump real amps in parallel and being switched off for final topping by Satiator. Bad idea?

 

[justin_le]

example : set open circuit voltage on the MW to 49v ... set the Satiator to 50v, let'er rip..
bad idea ?.. possible .. but with risk ?

This is what we were discussing at some length a little while back with liveforphysics etc. Short answer is that so long as the meanwell acts as a nicely behaved constant voltage source then it shouldn't be an issue. It would just look like you are charging a 50V pack that happens to have twice the voltage swing from empty to charged as a normal 50V pack. So when you generate the profile for this setup, you'd want to set your MinStart Voltage and trickle voltage thresholds accordingly, to values that are a a fair bit lower than normal. You would need to be very careful that you never accidentally set the meanwell voltage down to like 20V while it is in the midst of charging because then the satiator will see closer to the full pack voltage and have a fault.

 

[justin_le]

Would be cool, if Satiator could switch off something else on charge end or shortly before or in fault event, like relay. This way, some badass/badboy charger (Capacitive reactance type for example) can pump real amps in parallel and being switched off for final topping by Satiator. Bad idea?

A little ridiculous yes. BUT, if you get a satiator, you'll hear a definitive relay click at the end of charging when the power stage turns off to reduce the quiescent drain down to <1W in order to meet some california energy regulations. I suppose an adventurous user could wire into this relay drive signal for activating or disabling other stuffs. It's not exactly a supported usage case, but who's to stop ya..

 

[liveforphysics]

Would be cool, if Satiator could switch off something else on charge end or shortly before or in fault event, like relay. This way, some badass/badboy charger (Capacitive reactance type for example) can pump real amps in parallel and being switched off for final topping by Satiator. Bad idea?

A little ridiculous yes. BUT, if you get a satiator, you'll hear a definitive relay click at the end of charging when the power stage turns off to reduce the quiescent drain down to <1W in order to meet some california energy regulations. I suppose an adventurous user could wire into this relay drive signal for activating or disabling other stuffs. It's not exactly a supported usage case, but who's to stop ya..

I know I could pry one apart to find out for myself, but do happen to know if the device happens to have any currently unused A2D pins that could be used for reading total voltage output to pack, power supply and Satiator combined voltage at the battery?)

 

[neptronix]

24S seems to be common enough that not reaching that voltage becomes a deal breaker for many.

Of the high voltage users here, is a max output of 94-95V (ie 22S regular lithium or 26S LiFePO4) sufficient for 90% of you, or would the lack of 24S support (100V max) be a bummer? Keeping in mind that a higher max potential voltage does generally mean a lower max current when used at lower voltages. The top 75-80% of the voltage range will generally be 360W power limited. So at 90V it would be 4A, 80V would be 4.5A, and 72V would be the full 5.0A.