Lyen regl.: variable regen, maximal current from 5V

Hi all,

i design cyclocomputer(similar as cyclo anylist) based on arduino and LCD 128x64pixels.
I have these questions about regulator Lyen 12FET:

1.How much mAps(current) can I take from 5V regulator Lyen for supplying my cyclocomputer?
5mAmps take throttle and 20mAmps takes hall in MPie, my cyclocomputer need about 30mAmps
2.I want control intensity regenerative braking by PWM signal. The PWM signal controls gate of MOSFET connected on conector "regenerative braking jumper wire". Is it possible?



Thanks for comments,
Rob.

I think I asked this question a while ago.. pretty sure the answer is, no.. becuase of the way regen is handled in the software of the infineon controller IC.. it may work with limited results but I don't think the input is designed / programmed to receive a PWM signal..

others can confirm.

Good luck.

Andy

I have used the psu on these controllers ( +5 0v from the throttle supply but its actually more like 4.3v ) for around 30ma without any problems in the past but now I have started to use these.



They are so cheap its not worth considering anything else , I have not tested up to their full capacity (1A) but have test upto 500ma and are perfect they give out 5.4v and are tiny.
http://www.ebay.co.uk/itm/170969138830?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2648

as regards the regen I would would say try it and see what happens , what's the worst that can happen and it should be very easy to setup and program.

I've considered attempting that PWM of the brake line, but all of the controllers I have had so far, regardless of origin or design, respond with a significant delay to brake input, so that it takes some significantly nonzero time to begin shutdown of the motor and begin braking, and even longer to shutdown braking and begin restart of the motor.

I suspect that it isn't possible to do this because of it, but if your controller is either not delayed in it's operation of brake/resumption of motion, or it's electronics are such that the delay is caused by a removable filter (R/C, etc) on the input rather than in the software or MCU hardware, then you may be able to do this.

Gwhy:
Sales of psu was already ended. Do you have any suggestion with low power consumption
?

Yes, that true. Supply 5V dropped to 4.2V, the output of 7805 is probably a limiting resistor.

Rob,

As previous posters stated, turning the ebrake on/off using pwm is unlikely to work. Try it on a very cheap controller first as it's likely to mess up the controller. I somehow ended up with a pulsing signal on an ebrake line by accidentally attaching it to a wire for a speedometer, and while the +5V was ok for the ebrake signal, the rapid pulsing killed something on that part of the board. Regen never worked again, and during the short time it did the delay in regen made is sporadic not pulsed, which confirms to me that PWM of the ebrake signal won't work.

OTOH, varying the shunt resistance does vary the force of regen braking. On higher rpm motors like scooter motors used with smaller wheels, even when programmed to the highest regen setting regen force is often too gentle, because the current limits of the regen function are too low. On several bikes I reduced shunt resistance directly to fool the controller into allowing more current during regen, and then I reduced the programmed phase and battery current limits to get the actual limits back to where they were before shunt change.

As similar effect can be obtained by hacking into the current sense trace and introducing a voltage divider circuit, and attaching a variable pot to the ebrake handle. You'll need to be sure that when the variable pot on the brake returns to the off position that forward current limits return to normal, but overall the hack is relatively simple. I tried it once, but my lack of the tools, experience, and skill to do the fine soldering required is what prevented success, not an issue with the approach. I proved the approach by changing shunt resistance directly, and changing regen force permanently in that manner.

Though these discussions were directed at obtaining variable forward current, installing the pot on the ebrake will result in variable regen. see http://endless-sphere.com/forums/viewtopic.php?f=2&t=31643#p494749 and http://endless-sphere.com/forums/viewtopic.php?f=2&t=37068&start=25.

If you come up with a refine approach, please share the details.

John

rob.brno said:

Gwhy:
Sales of psu was already ended. Do you have any suggestion with low power consumption
?

Yes, that true. Supply 5V dropped to 4.2V, the output of 7805 is probably a limiting resistor.

Click to expand...

Yes the sales have finished for this listing but there are lots of sellers that sale identical ones, just look for the cheapest price . The arduinos will run ok on the throttle line of the controllers they work down to as low as around 3v .

Edit: the 5v outputs on the infineons i.e throttle line normally goes through a blocking diode and this is where the 0.6vi'sh is lost.


I expect these psu's are identical to the ones i posted eariler, just in a different enclosure :

http://www.ebay.co.uk/itm/EU-PLUG-Micro-USB-AC-ADAPTER-MAINS-POWER-WALL-CHARGER-FOR-APPLE-IPHONE-4-4S-5-5s-/231323734019?pt=UK_MobilePhones_MobilePhoneAccessories_MobilePhoneChargers&hash=item35dbf81003

http://www.ebay.co.uk/itm/NEW-EU-PLUG-USB-AC-ADAPTER-MAINS-POWER-WALL-CHARGER-FOR-APPLE-IPHONE-5-5S-4-4S-/161367187170?pt=UK_MobilePhones_MobilePhoneAccessories_MobilePhoneChargers&hash=item25923bdee2

http://www.ebay.co.uk/itm/Micro-EU-Plug-USB-AC-Home-Wall-Travel-Charger-Adapter-for-Apple-iPhone-5-5S-4-4S-/121427364269?pt=UK_MobilePhones_MobilePhoneAccessories_MobilePhoneChargers&hash=item1c45a289ad

I have not looked at this DC-DC converter but again I expect its the same as the ones above but with a output lead instead of a usb socket.

http://www.ebay.co.uk/itm/EU-Plug-Rapid-Fast-Micro-USB-Home-Wall-Travel-AC-Charger-For-Samsung-Galaxy-S4-S-/400767060245?pt=UK_MobilePhones_MobilePhoneAccesories_MobilePhoneDataCables_JN&hash=item5d4f940915

They have worked well down to 36v DC input

A couple of thoughts:

I have walked down this path not too long ago. The short answer is that you can enable some change in the Regen during operation, however the application of the techniques described in changing value of R12 or the Current (Shunt) Sense create a very unwelcomed and compromised riding experience. On top of which I blew up two or three good controllers during experimentation. Read all about it here:

Solving the Variable Ebrake-Throttle Problem

There is no way to make a silk purse out of a sow's ear. The controllers are the way they are, and without the ability to change the firmware (provided that we could), there are no hardware solutions to create variable regen via mods to the PCB that are useable or safe. Yes - we can affect regen on the fly by interposing voltage or changing resistance; however to get the controller to recognize the new R12 value (safer than messing with Current Sense) required a controller reboot which created a lag of 0.5 seconds or more.

The 7805 VR, like the LM317 are capable of supplying up to 1.5A. The voltage drop at the throttle will be due to either the built-in POT or HE device, which ever one style is employed. The link I provided above goes into more detail about that.

Safe travels, KF

Ah, the old "I couldn't get it to work, so it can't work." conclusion. You were told the R12 approach wouldn't work, but bagged your head against that wall for weeks anyway. I stopped following the thread before you got around to trying a variable voltage divider on the current sense, so I don't know what you did to make it unsafe. Based on the OP's original plan, he's not trying to make regen stronger, just make engagement smoother by starting with low force and increasing it to what the controller does now. That should be easy and safe for the controller by using fixed resistors to set hard limits for the range change in apparent shunt resistance. The easy way to prevent what you're doing with regen current from affecting forward current and the potential for handling pitfalls is to have a CA limit of forward current, since the CA's current sense is unaffected. That way the voltage divider only changes regen force. There are all kinds of ways to skin a cat.

John in CR said:

Ah, the old "I couldn't get it to work, so it can't work." conclusion. You were told the R12 approach wouldn't work, but bagged your head against that wall for weeks anyway. I stopped following the thread before you got around to trying a variable voltage divider on the current sense, so I don't know what you did to make it unsafe. Based on the OP's original plan, he's not trying to make regen stronger, just make engagement smoother by starting with low force and increasing it to what the controller does now. That should be easy and safe for the controller by using fixed resistors to set hard limits for the range change in apparent shunt resistance. The easy way to prevent what you're doing with regen current from affecting forward current and the potential for handling pitfalls is to have a CA limit of forward current, since the CA's current sense is unaffected. That way the voltage divider only changes regen force. There are all kinds of ways to skin a cat.

Click to expand...

Now John, play nice. You should go back and read my thread to the end; who knows... you might learn something.

The point is that the controller cannot be modified so that you can safely employ variable regen on the fly. I never discussed increasing it or any other nonsense; I wanted to do exactly the same as the OP desires and vary regen-braking.

The problem is that the controller reads the values on-boot or on first-signal from the ebrake – depending on which method is attempted, and does not read the change in value during the braking event; it reads the value once and handles it. Quite simple really, except that it doesn’t help us solve the problem dynamically. And given that the MCU is fused, there’s no way to make a proper fix.

I think the better solution is to pursue a higher quality controller offering variable regen-braking natively, or at least one that we can program to enable the feature, the latter of which IMO is the smarter move.

~KF