US 2.0 question, Alex of Ultrasabers! Need ya!

[Eandori]

Alex of Ultrasabers, can you answer this question? Before I fry a few LED's testing I would really rather just hear it from you first.

When you pick a certain LED to drive on the board, does the board ramp up power to the LED until a voltage is met, or a current value is met? If it's a current value target that the board shoots for.... then could I connect a Luxeon V green to the board and set it at Luxeon III setting so I would effectively run 1.4A through the Luxeon V? Yes I realize that I would need a battery around 8v+ to do this, but would that work?

Similarly, if I take a red luxeon III and config the board in "luxeon V mode" will it drive that red luxeon III at 850mA or will it try to drive the luxeon III red up over 6.8v?

My reason for asking is....

Perhaps I could set the board for Luxeon III overdrive mode (1200mA) and then turn on the flicker. So I would effectively have a luxeon V that hovers around the value of 850mA yet with flicker.

[Strydur]

I am pretty sure it is current..

I would shoot him a E-mail as I dont think he comes on here much

While you are at it find out of it will support multiple LED's

[Goodman]

On another forum, Ultra said that yes, you can set the board to the "Lux III Overdrive" setting to get more brightness on a Lux V. In fact, we found that doing so gives ~5-10% more visible brightness than the regular "Lux V" setting.

This test was done with two 3.7V rechargeable 14500 cells.

When one of the guys hooked up a third cell, for 11.1V, he reported an additional 60% more visible brightness.

'Hope that helps.

[Eandori]

Sounds good, I'll do that.

Also, I have an idea on how we can fix the battery problem. Here's a spec for the microcontroller that Alex used on the Ultrasound board. It's a microchip PIC
http://ww1.microchip.com/downloads/en/DeviceDoc/41303D.pdf
If you go to page 53 there are some notes about the reset pin, called master clear on this chip. MCLR_N is actually pin 1 on the microcontroller.

The circuit shown on page 53 would fix the problem of slow power up or ringing power when cells are inserted. There's only one issue... if Alex programmed Pin 1 to be an input and not an external reset line then this fix won't work.

I for one am having many issues with my boards. I'm using the 3 pin recharge port to kill power and MOST of the time when I remove the kill pin the board does not boot. I'm fairly certain that's because the 3 pin recharge port has a bit of ringing when the kill pin is pulled. As long as Alex didn't program pin 1 (MCLR_N) as an input this could work.

I also need to ask him how that pin is connected to the rest of the board. I'm guessing tied it high through a resistor... But I don't know for sure.

Anyways, the circuit shown on page 53 and a switch debouce circuit with a D-Flip flop might indeed fix the major issues with the US 2.0 board. Let's see what Alex says.

[Eandori]

--------------------------------------------------------------------------------

On another forum, Ultra said that yes, you can set the board to the "Lux III Overdrive" setting to get more brightness on a Lux V. In fact, we found that doing so gives ~5-10% more visible brightness than the regular "Lux V" setting.

This test was done with two 3.7V rechargeable 14500 cells.

When one of the guys hooked up a third cell, for 11.1V, he reported an additional 60% more visible brightness.

'Hope that helps.
Great news, thanks!

By the way, in that setup you just described the two 3.7v cells are not apparently enough to really drive a luxeon V saber properly. Remember that batteries don't work perfectly, each cell is built for and rated at a value of current flow. If you pull more current then it's rated current the battery voltage will drop significantly, even if fully charged. For example, if you stack up hearing aid batteries and measure out 8v with nothing connected, then try to pull 1 amp out of them, not only will the die quickly but you'll see the voltage drop VERY low. Probably it won't even get up to 1 amp for cells that small.

A 2 cell pack rated at 7.4v might not be able to source the 1 amp of current without dropping it's voltage below 7v. A luxeon V LED needs a voltage around 7v to actually drive about 1 amp through it.

Now the final part...

If your electronic driver board is getting 7v on input, chances are it will only give the LED around 6.2v to 6.6v. Boards cannot give out more voltage then they take in, it's always less. Unless the board has a charge pump for doing voltage up-conversion (only micro buckpucks at 350mA have that I think)

Long story short, if you are building a saber with luxeon V LED's I would use a pack rated at 8.4v or 9.6v. That's exactly why I build my Luxeon V sabers with packs made from 7x NiMH AAA's. 1000mAh of capacity gives around a hour of runtime, it's VERY compact and will fit in the saber well, and NiMH is very simple to charge with low cost peak chargers.

Don't go much above that with Ultraboard or it will get hot. Don't go much lower because then your LED is under-driven. Battery selection is MORE important on the ultraboard because it does less power management then other boards.

Cheers,

[Goodman]

I hear you on the power issue. I'm the one for whom Alex made the YouTube video showing that the board can indeed take three 3.7V cells (initial voltage is obviously above 11.1V, but he said on the video that it dissipates quickly). I asked specifically because I knew the Lux V's *minimum* is the 7.2V two 3.7V cells provided.

I looked at 8.4v and 9.6v NiMH packs, but the problem is all my sabers only have room for three AA cells in a triangle, which means I have to use the 3.7Vs.

The OTHER problem that this situation then brings is that I'm now having an issue with the custom triple cell holders I built. Probably lots of resistance in there somewhere. I may have also almost fried the voltage regulator on one of my boards (good thing I have three more boards, and that spare voltage regulators are only $1.25 apparently). I'm afraid to plug it back in with three cells, so right now I'm either using a spacer in the third port, or using a two cell holder.

THIS leads me back to the unfortunate fact that I'm back to underpowering the Lux V with my two 3.7V cells. *sigh*

For the moment it happened though, that Lux V on 11.1V was *BRIGHT*!!! 'Definitely worth another few hours of tinkering to get it all to work.

Let me know what Alex says about the Pin 11/input question.

[Novastar]

Add a (small) heatsink to the voltage regulator. :)

[Eandori]

I looked at 8.4v and 9.6v NiMH packs, but the problem is all my sabers only have room for three AA cells in a triangle, which means I have to use the 3.7Vs.
No, not AA, AAA cells. You use TCSS machined hilts right? Well you can get SEVEN AAA's in a "barrel roll" glue together and wired up to fit inside a TCSS saber body. In ONE cell length of space! One cell in the middle, 6 around it. Glue them together first (superglue) then wire them up in series. Use 1000mAH NiMH cells.

Don't bother trying to buy a pre-made pack. Nobody makes them like that. You will have to custom make it.


The OTHER problem that this situation then brings is that I'm now having an issue with the custom triple cell holders I built. Probably lots of resistance in there somewhere. I may have also almost fried the voltage regulator on one of my boards (good thing I have three more boards, and that spare voltage regulators are only $1.25 apparently). I'm afraid to plug it back in with three cells, so right now I'm either using a spacer in the third port, or using a two cell holder.
Not sure resistance is your issue there... Voltage regulators fry when they are not heat sinked and when they are over driven. Use a battery voltage that is around 1v above the needed voltage for your LED and don't insulate the board. Let heat flow off it the best you can.

[Darth Morbius]

Check your PM's Eandori. :cool:

I have sent you an email link for the man who built and programmed the boards.

[Goodman]

No, not AA, AAA cells. You use TCSS machined hilts right? Well you can get SEVEN AAA's in a "barrel roll" glue together and wired up to fit inside a TCSS saber body. In ONE cell length of space! One cell in the middle, 6 around it. Glue them together first (superglue) then wire them up in series. Use 1000mAH NiMH cells.

Genius! So I suppose I'm going to go and buy seven AAA sized li-ions now and do what you suggested. Yes, I do use the MHS pieces. This particular piece is the milled body connector that is 3 5/8 inches long. I'm cramming both the battery pack and the speaker in it.

Do you have a picture of your pack that you described? How did you connect all seven in series? Did you solder directly to the cells to save wiring space, or did you fabricate a holder with seven connections on it? *eep* Also, I must then assume that you're using a charge port for this custom pack...have you seen any issues with the cells during charging? I thought they slightly expanded during charging, which I imagine would be hampered by all the superglue (maybe I'll use hotglue instead, which has some flex to it). Or was that only in li-ions...



Not sure resistance is your issue there... Voltage regulators fry when they are not heat sinked and when they are over driven. Use a battery voltage that is around 1v above the needed voltage for your LED and don't insulate the board. Let heat flow off it the best you can.

I'm hoping that the board makers properly heatsinked the VR in the first place. Ultra's video showed the board can take 11.1V, so as far as the over driven thing I *thought* was taken care of. Maybe my board wasn't somehow...

And do you have suggestions on how to "let heat flow off" the board? I have it hotglued on the non-wired end (no components are covered by the glue) to the inside of an MHS choke piece, so the board itself is suspended in the center of the connecting piece (ribbed section). I thought this would allow air to flow both above and below the board. To keep the board from any possibility of shorting, I made a sleeve out of a pill bottle to slide over the entire board. There's 1/16" of clearance between the pill bottle and the ID of the ribbed section, and about 1/2" of clearance between the pill bottle and the top/bottom of the board. The wired end of the board is left open, for airflow.

What more can I do, drill a few holes in the body of the pill box?

[xwingband]

Quick note... he used NiMH for the 7AAA pack. 7 li-ions isn't going to help you much! I too like 7AAA of NiMH and have used it recently because it's compact and will run a higher voltage LED with regulation much better.

[Goodman]

Thanks xwing, yeah, I meant to acknowledge that. Seven AAA NiMH cells for 8.4V.

My concern, without having the actual cells in front of me, is that seven sets of wiring will hamper the fit into an MHS piece. But if you two gurus say it can be done, then I believe!

Pictures from anyone would certainly help.

*leaves for Best Buy to pick up some rechargeable Energizer E2 NiMHs*

[Leumas]

I've tried it with my setup and the brightness increase is very noticable. I was just wondering though, does anyone know how much of an impact on the LED's life span there is running it at 1200Mah?. I'm running my lux 5 at the lux 3 overdrive settings, and I have 2 trustfire 14500 li-ions powering it. the other thing is the battery life, running it at 850Mah at standard lux 5 settings i'll get about an hour, does anyone know how much runtime I can expect when running it on 1200Mah.

Cheers, Leumas

[Eandori]

Do you have a picture of your pack that you described?
Sure, here's a few pictures of the battery pack I made for the Proto3A and Proto3B sabers. Both of those sabers used CF with a variety of luxeons... but for US 2.0 I would only reccomend this battery setup with a Luxeon V LED.
http://home.comcast.net/~eandori/Proto3a_innards.JPG
http://home.comcast.net/~eandori/Proto3a_Battery.JPG

How did you connect all seven in series? Did you solder directly to the cells to save wiring space, or did you fabricate a holder with seven connections on it?
Yes, I soldered directly to the ends of the cells. Be CAREFUL when you are doing this, don't destroy the heatshrink insulation on the cells when you are doing it, don't heat up the cells too much, and DON'T solder many cells in a loop! Carefully plan out each solder joint before you do it. For me, I glue them all first, then I solder them.

*eep* Also, I must then assume that you're using a charge port for this custom pack...have you seen any issues with the cells during charging? I thought they slightly expanded during charging, which I imagine would be hampered by all the superglue (maybe I'll use hotglue instead, which has some flex to it). Or was that only in li-ions...
From what I have heard, cells only expand when they are overcharged, really hot, and venting chemicals as a result. I don't overcharge. Yes, I use a recharge port, and for my chargers, I use ones that are made to peak-charge NiMH packs at the proper voltage AND they only charge at 500mA to 800mA. These are only AAA cells after all, don't charge them at current above 800mA.

I'm hoping that the board makers properly heatsinked the VR in the first place. Ultra's video showed the board can take 11.1V, so as far as the over driven thing I *thought* was taken care of. Maybe my board wasn't somehow...
No, that's exactly the point I was making. On the US 2.0 board, the voltage regulator is soldered directly to the board, so heat flows into the board. Not a heatsink. That is NOT A SETUP that is designed to withstand a lot of heat. In short, don't try to sink the heat away, spend your time making sure the board does NOT heat up in the first place!!!!!! You do that by having a battery voltage which is very close, but 1-2 volts above that LED voltage. An 11v battery pack, with a red luxeon III LED will make that boar HOT!!! Don't do it!

My concern, without having the actual cells in front of me, is that seven sets of wiring will hamper the fit into an MHS piece. Picture posted above, seven sets of wiring? No no, the wiring is very short from cell to cell. Most of the wiring was only like half an inch long. As you can see in the picture, the pack is VERY compact. One reason I love it so much!!! An hour of runtime and enough voltage for a Luxeon V. Again, that pack is great for all luxeons with CF, but with US 2.0 only use that with a Luxeon V LED.

One final note, after I made the 7AAA pack, I insulated the top/bottom by cutting out small round pieces of cardboard, then I used a hot-glue gun to glue the cardboard pieces to the top/bottom of the pack. I insulated the sides of the pack by wrapping the pack with black electrical tape a few layers thick.


Check your PM's Eandori.

I have sent you an email link for the man who built and programmed the boards.
Gotcha, I'll send him an email today :) Thanks!

[Jedi-Loreen]

If you have a rechargable set up like that, why do you have 2 switches, and where is the 2nd one mounted? :confused:

[Darth Leximus]

He has 2 switches because he's using a CF with that saber. As for where it's mounted I have no idea

[xwingband]

Yup, CF so it needs two switches.

A tip I've had to adopt for making battery packs is epoxy on the ends. They're plenty sturdy when you're making them and all... shipping though, grr. Shipping does a number on them. I had one where the wrapping of the battery was ripped and forced one battery up. Even when not shipping it doesn't hurt to apply a generous layer of epoxy of the wiring. It keeps all the cells together and your wires won't break from movement either!

[Eandori]

Great tip Xwing, I'll need to experiment with Epoxy.

yeah those sabers were made with Crystal Focus boards. One switch was a latching (for blade ignition), the other was a momentary(blaster/menu/lockup). Those sabers are sold now, but each of them had TCSS switches on each side of the saber body.

Side note, Email sent to US 2.0 designer! Hope he responds :)

[Goodman]

Thanks for your comprehensive response, Eandori. 'Any chance you'd just sell me one of those 7 cell packs? I'm short on time, lol..

Seriously though, PM me :)


And from the picture in his sig, I'd say his second switch is mounted opposite the black on the bottom just forward of the grips. ;)

[Jedi-Loreen]

Oh, duh. I wasn't looking that closely at the board. :oops: I obviously wasn't paying attention, since this was in a US 2.0 thread, not realizing he was just showing an example of the battery pack. :rolleyes:









*Slinks back into the background*

[Eandori]

Still waiting for the lead designer of the US 2.0 boards to reply to me. I just need a few simple questions answered then I can provide you guys with some really solid hardware fixes to the board issues.

I'll come back to this thread when I finally get that info.

[Darth Morbius]

Jesse can take some time to reply. I too am awaiting a reply from him about some things as well.

The last time I emailed him it took about a week. He's pretty busy though, so it's not that he's ignoring anyone. ;)

[Eandori]

Still no reply from the designer, so I went ahead and did some testing. Here's what it seems...

The pin 1 of that microcontroller was not programmed to be a reset line. It seems to be used as just a data input line. It's tied high to VDD (around 3.3v) permanently. Here's what I did for my testing.

So I had to de-solder pin 1 so it was floating. Then I soldered a few wires onto VSS (ground), VDD (3.3v positive rail), and pin 1. Since he had pin 1 tied high normally, I tied it to VDD through a 10k resistor. Then after the board was booted and running, I shorted pin 1 to ground. I even verified that pin 1 was indeed going between 3.2v while the board was running to 0v. No effect on the system.

So even though that PIC microcontroller is designed for Pin 1 to be the memory clear reset line, it's NOT currently programmed as the memory clear reset line. which means... the circuit I suggested that's on page 53 of the microcontroller spec won't fix the power-on issue. The boards would need to be re-programmed for that to work. Back to the drawing board for a work-around. But here's some possible thought...

- Perhaps I could hold the clock lines low, so the chip is not trying to execute code until after the power lines have become stable.
- Perhaps I can look up a chip that removes VDD until some time after power has been applied. This is normally done on the reset line, but perhaps some of those chips are robust enough to work on VDD. If it's low enough power usage.
- Following the method of the above solution, a high power Op-Amp might also work... where the " - " pin is tied to battery positive input, and the " + " pin of the Op-Amp is tied to a capacitor-resistor circuit, set up to charge up in around 1 second. The output of the Op-Amp could be the power for the board and the idea would be that until the capacitor charged up the Op-Amp would supply no voltage to the board.

Hopefully the designer will respond soon. This would be a lot easier with just a few bits of data known about the board...

[Do-Clo]

I am also waiting for response from him. I need some info so I can try to interface some parks led strips to them.

[Darth Morbius]

Still no reply from the designer, so I went ahead and did some testing. Here's what it seems...

The pin 1 of that microcontroller was not programmed to be a reset line. It seems to be used as just a data input line. It's tied high to VDD (around 3.3v) permanently. Here's what I did for my testing.

So I had to de-solder pin 1 so it was floating. Then I soldered a few wires onto VSS (ground), VDD (3.3v positive rail), and pin 1. Since he had pin 1 tied high normally, I tied it to VDD through a 10k resistor. Then after the board was booted and running, I shorted pin 1 to ground. I even verified that pin 1 was indeed going between 3.2v while the board was running to 0v. No effect on the system.

So even though that PIC microcontroller is designed for Pin 1 to be the memory clear reset line, it's NOT currently programmed as the memory clear reset line. which means... the circuit I suggested that's on page 53 of the microcontroller spec won't fix the power-on issue. The boards would need to be re-programmed for that to work. Back to the drawing board for a work-around. But here's some possible thought...

- Perhaps I could hold the clock lines low, so the chip is not trying to execute code until after the power lines have become stable.
- Perhaps I can look up a chip that removes VDD until some time after power has been applied. This is normally done on the reset line, but perhaps some of those chips are robust enough to work on VDD. If it's low enough power usage.
- Following the method of the above solution, a high power Op-Amp might also work... where the " - " pin is tied to battery positive input, and the " + " pin of the Op-Amp is tied to a capacitor-resistor circuit, set up to charge up in around 1 second. The output of the Op-Amp could be the power for the board and the idea would be that until the capacitor charged up the Op-Amp would supply no voltage to the board.

Hopefully the designer will respond soon. This would be a lot easier with just a few bits of data known about the board...

Quite the interesting finds Eandori! Looks like I will need to sacrifice one of my boards for a little experimentation of my own.

[Logan Cade]

I’d like to commend Eandori for all the effort he is putting into this. You said early on that you felt it was a programming issue and I too wish they’d contact and work with you to help solve this instead of spending too much time on damage control.

[DizzyKungFu]

I’d like to commend Eandori for all the effort he is putting into this. You said early on that you felt it was a programming issue and I too wish they’d contact and work with you to help solve this instead of spending too much time on damage control.

Amen! Eandori is obviously a resource that should be used to full potential! :)

[Eandori]

Thanks guys, the kind comments are very appreciated.

I'll do my best to not only find fixes for these issues, but to make them as cheap and easy as I can. Then I'll make a detailed post on how to do it, unless Alex, Tim, etc beats me to it :)

Merry Christmas everybody!

[Eandori]

Update, still not a word from the US 2.0 designer. :(

I have waited plenty long enough. Hopefully these issues have been resolved by somebody else. I'll read through the threads and see. If not, I'll go implement some test-fixes WITHOUT any help from the board designer.

However... it would be pretty nice if Tim or Alex could back me up. Meaning if during my attempted work-arounds one of the boards gets damaged they could help me replace components? I don't suspect that will happen, I'm pretty careful about what I'm doing. But Murphy is indeed still alive in well.

[Strydur]

I am just reselling the boards and had/have no contact with the designer.

[Eandori]

What I'm asking though is this...

If I continue helping Ultrasabers resolve their board issues without their help... and as a result accidentally cause some damage to my board, will you help me replace it?

I'm offering free Engineering/Debugging help but with my own personal hardware. Typically, this is the kind of thing a person would be PAID for doing. I'm already doing it for free because I love this hobby and I like helping people out. But the designer won't give me the time of day to answer some simple questions that would allow me to SAFELY test out some alternative methods of bug fixing.

So the result is... I'm going to have to make ASSUMPTIONS about how he designed the board in order to try out a few more tests. I'll be as cautious as I can... but in the case something fries I would hate to lose my own personal boards as a result of this.

Because then I would not only be doing professional debug for free... but I would also be losing out on hundreds of dollars while I replaced my own fried hardware. Not very appealing...

Or... perhaps somebody could kick the designer in the butt and get him to either fix the issues or REPLY to me!

[Goodman]

Hey Eandori,

I have two boards left and I just ordered one more. I may be willing to donate to you for testing. PM me...

'Whatever helps the community!

[Eandori]

Thanks Goodman. I'll go ahead with the next tests then. Since now I have a backup just in case...

The next "trick" I want to try is holding the clock line down on the PIC Microcontroller. If the PIC is not receiving a clock, it will not try to boot. So if I hold clock down until power becomes stable, then let the clock go... hopefully that can solve the problem.

This of course cannot be done with a switch, because a switch on the clock line would also have bounce. No, I need to use a special circuit that will hold the clock line for like 200mS after power is applied. There are chips you can buy that do it... or I could wire up a Schmidt Trigger circuit. Or even perhaps a comparator with one line tied to a voltage divider and the other line tied to an R-C circuit set to charge in 200mS.

Lots of ways to try it... :)

[Logan Cade]

I’ll support you also, Eandori.

I have 3 of these boards that I don’t want to use until they’re right. If you run into problems and need one, let me know.

[Goodman]

Thanks for jumping in too, Logan.

And Morpheus, I literally ordered a USv2 from TCSS six hours ago. I did notice that they were out of stock at Ultrasabers' website, but that's why we all love Tim's store... ;)

[Strydur]

I am pretty sure I am the only one with boards left to sell and the next run will have these issues resolved or I will not be buying them. Overall the boards in their current state work fine for most people. Yes they require certain switchs etc and you have to monkey the battery pack at first but they do work. If I was to offer to replace any you blew up I would just be throwing away the money I paid for the boards. I am thankfull you are willing to give fixing these issues a shot but if you cant get the designer to respond to your E-mail then I doubt he will listen to your ideas on how to fix it either. This is something you need to take up with either the designer, Alex, or Michael (yoda) as they are the only ones "in the loop".

[Novastar]

Edwin... I'm going to have Yoda e-mail you shortly...

According to him, the USv2 problem *IS* being resolved. I believe there will be an upgrade offered... BUT... that is NOT the official word yet. At this point, I'm only trying to dissuade anyone from spinning their wheels on a solution that either will not be fully appreciated... or will simply cause problems with their own board.

The suggested upgrade is a software/firmware level upg., so... many of the hardware workaround headaches should (hopefully) not be needed.

Again... let's give Alex a moment to be able to announce this... but I'm saying all of this to help others not waste any time over it on their own!!

[morpheus1977]

Thanks for jumping in too, Logan.

And Morpheus, I literally ordered a USv2 from TCSS six hours ago. I did notice that they were out of stock at Ultrasabers' website, but that's why we all love Tim's store... ;)

I know I just realized I was having a major "duh" moment.

[Eandori]

Software update eh? I'll bet dollars to donuts that he's put more delay in the polling-rate which he reads the momentary pin to fix the bouncy switch issue.

For the power on issue... I'm curious to hear how that is being solved in software. Typical embedded systems design is to make a circuit that keeps the reset line held on your microcontroller until power rails become stable. In short, special circuits/chips are needed to ensure your system does not try to boot when power input is jumping around. Usually, issues like that need a hardware fix.

Great to hear theres fixes coming though!