Sold/Expired *Closed* Programmable Incandescent Regulator 1 sales thread

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

5V/470 = 10mA.. so with two parallel (5mA) might not be bright enough.. I actually personally would de-solder since with the 'talon' desoldering tool it takes less than 0.5 seconds to remove a 0603 device.. i only suggested tacking on 30ga wire as a thought because de-soldering a plastic SMD is normally a real piece of work.. it's nearly impossible to de-solder both sides to remove it. The MX-500 pays for itself daily! I used it about 20x yesterday to remove an 0603 capacitor to switch to a different value.

What i've been proposing is a mix of almost exactly those two concepts.. put the fiberglass 'wool' below a metal shield (since i'd be using with bi-bin bulbs).. I tried a test but made the bad mistake of trying to recycle some insulation that was previously installed in a house.. and it had DUST in it which instantaeously started to smoke! I plan to fabricate something out of tin or copper to make a conical heat-shield and put the fiberglass below that around the KIU. There can be no question that will help things.. normally i measure temps nearly 100C on the top of the KIU socket shortly after running the 62138 or 64625 lamp.

-awr

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

I don't find it that hard to desolder SMD LEDs, but they do tend to have a rater large surface area that causes a lot of capillary action, but it's fully doable.

Good news everyone!

Me and Alexander played with the regulator tonight and tweaked some settings. We tested using a battery pack instead of a PSU in hope that it would cause the losses in the capacitor to go down due to more stable voltage. It did not! In fact, the losses increased a lot. Powering a 12 V, 50 W lamp heated up the capacitor enough to fry eggs on it. This required a solution and we found two.
1. Adding a diode. Worked like a charm but took a lot of time to resolder and we didn't have 77 diodes on the shelf either.
2. Removing the capacitor. Also worked like a charm!

We laughed out loud when we found this trick and started to desolder the capacitors. It was China in practice, build it and start removing all capacitors and replacing all coils with short-circuits until it stops working. I can tell you that we won't learn that in school, even though we both study to become EEs!

Another funny thing was to run the same 12 V, 50 W lamp at 18 V. The losses in the regulator was about 100 mW because it had stopped switching (100 % duty cycle) and 100 W in the lamp. That's 99.9 % efficiency!

The software is almost done as well. Just some final beta testing and tweaking and we should be ready.

My plan is to solder, program and test the first 40 regulators tomorrow, possibly more if everything goes well.

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

I did the same thing (remove capacitor).. the specs for my driver chip want an input and output capacitor.. since 1) i'm using a faaaaar bigger capacitor called a 'battery' to drive.. no need for input cap.. (also no wires to act like radio antennas either).. and 2) output load does not change 'til the bulb blows.. no real need for output cap either.

This led to a couple issues during building.. sometimes the driver would not start with the power supply since there is a 38kHz ripple on it grrr.. note to self buy a better bench supply when you can.

I made 'tester bob'..

specifically because w/o the output cap and running from the bench supply, sometimes (very frustrating when not always, huh?).. it just wouldn't start.

great news

-awr

 

[wquiles]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Nice test "rig"

The extra two legs to the sides are for your voltmeter I presume?

Will

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

tester bob gets plugged into the lamp socket (his legs).. it's actualy his 'arms' that get clipped with the test leads from the DMM.. (somebody called them whiskers, that works too). The arms are horizontal so my miniclips can swivel and i can put bob down into a light or when i have to flip him around to orient + properly (cap is polarized).. it's really easy.

-awr

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Our plan to be ready tonight almost succeed. There was a bug in the calibration system that should have allowed us to calibrate every regulator prior to shipping so that you could set all voltages in volts, rather than sweets. It should be ready tomorrow but I don't think we will make it before the postal office closes.

Anyway, the hardware production is about 98 % done now, leaving only the 45 V regulators which I will do tomorrow.

The programming is about 99 % complete and the first PIRGUI interface is almost ready too, it only needs a touch of user-friendliness. You are only able to set all values with it and reflash it, and only using a parallel port, but all remaining features will soon be there.

We noticed one little bad thing also. I said somewhere that the standby-load was in the 1 uA region. It is not. Dividing a voltage using resistors in the 100 k-range and not reading a datasheet properly caused the standby-load to be about 205 uA at 10 V, with proportional in/decrease with the supply voltage. This is not a lot, but more than the self-discharge of your batteries.
As always, if you find this unacceptable, we will gladly repay your money, no questions asked.

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

with the size of the batteries being used to light a light using the PIR.. 200uA is absoltuely negligiable.. standby current on my single AAA cell nano when using the variable resistor is 80uA.. the Minimight which will effectively always be 'on' but in 'standby' mode and using only a single 1800 mAH cell.. will pull about 200uA at standby i'm sure.

Divide a typical battery pack for a mag85.. 1.4AH by 205uA.. you get 6800 hours.. 284 days.. that is much slower than self discharge in practice.. although you won't drain a battery 100% dead in 3/4 of a year it won't be useful enough to light a light.

here's where i get to say that hotdriver takes virtually zero uA at shut off since i actually disconnect the power to the chip and the entire circuit.. in shutdown it does problably take a couple mA (after low voltage kicks in).. well in the case of the models with the high-temp circuit which has to run the comparator..the driver chip takes like 10uA or something negligable. Talk about 'cheating' though, huh... unplugging power makes it take less power.

as long as it's in the uA range and not mA range, nothing whatsoever to be concerned about.. good to know though.. if you use it in a light that is stored long periods of time.. or in a light that uses primary cells.. might want to losen the tailcap when putting into storage.

-awr

 

[Sway]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Andrew I think you answered my question before I could research and ask

See I don't have a good grasp on uA and mA but I live with several nFlex lights that draw 0.2mA in the sleep mode and get along with them just fine, just to confirm the PIR is in the same ballpark or less, if so I'm good to go :candle:

Later
Kelly

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

0.2ma = 200uA so 205uA = 0.205mA.

a micro amp is 1/1,000,000th of an amp.. a miliamp is 1000 microamps 1/1000th of an amp.

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Divide a typical battery pack for a mag85.. 1.4AH by 205uA.. you get 6800 hours.. 284 days.. that is much slower than self discharge in practice.. although you won't drain a battery 100% dead in 3/4 of a year it won't be useful enough to light a light.

Really? I would way it depends on how you calculate. I was thinking that a 2.5 Ah NiMH battery pack would last 12000 hours, which is.... No, you are right. I did it wrong.

I'm off to the store now for getting pre-bubble waped envelopes so I can pack the regulators tonight.

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Mini update:

10 regulators programmed and tested, and still counting!

 

[wquiles]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Good progress dude

Will

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

We managed to program and test 20 regulators before the last bus left. We might be able to do some more tomorrow, depending how long school takes but the first 20 little gems will be shipped tomorrow. You can see in the list on the front page who are getting theirs fist. No hard feelings...

I really hope everything will work out all right this time. People might have some difficulties with our programming interface but I'll try to write a howto this week. The other problem that I can think of now is if someone set the cut-off voltage to 8.0 V. For some reason, the error margin for exactly 8 volt was +- 0.5 volt but about +- 0.1 or less for everything but 8.0 volt. Just set it for 7.9 or 8.1 V and you should be all right.

The default setting is as follows:
Mode 1: 5 V
Mode 2: 9 V
Mode 3: 12 V
Softstart time: 400 ms
Warning level (blink): 0.0 V
Cutoff level: 0.0 V
Shutdown temperature: 80 degrees Celcius

Please, before I ship everything, can you please check the list on the front page that you are getting the right setup and that you haven't moved/changed address or anything since you did your payment. Thanks!

We squeezed in one more menu as well... Here is the new options:

- Main menu -
1. Voltage 1 -> xx.y [V] x=[0 60] y=[0 9] 0.0 V => Mode disabled
2. Voltage 2 -> xx.y [V] x=[0 60] y=[0 9] 0.0 V => Mode disabled
3. Voltage 3 -> xx.y [V] x=[0 60] y=[0 9] 0.0 V => Mode disabled
4. Softstart time -> xxy0 [ms] x=[0 30] y=[0 9]
5. Cutoff voltage -> xx.y [V] x=[0 60] y=[0 9] 0.0 V => Cutoff disabled
6. Warning voltage -> xx.y [V] x=[0 60] y=[0 9] 0.0 V => Warning disabled
7. Shutdown temperature -> xxy [*C] x=[0 13] y=[0 9] 0.0 *C => Ignore temperature
8. Calibration -> Enter sub menu (yellow LED is lit)
9. Kitchen default -> x x=[0 9] 5 => confirm

- Calibration sub menu -
1. Calibrate drive - positive -> xx.y [%] x=[0 50] y=[0 9]
2. Calibrate drive - negative -> xx.y [%] x=[0 50] y=[0 9]
3. Calibrate cutoff - positive -> xx.y [%] x=[0 50] y=[0 9]
4. Calibrate cutoff - negative -> xx.y [%] x=[0 50] y=[0 9]

 

[Kevin Tan]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Please set mine to

1) 1185 on 3x3 17500

2) 1166 on 4s x 2p 14500 in Mag2D

Both to be installed in kiu socket.

Declare the package gift at US$50.00

TQ

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

for 3x3 17500 you want 11.1V operating.. 10.35 blink and about 9.5 shutdown.
for 4x2 1166 you want 12.5V operating.. 13.8 blink and 12.5 shutdown.

 

[Kevin Tan]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Thanks for the setup awr!!

It would be nice to have all the data for the regular WA lamp types like the 74, 11 ,60 and others. ( Hint, hint )

 

[aosty]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

for 3x3 17500 you want 11.1V operating.. 10.35 blink and about 9.5 shutdown.
for 4x2 1166 you want 12.5V operating.. 13.8 blink and 12.5 shutdown.

Andrew - you might want to double check your figures for the 1166?

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

Kevin Tan,


Sorry, I don't have time to program the regulators with everyones personal settings. It would be hard to keep tack of them then.

 

[andrewwynn]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

well at least it can help Kevin to program.. you'll just have to follow the directions, didn't seem to hard.. i didn't try it yet.

aosty.. the numbers are correct for the 1166.. interesting that shutdown and operation are the same voltage, it's just a coincidence.

I run the 1166 at 12.5V.. 913/594L.. from a 14.4V source.. means that with the hotdriver it will shut off at 11.5V but that is 2.785V/cell and i prefer to shut off LiON at about 3.1V/Cell which is 12.4V.. but no point to go below operational voltage if you can change the shutdown voltage... then you get 100% regulation.

the 'blink' is at 3.45V/cell.. the point where LiON become useless as alkaline... then you know you only have less than 1 minute left... you could change that to 3.5V/cell and get a 'two minute warning'.

-awr

 

[winny]

Re: Programmable Incandescent Regulator 1 sign up-thread *accepting payments*

All the 20 first was shipped earlier today. The transit time should be 4-6 days to the US.

We programmed and tested 12 more today and we should manage to do some more tonight. We are currently having some problems with our programming adapter and there have been about 10 regulators with soldering faults where the MOSFET driver lives its own life.

The missing components should arrive tomorrow according to our supplier but I personally take that with a pinch of salt. We'll see...

Time is not something that I have a lot of right now but I hope to write a manual before anyone receives their regulator so they'll know how to connect and operate it.