SALE: JM-PhD-D1 PWM Hotwire Regulator

[LuxLuthor]

Jimmy, how many of these have you sold so far? I am stunned that many hundreds have not been gobbled up, given that these (& AlanB's) essentially let you walk on water, they are that miraculous.

[JimmyM]

Jimmy, how many of these have you sold so far? I am stunned that many hundreds have not been gobbled up, given that these (& AlanB's) essentially let you walk on water, they are that miraculous.

Thanks for the kind words.
I've sold 68 but have less than 32 left. I have 3 to repair, sold one with a light, minus the ones I sent out for the 3+1 sale, and have one in my own 3xD-LiIon Mag5761.
Last week's sale was very successful. Perhaps the next batch will cost less to begin with. I just ordered a bunch of stuff from Digikey to start work on the X1 high power regulator. With proper heat sinking it should be good for ~80A (Dual 75V FETs). I'm still working on and off with the LCD Key idea. It's the smallest easily implemented graphic LCD I can find.

[secondtimearound]

Jimmy,

Received regulators. Thank you. Great installation instructions too. However the lamp insta-flashed upon start up. I measured full pack voltage at the blown lamp pins 10.5V

WA1274 lamp over 8 AA . You set the reg. to 8.3V lamp and 7.2V low

If I understand this correctly the voltage must be set using a load or resistor , so I couldn't possibly adjust the vlamp using the blown lamp. Is this correct?

[JimmyM]

Yes, you need to measure voltage with some sort of load.
Sorry about the instaflash.

Jimmy,

Received regulators. Thank you. Great installation instructions too. However the lamp insta-flashed upon start up. I measured full pack voltage at the blown lamp pins 10.5V

WA1274 lamp over 8 AA . You set the reg. to 8.3V lamp and 7.2V low

If I understand this correctly the voltage must be set using a load or resistor , so I couldn't possibly adjust the vlamp using the blown lamp. Is this correct?

[petrev]

Jimmy,

Received regulators. Thank you. Great installation instructions too. However the lamp insta-flashed upon start up. I measured full pack voltage at the blown lamp pins 10.5V

WA1274 lamp over 8 AA . You set the reg. to 8.3V lamp and 7.2V low

If I understand this correctly the voltage must be set using a load or resistor , so I couldn't possibly adjust the vlamp using the blown lamp. Is this correct?

Hi

I use an Osram 64458 to set my voltages but any bulb rated a bit larger than the max pack voltage can be used.

Also I would say 8.3V is way too hot - Lux shows he is going to test life at 7.9-8.0V and personally I would not expect to use anything better than 15hr point for real world use in a short rated life bulb (WA1274=40hr)

When Lux was testing lamps the voltage was ramped very very slowly and there was no major vibration (or shocks) as can happen to a bulb in-hand.

Regulated fun 4 all
Cheers
Pete

[secondtimearound]

Yes it is on the extreme edge, however I have been running even 8.4V using AWR's hotdrivers with success. Albeit bulb life is short but FET takes less of a hit. I dont recall what his softstart values were though.

[nitere]

Jimmy,

Received my regulators, my first impression is that the workmanship is outstanding. I just got a chance to do a bench test on each one to exercise the functions as I will not have the time to install them for a number of weeks. They work great. You really do not appreciate each of the attributes that the software gives you the ability to control until you whiteness it for yourself. FYI I used a variable power supply, DVM, scope and a couple of automotive bulbs wired in series (for the regulators set greater than 12v) for the load. Now I can not wait to upload new software to experiment with a couple of the parameters.

With respect to what Lux said about the sales, I too can not believe this is still your first lot. When people realize the possibilities these regulators give you for a hotwire (Battery, Bulb, Function) they will wish they had them. It is a great price for the hardware and the code is open source to boot.

Jimmy to you and the other contributors (AlanB, wquiles, petrev, LuxLuthor and others) Great Work, Thank You

[petrev]

Yes it is on the extreme edge, however I have been running even 8.4V using AWR's hotdrivers with success. Albeit bulb life is short but FET takes less of a hit. I dont recall what his softstart values were though.

Hi STA



Go as hot as you like . . .
PWM doesn't cause the same problems with the FET that AWR's LDO hotdriver did.

Hi nitere

No work here for me - all play ! ! ! Jimmy does all the hard stuff (and the others of course)

Cheers
Pete

[JimmyM]

Hi STA



Go as hot as you like . . .
PWM doesn't cause the same problems with the FET that AWR's LDO hotdriver did.
Cheers
Pete

This regulator works completely differently than AWR's. FET heat isn't a problem until you're up in the 13A RMS area.

I am going to be looking into a potential problem regarding Vbat/Vbulb difference. The regulator may go to 100% if the Vbatt is much higher than the Vbulb. In most cases it isn't a problem, but I need to do some experimentation to work out the details of this potential issue..

[Juggernaut]

This regulator works completely differently than AWR's. FET heat isn't a problem until you're up in the 13A RMS area.

I am going to be looking into a potential problem regarding Vbat/Vbulb difference. The regulator may go to 100% if the Vbatt is much higher than the Vbulb. In most cases it isn't a problem, but I need to do some experimentation to work out the details of this potential issue..

Should I worry about this? Since my battery voltage is 19.2, but my bulbV is only 15.0? I just got around to soldering up the board and I was going to test it, I’ll wait for your response.

[LED_astray]

Should I worry about this? Since my battery voltage is 19.2, but my bulbV is only 15.0? I just got around to soldering up the board and I was going to test it, I’ll wait for your response.

That should be fine. You can dig thru the design & discussion threads for details. But my recollection is that Alan B said the regulator would be most efficient when Vbatt was up to 1.3X Vbulb. (Right where you are.) I'm sure they've tested much higher drops.

[JimmyM]

Should I worry about this? Since my battery voltage is 19.2, but my bulbV is only 15.0? I just got around to soldering up the board and I was going to test it, I’ll wait for your response.

No worries there. The problem cropped up when starlight was driving a 7.2V bulb with a 6S Li-Ion pack. I'm going to dig into this and figure it out. I just got back from a week of travel for work and I have to travel this weekend as well, but I'll figure it out and issue a software update for those of you running a large difference between Vbat and Vbulb.

[Juggernaut]

Ok, I know I’m a total newbie when it comes to this electronic stuff, but I wired everything together today and every thing seemed to work great except when I hooked up my volt meter on the second run. After about two minuets the voltage coming out of the driver was 12.8 “not near 15” as time went by it very slowly increased, about 5 seconds before the bulb went into low voltage mode “pulsing” it was reading 13.5 volts, still no were near 15. So do you have any idea why?

[LED_astray]

I think you're seeing the effects of a meter that can't measure Vrms with a DC-offset accurately. (You need what Fluke calls AC+DC RMS measurement mode.) Or, Jimmy explains it in his opening post:

...
How it works: It uses PWM (Pulse Width Modulation) to control the RMS voltage to the bulb. By turning a switch (an FET) on and off several times (244Hz) a second, the "effective" voltage that the bulb sees can be controlled. The bulb is still receiving the full battery pack voltage while the switch is on, but it’s not on for the whole time. The frequency of the PWM and thermal mass of the filament prevent flicker. It can only control output voltage less than or equal to the pack voltage. It cannot boost battery voltage....

You probably seeing the DC average voltage, which goes up as your pack discharges and the pulse-width increases to compensate. If you use his formula and take two measurements you should be able to check:

...
Setting the voltage: Any voltage setting must be done with a bulb or other load in place in the socket. A 5W 1K Ohm resistor will work in place of a bulb. A True RMS reading DMM is best (Like the Fluke 189), but it can be done with an average reading DMM using the following calculation.
You need to now the following:
Vin = Battery voltage under load
Vrms = Your desired RMS voltage (The effective voltage to the bulb)

Vavg=(Vrms^2)/Vin

Once you’ve calculated Vavg, turn the voltage adjust pot until your meter reads the Vavg voltage you just calculated....

Hope that makes sense & helps. (I'm sure Jimmy or Alan will be along to explain better.... :-)

[LED_astray]

Found a section of the Alan B's discussion thread that gives more details:

http://www.candlepowerforums.com/vb/...10#post2959810

JimmyM, do you have a discussion thread so this doesn't clutter up your sales thread? (Sorry, I didn't see it, but haven't successfully kept track of all the threads.)

[Juggernaut]

I understand what you are saying, that was one of the things I figured my crap-tacular volt meter might not understand, the only thing I don’t get is why this “average” voltage changes “continues to grow very slowly. Also, I was getting extremely short run times, so I hooked up 2x 12 volt SLAs, and nothing happed? I thought we could hook up anything we basically wanted for voltage. Did you by any chance put a maximum voltage on this like around 20?

[LED_astray]

(I really hope Alan B or JimmyM arrive and save you from me! I'm thinking out loud as I try to fully understand this and hoping it will help you out. :-)

Your meter measures Vaverage. (At 50% duty-cycle that would be (Vbatt + 0) / 2, or Vbatt/2.) The regulator varies the duty-cycle as Vbatt drops to hold Vrms constant. (The bulb cares about Vrms.)

... 12.8 “not near 15” as time went by it very slowly increased, about 5 seconds before the bulb went into low voltage mode “pulsing” it was reading 13.5 volts, still no were near 15.....

I'll assume you have set Vrms to 15.0, your desired voltage, and you measure Vavg= 12.8 and 13.5.

Take JimmyM's equation: Vavg=(Vrms^2)/Vin,

rearrange it: Vin (= Vbatt) = (Vrms^2)/Vavg,

solve it for your values, I get:

Vavg = 12.8 => Vbatt = (15^2)/12.8 = 17.6V
Vavg = 13.5 => Vbatt = (15^2)/13.5 = 16.7V

If that is reasonable for your battery pack, I think you are seeing what is expected.

[LED_astray]

Thought of another way to look at it that might help fill in the picture. For this case:

Vavg = Vbatt * duty-cycle

rearrange & solve for duty-cycle with your Vavg & my predicted Vbatts:

duty-cycle = Vavg / Vbatt

12.8/17.6 = 0.73 = 73%
13.5/16.7 = 0.81 = 81%

the regulator increases the duty-cycle as the battery fades and that affects your meter reading. That's what the regulator is supposed to do, it's just faking your meter out.

[Juggernaut]

Thank you very much, your predictions are exactly what I’m seeing as my run time progresses. The only thing I don’t understand is why 24 volts won’t work?

[Juggernaut]

Thank you very much, your predictions are exactly what I’m seeing as my run time progresses. The only thing I don’t understand is why 24 volts won’t work?

Never mind that, I just went back to my room to try and figure out if I had done something wrong, and now it works? I didn’t change anything, well what ever I’ll probably switch to 2x small 12 volt SLAs. Granted my “low V” is to low now, worst case I could always drive over to JimmyM’s house to get that readjusted.

[JimmyM]

Never mind that, I just went back to my room to try and figure out if I had done something wrong, and now it works? I didn’t change anything, well what ever I’ll probably switch to 2x small 12 volt SLAs. Granted my “low V” is to low now, worst case I could always drive over to JimmyM’s house to get that readjusted.

I just got back from a trip visiting family/friends, playing golf. OK. I'm glad LED_astray's predictions match your observations.
It should work for any voltage from your Vbulb up to 40V. If it didn't work before, and it does now, without changing anything, I'd suspect a bad connection somewhere.

And... Yes, you can come by and I'll check everything out and reset things of you like. Hold off on that for a bit though. I have some repairs and software troubleshooting to get through first.

[JimmyM]

A discussion link is HERE.

[Bubblemaker]

Hi, I searched the net for a regulator like this and have a few questions. Since I'm 'technically challenged' I have great difficulty to understand the exact specifications of the regulator.

I have a xenophot 6V 35W divelight which runs overdriven on a 7,2V 12.000 mAh batterypack and uses a reed switch.
Can the hotwire regulater be programmed to achieve a dimmable light with a light output of 100% (7,2V overdriven), about 66% and about 33%?

If so, I'm very intersted in that configuration and would like to buy one.

Looking forward to your reply.
Thanks a lot.

Marcel

[Starlight]

Hi, I searched the net for a regulator like this and have a few questions. Since I'm 'technically challenged' I have great difficulty to understand the exact specifications of the regulator.

I have a xenophot 6V 35W divelight which runs overdriven on a 7,2V 12.000 mAh batterypack and uses a reed switch.
Can the hotwire regulater be programmed to achieve a dimmable light with a light output of 100% (7,2V overdriven), about 66% and about 33%?

If so, I'm very intersted in that configuration and would like to buy one.

Looking forward to your reply.
Thanks a lot.

Marcel

Yes, his regulator will allow you to set 7.2v, or any voltage up to 40v as your maximum. Right now his software will allow you to set up to 4 different voltages for dimming and switch between them and maximum voltage. The only thing you need to check is that the dimming features require that you have a "momentary" switch, rather than a click on/off switch.

[JimmyM]

Hi, I searched the net for a regulator like this and have a few questions. Since I'm 'technically challenged' I have great difficulty to understand the exact specifications of the regulator.

I have a xenophot 6V 35W divelight which runs overdriven on a 7,2V 12.000 mAh batterypack and uses a reed switch.
Can the hotwire regulater be programmed to achieve a dimmable light with a light output of 100% (7,2V overdriven), about 66% and about 33%?

If so, I'm very intersted in that configuration and would like to buy one.

Looking forward to your reply.
Thanks a lot.

Marcel

5 levels actually. Your xenophot bulb will be no problem. You need the above mentioned push-button to control the multiple levels. Go to the first post in this thread and read through the whole thing. Also, there are installation and wiring diagrams availeable ther for download.

[Bubblemaker]

Thanks for your quick reply. The divelight uses a reed switch to prevent the need to drill the body. Too bad the regulator isn't compatible with a reed switch so I'll have to look further.

Thank you and good luck with your business.
Marcel

[JimmyM]

Thanks for your quick reply. The divelight uses a reed switch to prevent the need to drill the body. Too bad the regulator isn't compatible with a reed switch so I'll have to look further.

Thank you and good luck with your business.
Marcel

No one said it's incompatible with a reed switch.
In standard mode, you can use the reed switch to control power to the regulator. Since you're only controlling the internal electronics of the regulator, the reed switch only has to pass ~ 5mA of current. That gets you a single level output. If you can use the reed relay in such a way as to only activate it momentarily, you can use multi-level mode for upto 5 levels of output.

[Bubblemaker]

I can activate the reed and almost immediately shut it if that is what you mean. I can't find the actual required (minimum and maximum) timing for this operation. Can you mention that?

Thanks

[JimmyM]

Less than 1 second. Longer will turn the light off. So, if you can just activate/deactivate quickly, that should work.

[JimmyM]

Updated software.
I made some changes to the software to eliminate the "jump to 100%" problem. This only occurred with the multi-level interface when I had set the bulb voltage to 7.2 and the Vbat exceeded 26V. I've tested this software from 7.2 to 40 volts going through each of the levels along the way. I now cannot reproduce Harry's issue.
The updated firmware can be found HERE