JimmyM's Regulator Earns It's Doctoral PhD Degree, Despite The Heat !

LuxLuthor · Jun 2, 2009

Jimmy,

I got everything all hooked up and triple checked with the two 64458's. I took several shots of my setup so you can see where I put everything, including temp probe with some thermal compound on the FET and angled it so when it does finally heat up and melt the solder it will fall off the board!

The last picture was after I had the same voltage settings and readings, but moved the voltage clips to the two poles of the shunt...in an attempt to figure out the amps by the voltage drop...but as it is 5 AM, I realize that I can barely remember my dog's name, so I'm not at all sure if this is right. I lightened the images so you can read the Flukes.

I'm puzzled by the Amps showing on the PS, since they were zero when we talked last night in setup with 5W 1k Ohm resistor in single bipin. I also took the Judco out of the equation, and just connected power control to first pole of shunt, so regulator is always on unless I remove wire.

So before I crank this up and potentially destroy your baby, (to find its limits) I thought I would see if I have this setup right, and try to find out how to measure the amps since that's what you really want to know. Anyone else can comment, I'll check these later today...again, this isn't my real review yet, until I get confirmation from Jimmy.

These are just 3 different views in case you need to see different angles. The black aligator going to Fluke 189 is clipped on soldered wires from KIU joining wire from PhD (at 3 O'clock position on your diagram)

I dialed up the PhD to output about 15V. Obviously it needs to keep going, but not sure of how to measure voltage and current to reach ideal target.

This is where I moved the voltage clips to the shunt, but the voltage reading of 0.0083 correlating to current didn't make sense to me. That is a 100A 100mV (A series) shunt I got from RC-Electronics here.

petrev · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

Gosh - You are having fun. Lux - AKA The 2-Fluke Kid.

100A - 100mV Shunt ? From what I know (?) that means it should read 1mV/A so your reading is 8.3A - could try the mA scale to get more acuracy ? ? ? ?

BUT . . .

Sure Jimmy will chime in with some actual real helpful knowledge but in the meantime looks good to this layperson. Lokking forward to the full review

Cheers
Pete

Al Combs · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

So the shunt is saying 8.3 amps while the PS gives a reading of 4.7 amps in the same circuit? Is the shunt exactly 1 mOhm? Or do they include a certification sheet telling you its actual resistance? It's not 1.766 mOhm, is it? That would be a pain in the butt if you couldn't use it without a calculator. At ≈ 7.5 volts per bulb, the 4.7 amp number does sound a little more like it.

It's not 5 in the morning for me but it sounds like maybe there is something wrong with the shunt. Have you had a chance to test it with something like a power resistor in series? A device that would give predictable results using Ohms law. Or is this the first chance you've had to use it?

Alan B · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

Do remember here folks that the current is also PWM and requires True RMS to read correctly.

niner · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

Alan B said:
Do remember here folks that the current is also PWM and requires True RMS to read correctly.

I agree Alan. I pretty sure the amp meter on the power supply is not Ture RMS.

I use a clamp meter to measure the RMS current, and sent the output to the scope for verification.

JimmyM · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

If you have your Fluke 189 set to AC+DC Volts, as it appears to be. That's correct. It looks like 8.3A. But a pair of 64458s in parallel on 15V should be pulling closer to 17A per your destructive testing results.

LuxLuthor · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

LOL! See, now this is why I knew not to trust my 4-5am work before going full out. I just talked to Jimmy who said that Al Combs pointed out that my bulbs were in series. Doh!

Now that he pointed that out, it's obvious that I screwed the pooch and put them in series by mistake. So the Amps on the shunt do make sense.

By the way, while waiting for Jimmy's call, I proudly marched the regulator voltage up to almost 32V, with the power supply pegged out at about 32.6V. Temp on FET never exceeded 54.6°C. RMS on 189 clipped on shunt showed maximum amps to be 10.8A.

So now we know three things.

1) Assume that working in the wee hours will result in missing something completely obvious.

2) At 10.8A, the FET is only 54.6°C

3) The measurements I was doing were perfectly logical, given my stupidity.

OK, now to hook up the bulbs in parallel. Let's see if we can destroy something. I also wonder at what point the PS will blow its fuse (but I did plan ahead and get extras.) Yikes, I cannot find the tinfoil icon. I'm doomed.

JimmyM · Jun 2, 2009

Re: Review JimmyM PhD (early stage)

LuxLuthor said:
I'm doomed.

We're all doomed. Since the moment we set virtual feet in here many many moons ago.
Things should get interesting at about 15A.

LuxLuthor · Jun 3, 2009

Re: Review JimmyM's PhD Regulator - The "Doctor" Survives !

Setup Display:

It makes a difference in your results when you hook up the two bulbs in parallel! OK, here is the real setup and results on the Power Supply. I have not yet tested this with a battery pack in a light, but I don't believe it is likely to get to the extremes unless you are overdriving a 36V bulb to some insane level.

Testing Results Summary and Discussion:

Everything worked out much better that Jimmy anticipated. I could not break his PhD circuit, and I tried my best torture techniques, although this did seem a candidate for water boarding, so I had to stick with more conservative techniques.

My PS outputs 30V; 20A, although with the PWM of this PhD board, the PS did not have to output that level of amps, so there was not a concern of overloading it.

As anticipated, the limiting factor on how much current can be cranked through this tiny PhD is determined by the heat buildup on the FET, which in turn radiates to the controller. The controller is programmed by Jimmy to allow a maximum of 70°C (158°F), then it ramps down to about 25% PWM output, cooling it off. To reset this heat protection feature, the power must be turned off. It could be bumped up a bit higher, but I don't see the point.

I found the highest current level I could deliver in this setup with two 64458 bulbs in parallel with the Voltage POT adjusted to maximum was 22.3 Amps. This correlated with a voltage measured in circuit of 13.7 Vrms. During this rampup, the temp of probe touching the (with some thermal paste) quickly reached a PhD FETmaximum of 87.1°C (188.8°F) While this is an extreme temp, it was not high enough to melt solder, and no heat related damage was visible on close inspection of PhD afterwards.

I started out taking a series of readings initially putting the power supply voltage on maximum output and turning the Voltage Adjustment POT clockwise while monitoring the shunt voltage to see how high the amps could go. I would occasionally switch the leads over to measure the actual circuit voltage (instead of amps), but Jimmy was most interested in the heat effects of Amps on the FET, as the limiting component.

Once I saw the high heat activate the heat protection feature, I let it cool down, and changed the Fluke 189 to the mV scale while measuring the shunt voltage (AC+DC RMS Mode). I repeated my high Amp measurements at 18.15 A, but again saw it quickly raise the FET temp to 88°C, followed soon by heat safety reduction.

The final set of measurements were to try and determine how many amps would heat up the FET to just under 70°C, thereby avoiding the heat reduction trigger. The answer was approximately 15.44 Amps were able to be run without FET temp hitting 70°C.

However, one must assume there will be radiated heat to the PCB from high overdrive bulbs, so the heat limit from bulb radiation may be more important to manage with the PhD than excessive current.

Testing Data:

Rather than displaying all the images, I'll just put a brief description of what the image shows, and you can click on link it if you want to see the result. There are 20 measurement images here, ranging in file size from 400-700KB

JimmyM · Jun 3, 2009

Lux, this is great information. Thanks for all your time that you put into this.

Some may note that the voltages being used here are all quite low. But you have to remember, it's the Amps that are the cause of the limitations due to heat. There is no difference between a 12V bulb drawing 10 amps and a 24V bulb drawing 10A. The amount of heat generated will be the same.

Alan B · Jun 3, 2009

Fantastic test results!

Jimmy, what is the on-resistance rating of that FET?

JimmyM · Jun 3, 2009

Alan B said:
Fantastic test results!

Jimmy, what is the on-resistance rating of that FET?

3.8 mOhm @ 5V gate drive. a 46nC gate chrage helps it switch quickly given the limited capability of the AVR. I also have a 150 Ohm resistor in series with the gate to keep current spikes at the switching events to a minimum. Had had good results during testing. I drive the hell out of a 64623 (15-16V) using a 220 Ohm resistor and had no heat issues. So I figured a 150 would help it switch more quickly and still not load the AVR output.

Alan B · Jun 3, 2009

Great choices. Pretty similar to the choices I have in the DropIn as well. I have not tested that hard, and the fuse will go somewhere above 15 amps, but it does look as though the fuse will go first, which is excellent.

Well Done!!! No :poof:

HarryN · Jun 3, 2009

That is quite an impressive setup - almost enough to make me go hotwire. :laughing:

Lux - please consider to somehow shield your eyes from those bulbs during testing. I have done enough table top testing with MUCH lower brightness sources to be made well aware of the possible dangers to vision.

I am pretty sure that the US has not yet adopted eye safety rules for incoherent light sources (like bulbs) but the EU has, and I can conservatively estimate that you are way past them. :poke:

JimmyM · Jun 3, 2009

HarryN said:
That is quite an impressive setup - almost enough to make me go hotwire. :laughing:

Lux - please consider to somehow shield your eyes from those bulbs during testing. I have done enough table top testing with MUCH lower brightness sources to be made well aware of the possible dangers to vision.

I am pretty sure that the US has not yet adopted eye safety rules for incoherent light sources (like bulbs) but the EU has, and I can conservatively estimate that you are way past them.

I've actually implemented a form of safety for my benchtop testing rig. Not for lumen level protection per se, but just becasue I can't see anything else while I've got a bulb cranked up. Not to mention a 24V (or was it 36V) bulb I tested for another CPFer. Huge thing. The filiment blew and fell into the bottom of the capsule where it began to arc. That softened the envelope and then it vented it guts in a jet of molten tungsten onto my days old clean white bench top.

LuxLuthor · Jun 3, 2009

HarryN said:
That is quite an impressive setup - almost enough to make me go hotwire. :laughing:

Lux - please consider to somehow shield your eyes from those bulbs during testing. I have done enough table top testing with MUCH lower brightness sources to be made well aware of the possible dangers to vision.

I am pretty sure that the US has not yet adopted eye safety rules for incoherent light sources (like bulbs) but the EU has, and I can conservatively estimate that you are way past them.

LOL....Now you tell me. Actually I tried to wear my Maui Jims during much of the testing, but trying to fit the tiny screwdriver on that tiny POT screw required a very steady hand, and no eye shades.

JimmyM · Jun 4, 2009

LuxLuthor said:
LOL....Now you tell me. Actually I tried to wear my Maui Jims during much of the testing, but trying to fit the tiny screwdriver on that tiny POT screw required a very steady hand, and no eye shades.

I just use a piece of heavy duty aluminum foil between me and the bulb.

LuxLuthor · Jun 4, 2009

JimmyM said:
I just use a piece of heavy duty aluminum foil between me and the bulb.

Hmmm.....I have been wearing the tinfoil as a hat. You use your hat as a light shield. Ingenious. I'm devastated that aluminum foil should have had a higher priority being put in the Homeland Security Survival Kit, over duct tape and plastic sheeting. Well if I go blind, I can now narrow it down to one of two sources, although generally mom has turned out to be right about most things.

wquiles · Jun 5, 2009

You guys are having way too much fun :grin2:

Nice photos and data Lux

LuxLuthor · Jun 14, 2009

Jimmy sent me a new hex code yesterday he is working on, and wanted me to find how many amps I could drive through without the FET heating up too high, and in turn radiating heat to the CPU chip.

Note: I took Harry's advice and used my Plumber's insulator shield when Benson Torch soldering copper pipes (protects up to 1370°C). This also gave good insulation of PhD from substantial bulb heat in close proximity to rule out that factor. These were open air tests, so I have to assume that inside a flashlight body, radiation heating would be higher & cause thermal shutdown at lesser Amps. (These results are not apples to apples with previous tests, because this software version has a lower thermal shutdown trigger.)

I started out with RMS AC+DC 14.217 Amps (mV scale used on shunt)

After about a minute, temp rose above 65° C and new (lower) temp protection kicked in reducing output to 25% PWM

I didn't know temp limit was lowered, but cooled down I reduced test to about 14.158Amps, in about 2 mins it also hit the 65° C reset.

Through trial and error, I found that 13.617 Amps was highest without exceeding 60° C

13 Amps was stable at just under 55° C

JimmyM's Regulator Earns It's Doctoral PhD Degree, Despite The Heat !

Flashaholic

Flashlight Enthusiast

Enlightened

Flashlight Enthusiast

Enlightened

Flashlight Enthusiast

Flashaholic

Flashlight Enthusiast

Flashaholic

Flashlight Enthusiast

Flashlight Enthusiast

Flashlight Enthusiast

Flashlight Enthusiast

Flashlight Enthusiast

Flashlight Enthusiast

Flashaholic

Flashlight Enthusiast

Flashaholic

Flashaholic

Flashaholic

Similar threads