PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2022-01-27, 02:10:29
News: If you have a suggestion or need for a new board title, please PM the Admins.
Please remember to keep topics and posts of the FE or casual nature. :)

Pages: [1] 2 3 4
Author Topic: A Simple Test for Large COP Claims in Pulsed Inductive Heater Circuits  (Read 49655 times)
Group: Guest
The purpose of this thread is to present a very simple way to determine the approximate COP (+/- 20% or so) in circuits that utilize a battery as input and an inductive resistor as load (or any device intended to output thermal energy as its primary COP claim).

The straightforward techique is a valid and quick way to prove one way or the other whether COP claims of say, 2 or more are true or false.  It requires only a regular high input impedance volt-meter, a thermometer and a variable bench power supply along with a couple of resistors and a capacitor and an SPDT switch (in addition to the D.U.T. or "Device Under Test" and its battery).  The exact same load element (Rload) is used in both setups, DUT and Control.  A clock with a second hand may also be required.

The test has been designed to sacrifice unnecessary efforts at absolute accuracy that might be required when testing the validity of small COP claims (well under 2).  It has some fairly small but known, predictable error sources.  The test has been designed, in general, to err on the side of showing a higher COP than is actually present where errors exist.  Thus it gives "the benefit of the doubt" to the claimant.

I, myself, would not normally use a battery to power the D.U.T. but there is often talk that the battery is an integral part of the operating mode and may be at least partially responsible for the high COP expectation, so a battery it is!

I will try to pose the usual expected arguments against the test regimen (be my own Devil's Advocate) and give logical arguments to refute or at least qualify the maximum error a given potential flaw in the setup/process might introduce.

Please bear with me and be patient.  I have apparently not mastered text entry on the forum; the text appears to scroll randomly once the box is full, thus making real-time editing nearly impossible.  Until I master this problem, I will do my text editing in Word and then paste it in.  It may take several days to present it all in several posts.

The first (immediately following) post will simply be the diagrams, with no text.  Many will find that to be a fairly complete description in itself.  Many will not.  Explanatory posts will follow in the next few days.
« Last Edit: 2011-01-15, 10:35:53 by humbugger »
   
Group: Guest
Here is the test setup.
   
Group: Guest
Preliminary assumptions and caveats:

1.  The black box contains anything you like except for ancillary batteries.  If it requires, as expedcted, internal bias power to run the internal circuits it is highly recommended that these be generated and taken from the source battery (i.e. that the D.U.T. remains a four terminal device, as shown).  If an external separate battery is absolutely required (if the D.U.T. designer cannot include an appropriate high-efficiency switching regulator to condition the battery voltage to an appropriate level for the internal circuitry for reasons of inability or otherwise and the internal drive circuitry draws significant power (more than 5% of the input power)), then a second battery may be used and its power draw may be measured by the same techniques and added to the power measured from the primary battery.  This requirement may be waived if it can be shown and proven that the auxilliary power does not show up in the output beyond a couple of percent.  It is much preferred, however, that all input power come from the primary battery to avoid this potential error and complication of the tests.  Even a lossy resistor/zener regulator will be fine.

2.   The variable bench power supply used in the control setup should have the ability to output at least as much power as is measured in the DUT/battery input power test and it should be able to be adjusted in output voltage from at least 1/10 of the battery voltage to well above the battery voltage (this may or may not turn out to be necessary but is a good starting point.  The supply should be of low impedance (i.e. a reasonable equivalent to an ideal voltage source with an impedance in the neighborhood of the shunt resistor Rshunt or lower).  It is acceptable to use more than one bench supply in series as necessary.  The supplies must have sufficient current ratings such that no internal current limiting is enabled.  In other words, they are not allowed to quit or go into current-source mode for self-protection.

3.  A standard "bulb" thermometer will be glued to the middle of the load resistor or at any location the claimant prefers using thermally-conductive "100% silicone rubber" (you will find that the standard GE products are quite thermally conductive, especially the black stuff GE 50B.03 but most any will do).  The thermometer range should cover from room temperature to at least 30 degrees F above room temperature.    A medical thermometer may be used if one is prepared to wait a half hour or so between tests to assure that the starting temperature is back to room temp (off the scale).  Around the 100F range, the markings should be spaced wide enough to read with reasonable (1 degree) accuracy.  It may be placed in an oil bath or inside a shoebox to prevent "wind" and temp gradients but there is probably no need for these measures since the very same load/thermometer is used in both the DUT and Control runs.  It should not be moved between tests.  It must settle to a standardized constant room temperature prior to beginning any test.



« Last Edit: 2011-01-15, 11:54:09 by humbugger »
   
Group: Guest
There is much more to come but I'd like to open the thread for comments and/or arguments regarding what has been presented so far at this time.  Even if it's just to indicate that there are interested readers.  No point in me babbling on if no one is reading or no one cares or anyone feels that the initial conditions are invalid inherently, so please chime in. ;D

If logically insurmountable flaws can already be seen by anyone and cannot be convincingly addressed by further explanation or modification, please state your objection and drop out of the thread.  If I cannot give a solid response to allay the objection, I will agree and terminate the thread.  Don't want to build upon a flawed foundation!  That would not be science!

But please don't jump ahead and speculate on what has yet to be described.  Let's keep argument and comment to what has been given so far before we proceed.
« Last Edit: 2011-01-15, 09:50:14 by humbugger »
   
Group: Guest
 Glad to see this thread. Just hope somebody will actually appreciate your suggestions. And use them!

 8)
Catlady
   
Group: Guest
Welcome Catlady!   O0

One more assumption:  

4.  The battery used in the DUT test setup should be robust enough such that its voltage remains constant within a couple of percent during each test run.  The tests are expected to be fairly short in duration; on the order of a couple of minutes maximum.
« Last Edit: 2011-01-15, 11:03:52 by humbugger »
   
Group: Guest
My own arguments with myself regarding 3. on the temp measurements:

Depending on how fast the DUT can raise the temp 30F, a higher number may be better here.  We will be timing how long it takes to achieve a given temperature rise, not the end temperature.  If the thermal mass of the load is small and the heating is too fast, it may be tough to get an accurate time reading from startup to the desired 30F above ambient test termination temperature.  If that turns out to be the case, a "lab" thermometer may need to be used, giving a higher range and typically they are nice and long for accurate readings.  I've seen cheap "outdoor weather thermometers" running up to 140F that might be useful here.  

Alternately, to retain simplicity and the ability to use the original 100F ~ 120F (and thus the medical or cheap outdoor weather thermometer) as a marker, the thermal coupling could be degraded on purpose between the resistor and thermometer bulb by putting the bulb near an end of the load or by using a thick built-up layer of the silicone rubber instead of contact-gluing the bulb right to the resistor body.  Since both tests use the same exact load/thermometer setup, this should not introduce significant error.

I would figure that test times at least twenty seconds would be adequate (the longer the better) to get a time mark within 5% (one second) and give time to read the meter.  Oh...we'll also need an on/off switch to start and stop the test (not shown).

Here I may have opened a Pandora's box if the claimant states that their black box takes more than a second to "warm up", "find resonance", etc.  A long test time will diminish any validity of those arguments but may require a higher end temperature.
« Last Edit: 2011-01-15, 12:52:05 by humbugger »
   
Group: Guest
Okay...maybe it's worth considering a meat thermometer and/or a big jar of water or a closed box with a fan even.  Anything to increase the thermal mass and slow the temperature rise.  As long as the total load setup is identical on all tests it shouldn't matter.  That means the same single load setup and thermometer in the exact same environment under both tests.

You see, the problem is I'm trying to be extremely general and keep it cheap and simple.  

If it takes any significant time for the black box to "come up to speed" so to speak, then this amount of time needs to be known.  

If it takes very long, a different approach must be taken because the control circuit (two wires) is ready to rock and roll immediately and will always show a COP very close to unity (1.00) right from the get-go.  It will follow and adhere to the "classic thermodynamics" laws and convert all the measured electrical power into heat (but no more).  Comparing the time it takes to get a specific temperature rise would give the control test an unfair advantage.  If the DUT is also ready to perform at full COP right away after being powered up then none of this is a problem.


« Last Edit: 2011-01-15, 16:54:58 by humbugger »
   
Group: Guest
A differential test using two seperate loads and run simultaneously with a servo loop on the control setup to force it to temperature-track the DUT load's temperature would be great but is too complex and introduces arguments that the two loads may not be identical or in identical environments.

The other obvious alternative is simple and, if there is concern about slow start-up of the DUT, is probably the right answer:

Let the DUT run as long as it takes for the load to reach thermal equilibrium with its environment and measure the final temperature.  This means a bigger battery and a longer test because we want to make sure the battery voltage doesn't change more than a few percent from start to finish.  Measure the power and then move the load into the control test setup, let it cool down some (an advantage to this method is that we don't care how long it takes or what the starting temp is) and then twiddle the power supply voltage until the control setup reaches thermal equalibrium at the same temperature.  Then measure its power input.  This takes time out of the equation (toss the stopwatch) but has one potential disadvantage.

As the load gets hotter and hotter compared to the surrounding air, it also dissipates the heat at a faster rate due to increasing convection.  This means the temp rise vs power in the load is not linear and gets flatter as the temp gets higher.  A given increase in power will cause less and less temperature increase as the load gets far above ambient temp.  This trends toward a more error-prone measurement and may take a much longer time to adjust the control setup to exactly match the DUT's final temp.  Someone please jump in here and beat me up on this if I'm wrong, but I don't think so.

So, those are the options and the reasons why it would be most effective and simplest if the "high COP claimer" will admit a DUT which starts right up immediately at full COP.  I trust the time measurement to be more accurate at the lower temps involved and the battery voltage drop is less of an issue.

So, rather than arrogantly pontificating a final universal solution here, I am laying out my thought process for all to see and critique, given these few unknowns..
   
Group: Guest
And now it really is time for me to shut up and grow ears for a while before I ramble on to the details.

 :-X
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
Thank you for your simple test measurement technique, and easily understood schematic test diagram.

Seems we are on the same page. I have also posted in this regard.

Tseung thread: post#38, post #46, (latest post #168 completely ignored)

Ainslie thread: post #5, post #7, post #11 and many other related posts.

While all the arguing was going on, I built up a crude blocking oscillator and tested its efficiency at 75.0% using the method in post #168 of the Tseung thread. This took less than an hour.

I drew up a two step method outlined in the attachment and had intended to post it early in the Tseung thread, but as the existing described methods were misunderstood by many, I realized the purpose of the thread seemed more about discussion.

Regarding some of your worry about temperature flattening at higher values, I have not found this to be a problem over the range you will expect to see in such low power devices.

I also do not worry about the time factor since with a fixed power input the temperature will stabilize as there is a fixed thermal resistance to the ambient in a semi insulated container.

I tend towards thermocouples as I have a lot of accurate TC instrumentation in my lab. I also can easily fashion tiny low mass thermocouples. Thin film RTD's are also very accurate and can easily be read or used in a simple bridge for differential measurements.

I like your idea of using a single voltmeter and the current shunt as it normalizes error sources.

You sir, are very obviously a person that is skilled in the art of practical real world test methodology, reading your posts here is most refreshing and familiar as I also walk that path.
« Last Edit: 2011-01-15, 18:38:07 by ION »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Guest
Right on ION  8)

I've seen many of your posts and count you among the most respectably knowledgable contributors here, right up there with Poynt99 and MileHigh.  Keep it up, bro!

Regarding your diagram, I would only make two "devil's advocate" type comments.  These are merely predictions of what might be challenges, minor as they might be in reality.

1)  There is an opening for argument that you are allowing the use of two heat sources that are physically different in shape, mass and location within the box.  Folks might yell and scream that the air temp inside the box may have strata and that those strata might not be all planar but instead rather random based on convection airflow inside the box.

2)  Some might object to the fact that you are putting the entire DUT in the box, thus allowing its internal heat (i.e. what are normally considered inefficiencies)  to be treated as "good DUT output power".  Maybe I'm missing something, but I'm guessing the DUT load must be a resistor (not an LED or a buzzer or something that produces other-than-thermal energy).  I don't get why the DUT itself and its load are not shown as seperate entities and why they are both inside the box.

P.S.  I love the turtles quotation.  Wonder how many here know the source and context of that one.  It's a classic!

   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
Right on ION  8)

I've seen many of your posts and count you among the most respectably knowledgable contributors here, right up there with Poynt99 and MileHigh.  Keep it up, bro!

Regarding your diagram, I would only make two "devil's advocate" type comments.  These are merely predictions of what might be challenges, minor as they might be in reality.

1)  There is an opening for argument that you are allowing the use of two heat sources that are physically different in shape, mass and location within the box.  Folks might yell and scream that the air temp inside the box may have strata and that those strata might not be all planar but instead rather random based on convection airflow inside the box.

2)  Some might object to the fact that you are putting the entire DUT in the box, thus allowing its internal heat (i.e. what are normally considered inefficiencies)  to be treated as "good DUT output power".  Maybe I'm missing something, but I'm guessing the DUT load must be a resistor (not an LED or a buzzer or something that produces other-than-thermal energy).  I don't get why the DUT itself and its load are not shown as seperate entities and why they are both inside the box.


As I have said before, it is important to know when a test is "reasonable" and this means using common sense. Yes, heat strata could be an issue, but heat rises and I always make my measurements at the top of the container at the same location. Air in the container will circulate and mix to a certain extent. What is reasonable in this regard?

As for point 2, I did separate the DUT and it's load early on and POYNT was dissatisfied with this and suggested a measurement of transistor temperature as an added variable. Then if we were to nit pick, we would have to measure the losses in each component, transformer, resistors etc. To satisfy the objection, I lumped the DUT and it's load and treated it as a single entity. Actually the separate heater shown in the last drawing could be the DUT's load resistor, and fed by bringing out a separate wire when the DUT is not energized as part of step 2, as the resistor is normally isolated by a "catch" diode.

To expand a bit, in thermal methods, we can treat the DUT and the load as heat producing items. Then we need only find if more heat is produced than we are inputting to the system, by comparing to a known equivalent heat producing power input. The system can always be broken down later to find the contribution of the separate elements. As long as COP=around 1 there is no OU.

To be devils advocate I also poked holes in my presented methods in order to pre-emptively hush some of those that would complain of energy lost in RF radiation, audio, and proposed particle creation etc. There is always someone to complain about something....but what is reasonable?


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Guest
"Reason" is not always present when the feet of the COP>1 claimant are held to the fire.  Your explanations are completely reasonable and satisfy me!   O0
   

Group: Administrator
Hero Member
*****

Posts: 3173
It's not as complicated as it may seem...
ION,

I like your last diagram. I think it captures all the minimum requirements.  O0

The only potential hitch, or "devil's advocate" point I see is the measurement of the input power. The current is pulsed, and it is an unknown how the power supply will deal with that. RF or HF can feed back through the power supply lines possibly messing with the supply's regulation and metering.

.99
   
Group: Guest
Imagine running a poorly-terminated (high VSWR) 5KW 40MHz RF amp on the bench with no shielding.  We used to have to use twisted pairs and big ferrites and super-lowZ caps all over our power supply leads to keep supplies in regulation.  One time a big Instec supply actually just blew up and went flaming despite all that!  Not to mention DMMs!  We always used old-fashioned Simpson 260's but even then had to twist-pair the leads and use ferrite cores and caps on them near the meters.   :o (my brain after five solid years prototyping RF generators at close range)
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
ION,

I like your last diagram. I think it captures all the minimum requirements.  O0

The only potential hitch, or "devil's advocate" point I see is the measurement of the input power. The current is pulsed, and it is an unknown how the power supply will deal with that. RF or HF can feed back through the power supply lines possibly messing with the supply's regulation and metering.

.99

Proper bypassing of the DUT should prevent any backfeeds of RF without affecting the operation of the DUT.

After all, a good battery has a very low impedance, so a good low ESR electrolytic bypassed further with a quality ceramic or film capacitor should present an equivalent transient low impedance. The power supply will then see only the actual steady state current and voltage being used. A common mode choke might also be helpful.

I have not shown all these things in the drawings as they are assumed in good laboratory practice.

We are dealing specifically with a device that is most likely under one Watt capacity. 5 kW is another matter indeed.

Regarding meters: I've got dozens of DMM's lying around ...all levels of quality....but my Simpson 260's? some quote about cold dead hands rings a bell LOL
« Last Edit: 2011-01-15, 19:56:21 by ION »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Guest
But I heard that transverse longitudinal radiant energy zipons are diverted abruptly to far parts of the universe by conventional "good bench practice" bypassing and filtering.    ;D

So you feel the same way about your 260 as shooters feel about their guns and silver-stackers about their stash. huh?  Me too.
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
But I heard that transverse longitudinal radiant energy zipons are diverted abruptly to far parts of the universe by conventional "good bench practice" bypassing and filtering.   ;D

Those "zipons" can ruin a good day  :o


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Administrator
Hero Member
*****

Posts: 3173
It's not as complicated as it may seem...
Imagine running a poorly-terminated (high VSWR) 5KW 40MHz RF amp on the bench with no shielding.  We used to have to use twisted pairs and big ferrites and super-lowZ caps all over our power supply leads to keep supplies in regulation.  One time a big Instec supply actually just blew up and went flaming despite all that!  Not to mention DMMs!  We always used old-fashioned Simpson 260's but even then had to twist-pair the leads and use ferrite cores and caps on them near the meters.   :o (my brain after five solid years prototyping RF generators at close range)

At work we have these old 20V, 30A power supplies. Built like a tank (15x 2N3055's) and has current limiting. The story goes, that while powering a Motorola 30W VHF mobile radio, while keying the transmitter, the PSU would lose it's mind and become erratic, and that is even with correct 50 Ohm termination!

The PSU company subsequently came out with a mod to filter out this RF backfeed, and they worked fine after its installation.

On a side note, I try never to assume anything when it comes to what "unseasoned" electronics hobbyists may know.  8)

.99
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
At work we have these old 20V, 30A power supplies. Built like a tank (15x 2N3055's) and has current limiting. The story goes, that even while powering a Motorola 30W VHF mobile radio, while keying the transmitter, the PSU would lose it's mind and become erratic, and that is with correct 50 Ohm termination!

The PSU company subsequently came out with a mod to filter out this RF backfeed, and they worked fine after its installation.

On a side note, I try never to assume anything when it comes to what "unseasoned" electronics hobbyists may know.  8)

.99

Often stuff that is "built like a tank" has an Achilles Heel.

Regarding your side note, I'm not really interested in teaching sub 101 electronics to unseasoned hobbyists, although there are many here that enjoy that very thing.

I pointed out in one of my rants that if we have unfortunately the very seasoned guys that have been in the industrial trenches designing and testing equipment all their lives thrown in with the unseasoned hobbyists, it is like expecting a successful outcome to a Manhatten style project by a lottery of citizens. What are it's chances for success?

Is this an "overview of electronics" teaching forum or a "research forum" where reasonable expertise is expected and assumed?

Delete if off topic, but the point should be addressed. What level of detail or explanation is required in each post?


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Guest
That's the trouble with us old farts.  We are so used to hunting for the Achilles heels in our own designs that full abandon to wild optimistic enthusiasm is sometimes hard to muster.  But what would a forum be without the young and innocent?

I suspect your last question was directed to Poynt99 but my answer would be NONE!  It's all voluntary until you start making extraordinary claims, I'd say.  It can get pretty frustrating, though, to be run around in circles by wild-eyed new-age jargon-spewing claimants who know nothing about making things that actually perform their specified functions reliably every time.
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
I agree, but have often found myself spending too much time in explanations to the point of frustration.

On another note, here are pics of the styrofoam containers that I use to house load resistors and DUT. The simple blocking oscillator fits easily. I use them upside down so the tapered long end is at the top.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Administrator
Hero Member
*****

Posts: 3173
It's not as complicated as it may seem...
Often stuff that is "built like a tank" has an Achilles Heel.
It's relatively simple, but uses a fairly sophisticated Motorola regulation IC.

Quote
Regarding your side note, I'm not really interested in teaching sub 101 electronics to unseasoned hobbyists, although there are many here that enjoy that very thing.
I suspect that Lawrence and the Professor fall into this category, as do many of the members here I believe. I suppose it comes down to how far the offer to help these folks extends. You offered a solution for measurement to Lawrence and the Professor. Clearly they are not aware of many of the factors and subtleties involved regarding any measurement method, except perhaps basic DMM measurements. Do we offer them a somewhat high level solution, then abandon them when they ask for further clarification and help?

I suppose we all have our own approaches. I've always tried to carry through with support until all their questions are answered and they are flying on their own. IMHO, if we don't, our offerings may have been made in vain. But each to his own in this matter.

Quote
Is this an "overview of electronics" teaching forum or a "research forum" where reasonable credentials are entry requirements?
To date there are no real credentials for entry to OUR. We felt that all should be welcome, but all will be moderated. Most of the folks that join these forums are hobbyists at best, therefore learning will always be part of the forum content. In fact, none of us stop learning here, and that is one of this site's assets I feel. There are several threads being discussed at several different "levels", and the beauty of that is there can be something of interest for everyone.

.99
   
Group: Elite
Hero Member
******

Posts: 3539
It's turtles all the way down
POYNT

I had earlier edited that post to read "reasonable expertise is expected and assumed" as credentials often are not the measure.

It was not picked up in your reply.

Quote
I suppose we all have our own approaches. I've always tried to carry through with support until all their questions are answered and they are flying on their own. IMHO, if we don't, our offerings may have been made in vain. But each to his own in this matter.

Actually, I have not found this to be the case. Despite all of the many meaningful measurement explanations you gave on the Ainslie thread, the result was more and more disagreement and argument to no conclusion. I see the same trend developing in the Tseung thread.

There is a saying in the I CHING that in order to teach you must have a willing student. In the prior cases mentioned you don't have students, rather people convinced they are right before any realistic measurements have been made, and will argue against all your good intentioned help vehemently.

Just observing trends here. We can always hope though.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Pages: [1] 2 3 4
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2022-01-27, 02:10:29