Mosfet Heating Circuits

[wayne.ct]

... an HHO heating system....

The HHO threads are also very interesting.

Thanks for the additional background. Replication is often next to impossible. A few small successes from time to time keep me interested and always learning.

[FuzzyTomCat]

Load Inductor Heat Profile

Hi everyone,

There appears to be questions on my documentation on "Temperature Profiling" of the made custom 10 Ohm resistor inductor that was used for the majority of my prior testing, and I forgot to even post it from what I've been told .... so sorry if I did .... and here it is ....








I hope this helps for calculations anyone else wants to try or just for documentation reference ... good luck !

Regard's
Fuzzy

[Harvey]

Thanks Glen,

I had posted the list on page two - but it's nice to have the photos too

(http://www.energeticforum.com/induct...html#post93202)

[FuzzyTomCat]

Thanks Glen,

I had posted the list on page two - but it's nice to have the photos too

(http://www.energeticforum.com/induct...html#post93202)

Hi Harvey,

I totally forgot that you placed my custom made 10 Ohm "load inductor" temperature profile data sheet in one of your postings and your right about the documentation including the dated photograph with the data sheet seen in it.

I am also very surprised no person in the entire alternative energy community has reviewed any of the "extensive" collected data and videos that was compiled on my testing and evaluation and has made any kind of remarks on the results of twenty two (22) tests .... if wrong or right ??

Best Regard's,
Glen

[waltersm]

MOSFET Switched Circuit

Hi My experience is from circuits very similar but not setting out to heat the load! My circuit is very similar but there is an inductor (intended) in place of the resistor in Rosemary's circuit. The MOSFET various types from 600v to 1500v internally limited devices. My driving circuit has a PIC and a MOSFET
gate driver. Does the same job.

In one case (1) I use a pulse width modulator PWM controlled by a HEX switch setting the mark period. Tests on the related circuit is usually limited to 50% M/S maximum. PWM is @1.6KHz.

In the other case (2) a square wave is frequency set to that desired by programming the PIC. Currently using 10KHz.

Both have been designed after making several Adam/Bedini type motors.

The main observation was initially concerning the maximising of the BEMF amplitude collected in a capacitor (many values). From the driving coils initially the after realising no moving parts were needed - the switched coil BEMF was harvested.
I soon realised the value of the BEMF was being limited (as seen by a flat top on the pulses seen at MOSFET switch off - when viewed on an Oscilloscope. In short the value reached was mainly limited by the internal MOSFET diode.
The highest value I have purchased at an affordable price is 1500V. Using this I charge a suitable capacitor to 2000V.

Back to the similarity to the Rosemary Circuit - I have experienced the inductor used to generate the BEMF getting very hot where it should not.
The lower the inductance used the greater the heating effect is. I have resorted to using a large ferrite bead as a former as this allows for heat sinking by mounting it on the box/case surface. I can also reduce the heat by
increasing the inductance but the power of the unit is reduced. This was observed by monitoring the amplitude of the current pulses seen when recharging a car battery. A small value (1ohm) was put in series and scoped the voltage I=E/R. I was seeing greater than 10A pulses at either 1.6 or 10KHz rate depending on the drive used.

Currently not tried to maximise the heat as yet - hence the interest in your tests.
This might not be the thread to outline my observations but I am finding a common link in many of the reported experiments that are aiming at different ends!!
If this is in the wrong place please let me know as more observations and feedback are of great interest.
As a ex large company Electronics Engineer I miss the discussions of old.

Mike

Ham call G3JVL

[Harvey]

Hi Glen,

Yes - it is a bit perplexing that no persons have downloaded the custom database program I wrote to assist in evaluating your data since it consists of 100,000 rows of data which exceeds Excel limits for versions prior to 2007.

I can only conclude that either they are willing to accept my limited analysis of the data or they simply are not interested in it. There is no excuse for not downloading the data and reviewing it themselves since it has been widely advertised throughout the community as to its existence and precisely how to download it.

I never received even a single PM here or elsewhere regarding the program, the tests or the data which speaks volumes regarding the lack of interest.


I still think we need continuous uninterrupted data on a single circuit design (where the oscillator and power stage draw power from the same source) with a run long enough to exhaust the supply source by at least 1.25% and academically acceptable calorimeter and hydrometer readings included in the data. IMHO, it is the only way this circuit will ever be taken seriously. Failure to at least meet that minimum puts the system well below commercially available heat pumps making it unworthy of further development by commercial standards as there needs to me some redeeming value or competitive edge.

Your custom resistor, however, stands on its own merits completely independent from any particular method of driving it. In this thread, I did a comparison of that resistor to my standard 1500W Oil based heater and your resistor wins that analysis as being more efficient even with DC. So while the driving circuit seems to be fraught with issues, I think your resistor has some viable future application worth developing. There are plenty of ways to drive it without infringing on anyone's intellectual property.

[Harvey]

Hi WaltersM,

You may want to look into the work of Charles P. Steinmetz and his Induction Furnace Patent. Understanding of Steinmetz Law along with understanding inductor core eddy currents can help in pinning down the cause of anomalous heat.

It would be interesting to find out if your inductors heat up in a vacuum or oil with the same thermal curves as they do in free air. Especially if the wire has any Nickle exposed to the surface.

If you are using a Mosfet and your inductor is acting like a resistor, then this would probably be the place to share your information and ask questions . This Mosfet Heating Circuits thread is a subset of the Inductive Resistor section.

The clamping from the internal avalanche diode wastes the energy into the Mosfet and is generally undesirable. There is also an effect present in AC systems that does not exist in the DC profiles, called the Proximity Effect and this too can lead to anomalous heating that is not otherwise explained.

Apart from these things, it surprises some to find out how much power is being dumped back through the inductor during the BEMF pulse ringdown. Simply analyzing that power across the internal resistance of the wire can explain much of the heating. An lastly, Harmonics are notorious for adding to the power losses and causing heating in inductive systems - being a Ham I'm sure you're familiar with that in the antenna systems and SWR scenarios.

Cheers,

Harvey

[waltersm]

MOSFET Switching Circuit

Hi Harvey Thanks for the comments. I am not sure my inductor is behaving as just a resistor. Firstly it is constructed on a ferrite bead about 18mm outside dia and uses 100 strand Litz wire - measures 0.193mH and 0.5ohms.
My test meter checks at 20kHz so not too far from my current 10kHz op Freq.
With just this as the MOSFET's drain to supply load I see BEMF pulses of 1500V on my scope (good job the probes from China are cheap!).

In order to collect this BEMF I am using a separate coil and capacitor series tuned by adjustment to 10kHz. (about 5mH),t with a moveable Ferrite core. This is connected from Source to Drain via 0.07uF 2kV wkg cap. & from the mid point there are two 1kV diodes in series to a 1uF 3kV oil/paper cap. The DC measured is 2kV across this capacitor. This charges in very quick time 1-3 seconds hard to determine. I have at one time earlier had either 25uF @ 4kV or a string of 2000uF 450v El Caps to similar voltage. But a bit lower with the Electrolytics due to their leakage. As I am using my AVO meter on the 2.5kV range, and the needle does not move too rapidly any time. I am also seeing (at the Centre of the Tuned Circuit) approx 200v sine wave which I can maximise. The input current from the (12v nom) PSU minimises and the voltage measures 15v (about 16v no load). Input current checked at between 150 & 250mA varying with tuning.
I will report further after some comments please.
Yes very familiar with Antenna's (Google G3JVL)

Again many thanks for your references very interesting.

Best regards
Mike

[Harvey]

Hi Mike,

You are correct, your primary inductor is not designed to be a heater. The fact that you are using Litz wire is a really good way to minimize the proximity and skin effects. This tells us that the heating is coming from ohmic loading of the smaller gauge strands and core heating. To reduce heat, you don't want to exceed the amp rating of 1A per 700 circular mils on any given strand. Another great thing about the Litz wire is that it is individually insulated - this eliminates inter-strand eddy currents that naturally occur in other stranded wires. So this leaves us with core losses. The core will experience both the Steinmetz losses and eddy losses and those are very material dependent. For more on calculating those losses especially with nonsinusoidal waveforms see:
http://engineering.dartmouth.edu/ind...apers/IGSE.pdf
http://engineering.dartmouth.edu/ind...papers/gse.pdf
http://www.micrometals.com/appnotes/...lossupdate.pdf

The last link is pretty good as it provides a much better match to the empirical measurements than the Steinmetz formula alone - at least for that product line at Micrometals.

I think you may still be losing energy into the primary inductor on the BEMF reversal. You may want to experiment with placing two of your diodes between your primary inductor and the Mosfet drain so that the BEMF energy cannot fall back into the primary. The way you have described your circuit, that energy is split between the two paths in the two inductors and the diodes would block the reversal into the primary forcing the energy to dump into the secondary inductor. The problem I see though is that the tank junction will elevate to just a bit above the 3 kV output capacitor and will act to waste that tank storage on the next Mosfet trigger. This is where using a smaller capacitance in the tank works in your favor. You will still want a low resistance secondary inductor to dump the BEMF into the 3 kV output cap.

It sounds like you have a relatively good design there to recover the energy (I haven't checked your values and resonance). But you do have to keep the BEMF below the avalanche threshold of the Mosfet - that's the only real caveat that I can see in your setup apart from what I have mentioned above.

Cheers,

Harvey

[Guruji]

Heater

Yesterday I did two resistors in parallel and as Harvey told me the amperage remained the same.
Interesting I will add more and see how much heat will I can get from more resistors.

[Harvey]

Yesterday I did two resistors in parallel and as Harvey told me the amperage remained the same.
Interesting I will add more and see how much heat will I can get from more resistors.

Hi Guruji,

Are you referring to this post?
http://www.energeticforum.com/induct...tml#post121204

If so, I wasn't indicating that parallel resistance does not increase current although I can see how my post may have been taken that way.

Instead, I was stressing that we can add more resistors in parallel provided the resistance in higher in each so that the resulting current stays below the MOSFET amperage rating.

The total resistance formula for parallel resistors is:

1/((1/r1) + (1/r2) + ... (1/rN))

So, lets say r1 = 5 Ohms, r2 = 7 Ohms and r3 = 10 Ohms:
1/((1/5) + (1/7) + (1/10)) = 2.26 Ohms

Clearly, the current will be more than any one of them by themselves because the formula for current is Volts / Ohms = Amps and the lower the Ohms the higher the amps.

So, if we keep our amps to 6A, the voltage would have to be no more than 13.56V with that setup. But what if we want to switch 900V instead?

Now we could put 66 sets of those 3 parallel resistors in SERIES (900V / 13.56V = ~66) and the result would be about 6A for the entire SERIES / Parallel Matrix of 198 Resistors.

900V
\ \ \
r1 r2 r3 (set 1)
r1 r2 r3 (set 2)
. . .
r1 r2 r3 (set 66)
/ / /
0V


Each parallel set is 2.26 Ohms and that in series 66 times = 149 Ohms

So the 900V across the entire matrix of 149 Ohms = 6.04 A

And the power would be 900 x 6.04 = 5.4KW

Cheers

[Guruji]

Heater

Yes I was referring to that post. I am trying more resistors in parallel to see if heating increases.
Thanks

[Aaron]

parallel coils

Guruji,

Can you verify that the field collapses slower when you have them
in parallel? You might not even tell a difference compared to coils with
a lot of inductance that have iron cores for example.

[Harvey]

Yes I was referring to that post. I am trying more resistors in parallel to see if heating increases.
Thanks

Hi Guruji,

If you want to provide the values associated with your available resistors and the power supply information (voltage and current) I can help you arrange them for optimal heat.

Also, as Aaron mentions - if you have a means to determine the inductance that would help as well as the inductance plays an important part in the resulting impedance at specific frequencies.

Lastly, rather than assume you are using the IRFPG50 - it would be good to know for certain what the specifications are for the MOSFET you are using, so providing that will be important as well.

Cheers

[Guruji]

Heating

Thanks Aaron and Harvey for your response. Aaron unfortunately I don't have an oscilloscope cause I cannot know certain specifications but I can measure resistor. Last time I did two in parallel the thing is that amperage remained the same on the circuit as with one I was amazed
Heat was quite good on both resistors.