# Help with transformer

Discussion in 'The Projects Forum' started by jcole, Dec 30, 2012.

1. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
I am conducting electrolysis experiments, and need to step up the voltage output from a halogen lamp transformer. The output of the tranformer is 11V and 2 to 7 amps with AC current at ~94Khz frequency (square wave).

I would like to find a transformer that I could use to step up the voltage approximately 15-20x. My impression is that I will need some type of RF transformer because of the high frequency alternating current.

Anybody have any ideas about what I could use to do this?

2. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
Why do you need such high frequency for electrolysis? I thought it does not really affect anything, so standard 50 o 60 Hz would be ok?

3. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
I am actually conducting research on low energy nuclear reactions (LENR) with electrolysis and the high frequency/high voltage AC is a necessary addition to my experiment in addition to a second circuit utilizing DC current (for the electrolysis).

4. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
Maybe you would better off if you use some standard supply like an ATX one, use the 12V rail and make your own transformer and switcher that suit that, or is the 97khz frequency exactly what you need? My thought being that if you want to conduct some research, you should have some criteria on the frequency, voltage and power you need to do it the best way possible. And a stable DC supply and your own DC/AC converter should give you broader options than using an off the shelf halogen supply with a transformer.

5. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
I have a nice programmable DC power supply that could be used at some point. I was just trying to keep it simple since I'm doing amateur research.

What I really need is something that will do 400v (peak to peak) AC with pulse widths down to ~80 ns and an adjustable duty cycle. I can get down to 60 ns with a pwm unit I have operated by an Android device, but I don't know how to turn that into AC high voltage. I'd like to be able to hit frequencies between 80 and 120 khz.

I've looked at the possibility of arbitrary waveform generators + amplifier, but the amps are very expensive to get 400Vp-p.

6. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
Exactly, so this is where you need to decide what you actually need. 60ns pulse width with a sort-of square wave ouput would require amplifier capable of at most 6ns rise time, which lies in the order of 100MHz bandwidth. I don´t know what ranges you want to cover, but turning this task from a square wave at 100khz into a variable pulse width wave downto 60ns is making this way harder and more complicated than it was in the first post.
Not to mention 400Vpp, which adds another 2 or 3 orders of difficulty. And for obvious reasons you will not be able to use one tranformer for a range from 100khz to a 100MHz. And you stil haven´t mentioned what current do you expect from those 400v.

7. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
Yes, those are reasons that is on the shelf for now. Those are some specs suggested by an electrical engineer doing this kind of research. I'm not convinced that this level of precision is needed and wanted to try something that was closer to what the original researcher suggested, but not incredibly difficult to achieve.

Anyway, I appreciate your thoughts on the matter. I'm just having a hard time figuring out what kind of transformer I need from my original question.

8. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
Like I said, without knowing the current or power transfered through the transformer, it might just as well be a ferrite bead with one winding on one side and ten windings on the secondary side like in an antenna balun or similar device. The devil is in the details.

9. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
The supply side will be 11V, 2-7 amps in the 90 khz range of frequency.

10. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
I don´t understand why you specify output power in terms of input current.

11. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
Probably because I am a novice with transformers! sorry. I need it to be around 200V output. I'm assuming the power output will be grossly equivalent to input (minus losses).

So the input power is 20-70 watts. So presumably, with 200V, the current output would be between .1 and .35 amp. Please forgive my ignorance about this. Should I just wind one myself? One wind on one side and 20 on the other?

12. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
Well, do you have a plan what you want to research, or do you have just an idea to use the supply you have and try it out?
If the supply outputs symmetrical square wave you should be able to find a suitable core just by finding equations for full bridge smps on google and using the voltage and frequency, then the size of wires based on currenat and if they fit the core.

13. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
I am studying heat production in the electrolytic cell. I'm monitoring the temperatures with sensors from an Android device. The HFAC is theorized to be one triggering mechanism of low energy nuclear reactions in electrolysis. I'll be running that through the cathode in the cell and studying temperature effects.

The square wave is symmetrical based on what I've observed with the oscilloscope. I'll follow your suggestion above. I just really didn't know where to start with this. Thanks again for you time and help.

14. ### kubeek AAC Fanatic!

Sep 20, 2005
4,689
806
You're welcome, I just seems to me that studying some phenomenon in depth requires a careful thought about what, how and why is to be measured, and whether the measured value actually gives enough information about the process you're measuring, or if there is some other variable unaccounted for that could change the view.

Also, in post #7 you mention "the original researcher", care to post what you're referring to?

15. ### AfdhalAtiffTan Active Member

Nov 20, 2010
117
11
Maybe a normal ferrite transformer used in SMPS could help?

16. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
kubeek,

You can read a little about what I've done with this on my blog.

http://www.lenr-coldfusion.com/category/hobbyist-lenr/

The original researcher I am referring to is Robert Godes of Brillouin Energy.

You can read his most recent presentation here:

http://www.brillouinenergy.com/docs.php?doc=iccf17

Because LENR is not well understood from a theoretical basis, I am taking an Edisonian approach in attempting to trigger the reaction.

17. ### #12 Expert

Nov 30, 2010
16,704
7,354
"I haven't failed a thousand times, I just know a thousand ways that won't work". Right?

18. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
Exactly #12. So far, I've tallied up quite a few ways that don't work, but it's been very interesting and I've learned a lot.

19. ### wayneh Expert

Sep 9, 2010
12,394
3,246
You'll have enormous difficulty convincing anyone of your findings, if you do hope to claim excess power production. Such experiments, when conducted carefully in a well-funded lab using state-of-the-art calorimeters, are still subject to intense scrutiny and suspicion. As they should be. Extraordinary claims require extraordinary proof, and DIY calorimetry isn't going to come close. Can't you look for neutrons or something a little more definitive than heat flows?

20. ### jcole Thread Starter New Member

Dec 30, 2012
12
0
I'm not really concerned with convincing others at this point. I would have to be convinced myself first, which will be difficult to do. This is more of a hobby for me, and if I find something impressive, then non-amateurs could test it out. I've demonstrated some methods that fairly convincingly show the absence of excess heating with what I've tried so far. I'm not looking to measure milliwatts of excess heating as those with advanced calorimetry do. I'm only interested in large amounts. Also, it's just enjoyable to learn and experiment.