Matt
Hello Matt! I made a VIC Bobbin from old USSR TV. Looks funny! I hope it works! Could you share a schematic logic diagram of VIC v5 Rev0 driver for a better understanding. Thank You! :)

https://drive.google.com/file/d/0B2WHxX-dD1NhTV9uN0lSVWVWckU/view?usp=sharing

Quote from DimDimuch on May 21st, 2017, 11:30 PM

Could you share a schematic logic diagram of VIC v5 Rev0 driver for a better understanding. Thank You! :)

Funny, I would have thought I did already, but scanning through the thread, I didn't see it.

Here it is.

I've started testing the VIC driver V5 rev0 board. I though I'd share some preliminary results and scope shots, and ask some questions as I could definately use some input.

So, my current setup is as follows: 11 cell WFC. I'm using the 5 coil VIC design. The coils are wrapped according the Don's impedance measurements. So Prim: 10,5Ω, Sec: 71,3Ω, Choke+:76Ω, Choke-:70,9Ω, feedback:+-11Ω. It's not 100% accurate (resistance measurements vary with temperature), but close enough?

The core is N27 core material, with permeability of 1820 and should be able to handle frequencies up to 20 kHz. No air gap between the cores. I did some basic tuning of the VIC driver board as described in Matt's tuning video. TIP120 roll-off is set to around 10kHz.

First, I did a test with all 11 electrodes hooked up. See attachments for the screenshots with prefix '11 cell'. Measurements are taken across L1. It tuned it as Ronnie described in the Understanding How Stan Meyers Fuel Cell Works, as in take it to 2 volts, tune it, take it up to 4V and repeat up to 12V.

My second test was with only 1 electrode connected. Screenshots prefix '1 cell'.  In the screenshots the PLL locks quite nicely to a solid 6,1kHz and at this frequency the core purrs like a kitten, but no gas production... I did notice the tuning is extremely sensitive. A slight change on the pots and the signal is all jibberish. Also, I needed to adjust the frequency every time I increased to voltage. At 6.1kHz, is this a step charge of the cell I am seeing in the screenshots?

The max input I tried driving the VIC was about 15V while consuming about 700mA. The top voltage I saw during testing was about 450V peak to peak. I need to order new probes that can handle higher voltage...

Obviously I need to do more tests, but I would ask all of you for some input on these results. How does this waveform look to you? What should I try next? How and where should I take meaningful measurements to share with everyone? Is it ok to take measurements across the primary (as in scope probe and probe ground connected to + and - output of the primary)?

@Matt, how high voltage could I put in this board? I stopped increasing the voltage to the board at around 15V, but is it safe to go higher without blowing up the board? Is it supposed to go higher, or is 12V the maximum?

I know, some noob questions here, but my eternal gratitude for any feedback. Now more than ever. :)

Cheers,
Henne

Quote from Henne on June 30th, 2017, 06:21 PM

@Matt, how high voltage could I put in this board? I stopped increasing the voltage to the board at around 15V, but is it safe to go higher without blowing up the board? Is it supposed to go higher, or is 12V the maximum?

I think the regulators will allow you to go clear up to about 30 volts, but you'll need to keep an eye on the current draw--don't let it get beyond more than 1.5 amps.  I think pushing it that hard you'll definitely notice regulators & transistors getting hot.  Also a good chance of you smoking your primary.

The other thing you have probably noticed already is when you increase voltage, tuning becomes more and more difficult.  Ronnie stressed this early on and is why you tune a little, then slightly increase voltage and tune some more.

You've done really great work Henne coming this far.  I sure wish Ronnie would jump back in here and help you to the finish line.  The only thing I can add at this point is the core gap--that's another tuning variable you can work with to maybe begin seeing some results.

Thx Matt for the feedback and your encouraging words. The past few months I've been focusing heavily on getting a test setup built. I'm excited I got to this point and curious what comes next! Although it has to be said, I would never have gotten this far without this forum and the open-source spirit of its members, so you guys deserve the credit really. I can only hope to ignite the momentum on this project again.

Good to know 30V is possible, as this gives some margin to play with. I'm using 29AWG, so I'll burn out the coils (at around 1,2A) before the board can blow up. So it's a built- in safety feature right there. :) In my first tests I did notice the voltage regulator getting hot. I'll install some heat sinks just to be sure.

I've taken your advice and introduced a small core gap. Quite a difference! At just 9V into the primary, the voltage across L1 went off the scale (see attachment). I ordered new probes that can handle higher voltages. Wouldn't want to blow up my oscilloscope... Now it's waiting for the new components before I can continue. More info on the progress as it comes!

Cheers,
Henne

Quote from Henne on July 2nd, 2017, 05:19 PM

In my first tests I did notice the voltage regulator getting hot. I'll install some heat sinks just to be sure.

Yes, do install a heat sink to the LM338 regulator for sure.  It's a linear voltage regulator and rarely used any more because of its inherent inefficiencies, but plenty good for this application.  Once you go above about 30 watts of power handling, then it's time to look for something more efficient in the switching regulator category.  The goal here though, if everything is working correctly, you should never have to push more then 10 to 15 watts to get your cell producing gas.

Quote from Henne on July 2nd, 2017, 05:19 PM

I've taken your advice and introduced a small core gap. Quite a difference! At just 9V into the primary, the voltage across L1 went off the scale (see attachment). I ordered new probes that can handle higher voltages. Wouldn't want to blow up my oscilloscope... Now it's waiting for the new components before I can continue. More info on the progress as it comes!

I really don't know how accurate you'll need to be with this version of the VIC in adjusting your gap spacing.  It might be good to find a set of plastic feeler gauges.  You are using much thicker cores than Stan's original, so it could be that you'll end up having a quite large gap.  Or the reverse could be true that you'll end up having a very tiny gap that is difficult to precisely set.  I just don't know.  You're in uncharted waters unless Ronnie decides to jump in here and assist.

Something you guys may want to snag is a suitable isolated DC 2 DC power converter for powering your VIC Driver Board.
http://www.mouser.com/Power/DC-DC-Converters/Isolated-DC-DC-Converters/_/N-brwkv/

I see some 25 watt devices that should do the job.  A little pricey but still much cheaper than even budget differential scope probes.

If you don't have differential probes and your scope isn't a high-end power analyzer type (like Russ') with differential inputs already built-in, please do use a battery with a fuse and preferably an inline amperage meter.

I mention this because some people missed the fact the two output conductors from the VIC Driver Board are NOT ground referenced.  Most benchtop power supplies are as well as cheaper benchtop oscilloscopes.  If you forget about this hidden connection between your power supply and scope, problems will ensue.

 quote of Henne :
Also, I needed to adjust the frequency every time I increased to voltage.

Henne what you use to adjust the frequency?

Thank
andy

Quote from andy on July 4th, 2017, 11:40 AM

Henne what you use to adjust the frequency?

Yes, good question Andy.  I would have thought the PLL would lock automatically.  When you say frequency are you referring to the gating, because that would stay fixed regardless of what the PLL does?

Can you answer , Henne?

I think he's travelling for a couple of weeks Andy.  He should be back after a bit.  I'm sure he is far from done testing his new setup.

Matt
Do you have tested your board with the coils from Russ?

Not as yet.  Still learning a few things and working out a strategy for acquiring a cell.

I'll send you my cells for testing Matt if you wish. ~Russ

Quote from andy on July 4th, 2017, 11:40 AM

quote of Henne :
Also, I needed to adjust the frequency every time I increased to voltage.

Henne what you use to adjust the frequency?

I see how this could be interpreted ambiguously, I was not very clear in my explanation. What I meant by that is that I need to tweak the center frequency of the PLL ever so slightly (by turning the potentiometer on the board) when increasing the output voltage of the VIC driver. At a certain point I though perhaps this could an isolation/ground loop issue, but even when powering the board with a 12V battery, I need to do this adjustment. At increasing voltage the adjustment of the center frequency becomes very sensitive.

However, I did notice that increasing the voltage somehow distorts the output signal (see attachment). This happens noticably more when running of a power supply than when running of a battery. Unclear at this time what is causing this. Though I still need to resolve the isolation issue Matt mentioned above...

In other news, I've run some more tests with my shiny new high voltage probes. I've installed a Schottky diode rated to 1,7kV repetitive reverse voltage in the VIC. The result is an increase of the voltage to the cell to about 1,4kV peak-to-peak (see attachment), consuming about 0,9 amp. No gas output, no surprise. Is this the correct waveform? I've seen different results with other people, so I'm a bit unsure...

I've been looking into the charge ratio to the cell as explained by Ronnie. Currently, the signaI across L1 and L2 are very similar. When introducing an air gap between the cores, the voltage across L2 (-) actually becomes lower than the voltage across L1 (+), which is not what I was expecting. I'll have to dig in deeper in how to achieve a 2:1 charge ratio by changing the # of windings on the coils and the core gap. Any suggestions are appreciated. :)

@Daniel: I have no problem at all if my scope shots are posted on fb (sharing is caring). However, I would kindly ask you to add a link to the corresponding post on this forum instead of to your own youtube channel, as this prevents people to participate in this discussion in the true open-source spirit.

Cheers!

Quote from Henne on July 20th, 2017, 06:22 AM

the center frequency becomes very sensitive.

increasing the voltage somehow distorts the output signal (see attachment). This happens noticeably more when running off a power supply than when running off a battery. Unclear at this time what is causing this.

My interpretation is that when you increase voltage, you in-effect increase the Q-factor of the circuit, because any wire resistance in the circuit does not change.  A higher Q-factor means tighter/narrower bandwidth and any external noise can/will pop the system out of resonance.  The PLL will attempt to regain resonance, but this doesn't happen instantaneously.

A battery will have pretty much zero ripple current, not true with a DC power supply.  That ripple current is probably introducing just enough noise when your Q-factor is high to manifest the symptoms you are seeing.

Henne, are you getting 1,4kV peak-to-peak with water in your cell?

Thanks,

~Russ

Henne, are you getting 1,4kV peak-to-peak with water in your cell?
 Measurements are taken across L1 or  across your cell?
thank

Yep, there was filtered tap water in the cell during testing. That scope shot was take across L1. It's peculiar that 2 people ask. Are these surprising results? I could put together a video if people are interested.

I was putting in about 23V in the primary, amp usage was about 1A. This was to see how hard I could drive it. I think I could go a little further, but 1,2A is the maximum 29AWG can handle and this was the maximum I felt comfortable with at this time. The primary coil already gets really hot!

Next up, I'm going to look at the charge ratio of the cores.

Henne, are you getting 1,4kV peak-to-peak  across your cell with water in your cell?
Thank for answer.
andy

Henne
When you introduced an air gap between the cores, the voltage across L2 (-) actually becomes lower (fall) than without the air gap?
Or with the air gap  the voltage across L1 actually becomes higer ( rises) than without the air gap?
thank
andy

Quote from Henne on July 22nd, 2017, 08:53 AM

It's peculiar that 2 people ask. Are these surprising results? I could put together a video if people are interested.

Years ago members here at the forum had difficulties to get voltage into water up to the hundreds volts range by using transformers. When Ronnie posted information that high voltage should not be needed to create (an unknown amount of) gas in the water  they focussed on other aspects than voltage. we never got information what voltage Ronnie used over the cell to create moderate amount of gas.

AFAIK Ed Mitchell was the only person who reached 8.9 kV in water in 2013. Once he created a huge amount of gas and then the transformer failed.
The assumption is made that it takes 10 kV for a 10 electrodes serial cell to get over the ionization point and create gas. Until then there is a minimum amount of bubbles but then an avalanche effect starts and creates lots of gas.
Ed´s waveform was an AC waveform with 8.9 kV peak to peak at 5 pulses and then a gating pause. It´s unknown yet if the waveform Meyer created was a step charging DC or step charging AC waveform.
Interesting to me is that also an AC waveform creates a step charging effect from pulse to pulse. So the transformer stack or the water seem to change their electromagnetic parameters from pulse to pulse.

Quote from Henne on July 20th, 2017, 08:11 AM

Is this the correct waveform? I've seen different results with other people, so I'm a bit unsure...

can you please stretch the pulse sequence you posted to show the shape of a single pulse or a group of 5?

Something to note:  A step-charge may also be considered a parametric oscillation.  Controlled series/parallel resonance is a key to getting it.