Set a jar of mixed up biodiesel/glycerin on top of the magnets and see what happens. I don't expect anything, but it won't hurt anything.

-Jim

quote:

Originally posted by Dogma:

quote:

Using a large number of lower value resistors is essential, to distribute the voltage stress and power dissipation across each.

In fact, I'd probably be a little more liberal and limit current to 10mA vs. 1mA.

I thought from Graham'd early postings that the aim was for 0.1mA? Milliamps that is.

Would these be ok for rectifying my transformer outputs?
HV diodes on ebay from China

can I use them in series if I have higher voltages ie two per leg of a bridge rectifier for 30kv etc? I would prolly pot them in epoxy or silicone once soldered up.

Ok did a quick read up on diode strings and and found you get a higher DC voltage out of your rectifier than AC in. approx 1.4 times. Then you want to add a safety factor of 1.5 or 2. so for 30kv you would want 42Kv plus 21Kv or 42Kv. 63kv to 82kv per leg. However as two legs share the load at any one time... 42Kv would do and 63Kv is well over engineered. Having a resistor or choke on the output should protect the rectififier from any possible 'ringing'as well as limiting the current safely. So four diodes per leg (64kv) would suffice for 30Kv ac input and give a 42Kv DC output. I'm just confused about forward and reverse voltages. Are the linked diodes 16kv or 50v for our purposes?

Hi Ant

Don't worry about DC just yet - see how you get on with AC first - you should be pleasantly surprised.

0.1mA is fine in the small batch experiments.

Hi Dogma,
I'd recommend you don't go above 1mA, for safety's sake - current is the killer. Limit that to 1mA and you should be reasonably safe.

I stick firmly by my recommendation to use 10 x 100K resistors per kV if using general purpose resistors readily obtained from Radio Shack etc.

Beware when looking at voltage ratings - there is a "common-mode" voltage rating, which means the whole resistor can be floating x amount of V above its surroundings and there is a "series-mode" voltage rating, which determines the maximum permissible voltage between the resistors two terminals. THIS is what I am concerned about, the series-mode rating.

The resistor you quoted has a common-mode voltage rating of 1000V, yet it is 56 Ohms and rated 1 Watt. You will immediately see that the 1000V is NOT rated from end to end of the resistor. Do the math to see what power would be disippated with 1000V across 56 Ohms!!

OK, I'll do it for you:-
1000 x 1000 / 56 = 17900 Watts.
That is NOT going suit a 1 Watt resistor, is it?
A bright flash, a loud bang, a brief puff of smoke and then ... nothing - an empty space.

Your resistor will have exceeded its power rating with only 8V across it it. It has a series-mode rating of only around 7.5V

Your 1W resistor WILL be able to handle 200V when at 100 kilohms. It will have 0.4W within it, at that voltage, within its design rating. But we don't run AT design ratings when safety is concerned, we design conservatively. So I say, run at no more than 100V across your resistor. See where I'm coming from?

Resistors cost pennies, and the extra security is worth a lot. Trust me, use 10 x 100k 0.6 watt resistors in series per kV. It just makes sense.


You are dealing with frighteningly high voltages here, Ant - are you sure you want to go that high?

I don't think you NEED to.

Graham
A picture tube on a TV uses a HI Voltage DC.
Electronic air cleaners

Hi everyone,

There are plenty of high voltage devices, but we're not in a competition to find the highest voltage device here, with maximum killing potential.

Sorry if I sound like I'm being sarcastic, I'm not. I'm being serious.

I'm gonna suggest we stick to no more than 1000V, current limited to no more than 1mA, for the purposes of these experiments.

It does the job, it's enough, it can be managed reasonably safely.

You don't need a stick of dynamite to kill a mosquito. You'll probably hurt yourself and others too, if you do.

That really concerns me.

Graham,

Thanks for continually pointing out the safety concerns. It sounds like you are moving away from DC to AC. Are you seeing better results with AC current?

Hi Marvin

AC seems to work as well as DC for ...

1. Separating glycerol from BD
2. Separating water from washed BD
3. Clearing emulsions in washed BD
4. and perhaps clumping soap too.

But DC works fine too.

So, if you have 1000V DC - fine, that'll work.
If you have 1000V AC - fine, that'll work too.

I wouldn't worry about trying to convert your AC to DC.

DC may be better with soap clumping, but I'm still not sure about that.

whoops,

i meant to pull down the 56k ones, not the 56 ohm ones. My bad.

None the less, I am totally a believer in safety, but, I am also a believer in operational reliability. 10 of anything in series is not a great idea. My preferance would be to go to a wire wound resistor at the appropriate power dissapation level. You could wire in an led (or other device) around it in parallel to verify that the wire wound was still intact. I still think limiting to 10mA is fine. It won't kill anyone, I can't think of anyone anywhere at anytime that has been killed by anything less than 15mA - and that is AC. DC lethality levels tend to be almost 2x those of AC.

Ya, again, my bad. I blame it on the flu. That's my story and i'm sticking to it.

Here's a better solution right here. Use two in series.

http://www.vishay.com/docs/30215/cw.pdf

56kOhm wire wound resistors from newark. That's the link to the datasheet.

By their formula, (P x R)^(1/2) the maximum safe working voltage is something like 28000 volts. Two in series, at 97 cents a piece is an inexpensive and very safe solution and will limit current to under 10mA (8.9 ish mA) These resistors are rated at 5W. Two from each terminal in series for each 1kV of voltage applied.

Note: higher resistances are available, but, as a minimum (because I agree with Graham), you should use at least two resistors at at least 2x the equivalent voltage rating.

soap clumping KOH soap from bio is my main aim here. And in a tank bigger than a jam jar.

The voltage needed is based on the distance between the electrodes. You reported faster results with 4KV than 2KV at a relatively small spacing. As I understand it your results mean I could find anything between 15KV and 100KV usefull depending on where I space the electrodes. 4.5KV or 9kv is still worth playing with to see what it will do. Lower is better if it works. Likewise you have reported DC works better for soap, plating the anode as well as clumping the bulk and generally working bettter than AC. You seem to be qualifying that report now but without details of why. HV rectifiers are not too tricky if you dont intend to smooth the output with capacitors and use a current limiting resitance after the rectifier. Would I still need to add a resistance to both terminals when using DC? Or to put it a better way would it hurt to add resistance to both terminals if I were using DC?

Re the wirewound resistors they say the dielectric is only good for 1000v max?? what does that relate to? The insulation?

I have my eye on six 10meg 30KV resitors which have a 24 high volts diodes with them. Came off a 50KV supply running on three phase. no real info on the diodes though. Any idea what they would be from the info given?

Might be better to just use stock components that are easily and cheaply replaced though.

I admit to learning as I go but thats how I learn most things. And I do read ahead of my actions. And ask here. I fully understand the basic concept of make sure you cant possibly touch the HV lines and limit what can happen if the impossible happens in some previously unimaginable way. Keep all non limited sections locked away in an earthed steel box supplied by an elcb that will trip the power off if you sneeze too close to the box. Not all that hard to do if you know what you have to do. I know HV tends to track and have discharge coronas if you dont submerge it in oil. So I probably will. If it tracks anyway it will trip the elcb. Maybe too sensitive to operate for more than 5 seconds at a time seems to be a possible problem but time will tell.

quote:

Originally posted by GrahamLaming:
Hi everyone,
highest voltage device here, with maximum killing potential.
That really concerns me.

Whilst gratefull for your concern and advice current kills not volts. Thats why you can get a sting of a van da graph generator and live to tell the tale. There is nothing wrong with using the volts that the condiitions indicate so long as you limit the current and stand well back anyway. Maybe I will even adapt the infra red key fob and reciver from a car central locking to turn it on and off just to be flash as well as safe.

quote:

Originally posted by GrahamLaming:
Hi Ant

0.1mA is fine in the small batch experiments.

Hi Dogma,
I'd recommend you don't go above 1mA, for safety's sake - current is the killer. Limit that to 1mA and you should be reasonably safe.

So a larger batch does consume more current?

Ant

Current should increase at a rate proportion to the increase in surface area. Current will decrease proportional to the distance between the +/- electrodes increases.

Current is not a necessity in this - just a by product. The only "current" you are interested in is the flow of the polar impurities molecules towards your bottom anode. As you increase voltage, the rate at which they will travel downward will increase.

so I can limit to 1mA as Graham suggests and all will be fine if I turn the voltage up enough to bridge the distance? I can always change to 10mA if 1mA dosn't work out in a way that suggest more would be better. No harm in being over safe. Better a light sting than a heavy one if the unconceivable actualises.

There should be no difference in performance between 1mA and 10mA. Also, that limit - is NOT the actual current that is flowing through the biodiesel mix. Its a "worst case" amount that would flow if you attached yourself directly to your high voltage source. (it would actually be less because it doesn't include the resistance that your body would have).

I suspect that the current actually flowing through the biodiesel would be in the microamp range.

I suggested 10mA vs. 1mA because it would simplify construction and reduce the number of potential failure points which would increase reliability without undully increasing the safety hazard. 10mA is still well below anything that would be considered lethal. (100mA is generally accepted as a lethal amount of current - this is for AC current, DC current most people usually consider something closer to 200mA to be lethal). I'm certainly not knocking on Graham at all, just stating that we have a small difference in opinion.

quote:

Originally posted by welder:
Hi Graham. Nice experiment. Have you considered ultrasonic seperation? Solid state piezo-electric transducers may prove equally effective, with less current draw and a higher margin of safety. Might be worth a shot. See www.sonosep.com.

Hi, A while back I had the idea of trying ultrasonics for separation. But my breif web search for transducers turned up references to their use for emolsification so I droped the idea. I guess it must be very frequency and amplitude dependant. I would love to try if I can come up with cheep enough hardware.

Tim

quote:

Originally posted by GrahamLaming:
Hi folks,

I have been experimenting with the effects of high tension electric fields on suspended soap, glycerol etc. in biodiesel, with some interesting results.


+++++++++++++++++++++++++++++++++

Hi Graham, Just found this thread. Thanks in advance for robbing me of another nights sleep.
This is good stuff. Has anyone implemented a production system yet?

I had an idea for a continuous version. Imagine a tall colum of concentrric tube electrodes. This would encourage laminar flow with breaks in the tube stack at suiatable pionts for feed and outputs. carefull design of the ports would be required and the tubes sized to achieve equal low Reynolds number. What do the coalesers used by petchem ind look like. Are they batch or cont,

I want to build one! My real job is suffering badly enough already due to my continous reactor project, now im realy in the ****.

I found some big plastic filter housings with electrode connections today ( from some electroplating app) while searching for a centrifuge. think I may be compelled to go an buy them

Tim

Tim

Hi Wes & Ant,

Wes = For lethal, please note that less than 4mA is considered the safetly limit in industry.

Let's not restrict our designs to what is easy and what is at a lethal limit. Let's design responsibly, safely and concientiously. If it needs something, it needs it. Full stop.

Fewer components = good. Agreed.

Keep the volts low and use a few resistors. But, if you increase the voltage, please use proportionally more resistors.

Most competent electronic technicians can reliably string many resistors together. It is nothing special! If someone can't do that, they shouldn't consider themselves experienced enough to experiment with such high voltages.

I suggest you don't skimp on important components just because you consider them a hassle to install reliably.

And certainly don't put an LED across a resistor, because that will defeat the object of providing current limiting.

If you use voltages much above 2kV, I believe you WILL find that you suffer from all kinds of leakage currents which will make the whole setup a nightmare to keep operating reliably.

There is so much likelihood for polar contaminants finding their way into your wiring, your insulation etc. I'd confidently predict you'll soon quit using extra high high voltages and venture down in the realms of 1000V or so.

You can successfully use lower voltage in a bigger tank by interleaving electrodes.

Same end result.
Simpler PSU.
Less components needed to make it safe.
Less tendency to arc.
Less ignition risk.

Let's work towards an easily manageable, reliable system, instead of figuring out how to train the rabid tiger of 2kV and beyond.

Hmmm I suspect interleaving close spaced electrodes to fill a big tank would provide its own reliability nightmares and logistical difficulties.

I agree it's no great trouble to use enough resistance and will stick with the 1mA design for safety's sake. I see no reason not to use larger resistors and fewer of them though. Just to save space if need be. Will it hurt anything to add them to both sides of a DC supply?

Bear in mind the 4.5Kv is in a device designed to sit in commercial premises open to the public and works safely as is. So it can be done and much of the work there is done for me. Steel case and all. Likewise the neon transformer is from a decorative neon sign designed to hang on a pub wall. Double insulated and presumed safe enough to have in a public place. So again it can be done.

I'm still a bit concerned that a plastic IBC full of bio could become a giant capacitor with enough charge to blow me through a wall.

I seem to remember you suggesting this would not happen but can't remember why. I'm tired but will re-read the thread later when I'm fresh.

I'm also interested in why you seem to now be unsure if DC is needed for soap removal. You were fairly unequivical originally.

quote:

Wes = For lethal, please note that less than 4mA is considered the safetly limit in industry.

My data is going off of US Mil Spec. I agree its important to use a reliable set of standards in an design.

quote:

Let's not restrict our designs to what is easy and what is at a lethal limit. Let's design responsibly, safely and concientiously. If it needs something, it needs it. Full stop.

Agree, but, again, depends on the standard, UL, Milspec etc. Its a question of definition. How safe is safe enough? 10mA is 1/10th of the lethal amount, 5mA is 1/20th, 50mA is 1/2 - how safe is safe enough? Should we limit current to 1 micro amp because that's even safer than 4mA? The only time I've ever heard of something under 100mA killing someone was when this moron stuck multimeter leads through the skin of his temples on either side of his brain and then measured resistance. We cannot protect against stupidity. I understand the concept of responsible design, but, you can only make something so safe.

Fewer components = good. Agreed.

quote:

Keep the volts low and use a few resistors. But, if you increase the voltage, please use proportionally more resistors.

Agree with you to a point. Totally agree that as voltage goes up, risk goes up, more weirder things tend to happen. Ideally you want the voltage to be as low as you possibly can and still get the job done. It just makes predicting what is going on so much easier. Above a certain voltage threshold, air gaps aren't air gaps anymore, insulators conduct current, etc. etc.

quote:

Most competent electronic technicians can reliably string many resistors together. It is nothing special! If someone can't do that, they shouldn't consider themselves experienced enough to experiment with such high voltages.

Agree. But, I don't think this 'argument' is necessarily about the ability of an electronics tech to solder multiple resistors together. I think what it's about - at least for me, is making sure that components that are used are rated for the use intended. Insofar as that goes, my preferance would NOT to be limiting short circuit current by resistors, my preference would be to employ a power supply that is properly spec'd and current limited for the application I was intending it to be used. i.e. if i want no more than 10mA, or 4mA or 1mA, that i should be able to control that at the source. Then, I can use a couple of resistors as my safety back up. But, in the course of the design and the construction, I'd make sure that the possibility of personnel exposure was as restricted as possible. Additionally, we've done alot of talking about redundant resistors etc. But, one thing that has NOT been considered is the insulation of the wire. Most insulation is only rated for something like 600V. I have first hand experiance with a guy who died because he tried to measure the open circuit output of a generator (approx 24000 volts) with a digital fluke multimeter. The fuse on the meter blew, but, the guy died because the insulation on those leads was only rated to something like 1000V. In this case, even 10000 resistors in series is not going to save you.

quote:

I suggest you don't skimp on important components just because you consider them a hassle to install reliably.

A point of failure is a point of failure. Most of the time resistors will fail open vs. shorting out. They tend to work alot like fuses. To me adding more ("more" is up for interpretation) than "necessary" is asking for failures post installation.

quote:

And certainly don't put an LED across a resistor, because that will defeat the object of providing current limiting.

Ya, I'll agree with you on that one - lol.

Hi folks,

Just a reminder, there are 2 main reasons for limiting the available current - shock risk and ignition of fumes risk.

Ant,
Instead of interleaving, you could use the simple close-spaced dedicated cell to process a flow of product - I think this has a lot going for it.

Now that I have had more experience with AC, I see that soap clumping is effective with AC too.
I have tried various waveforms, and a square wave has more effect than a sine wave of same RMS value and frequency. My guess is around 50%, so a 650V square wave seems about as effective as a 1kV sine wave.

I suspect this is something to do with rate of change of the field, causing a greater stress impulse within the soap lattice with square wave. But a sine wave still does the job.

You say...

quote:

the 4.5Kv is in a device designed to sit in commercial premises open to the public and works safely as is.

That device isn't intended to be used in a flammable vapour atmosphere, connected to a 1000 litre IBC, with wires running to a pair of humungous plates. Another good reason to consider the small cell approach.

Wes, which MIL spec are you referring to?

I'm using IEC1010 as my design guide, as I do in my day job. Electrical and electronic design isn't just a hobby of mine. It is my profession, and has been for over 25 years. Also specified by UL in the States, CSA in Canada. We design and build industrial electronic equipment which we are bound by law to ensure meets international safety specs, so I'm not trying to BS anyone here, I'm just telling it how it is. I'm no ego tripper, I'm just stressing some important points which should not be diminished for the sake of simplicity. It is in no one's interests to either over-engineer or under-engineer a solution. I feel I have been appropriate in my judgement of what is sensible, for this application.

You confirm exactly what I've been trying to tell you, when you say...

quote:

Above a certain voltage threshold, air gaps aren't air gaps anymore, insulators conduct current, etc. etc.

This is EXACTLY why you want to aim for 1 resistor per 100V. The 'insulation' of a resistor ceases to insulate above a certain voltage.

You also say...

quote:

my preference would be to employ a power supply that is properly spec'd and current limited for the application I was intending it to be used. i.e. if i want no more than 10mA, or 4mA or 1mA, that i should be able to control that at the source.

Me too! But we have seen mention of transformers with 15mA current rating. So, I proposed a safe way to limit that current to a more reasonable level.

You say...

quote:

Most of the time resistors will fail open vs. shorting out.

NO SIR! Not when they are subject to more than their rated terminal to terminal voltage, or if they are damp or surface-contaminated. An arc-over in a resistor REDUCES the resistor's effective resistance for an appreciable time, until burnout occurs, if the power supply has enough energy to cause full burnout. If not, the arc will persist and the available current will INCREASE.

For the 56K resistor you propose - do the math on that also - what power would that resistor have to dissipate with 1000V across it? The 1000V is NOT rated for terminal to terminal. 1000V across it WILL result in a charred mess.

With the resistor types you quoted, I'd suggest a string of 10 x 100Kilohm versions per kV. For all the reasons I've mentioned.

Wire insulation quality is certainly a fundamental design requirement. Of course, use the correct rated wire for the job, whether it's wiring a motor, lightbulb or a spark plug.