Go Down

Topic: reading voltage of an N channel mosfet (Read 3104 times) previous topic - next topic

allanhurst

#90
May 07, 2018, 05:35 am Last Edit: May 07, 2018, 05:45 am by allanhurst
Hooray - signs of life.


Yes - tants are polarised.

If you're trying to burn a hole I'd say a fluid flow would be a good idea.

Bubbles... implies electrolysis. You may be doing a bit of erosion as well.  Try a non-polar fluid such as paraffin. The eroded particles may react with water to give an electrolyte.

Keep us posted.

Small revision - the resistor driving the TL431 - R5 - should be a bit lower - say 2k2. Not critical.

Allan

mtraven

i put a fan blowing on the water to make some turbulence...it helps, but probably needs more...I am working out how I want to set that up.  The h20 I am using is reverse osmosis water, it tests under 20ppm, I thought that'd be good enough, but I have my doubts too.  I thought paraffin was a wax, and it is, but a quick search revealed to be what we in the states call kerosene--its sold primarily as fuel for heaters...is that what you are referring to? my first impression is that I'm gonna start a fire, no?  I'll try try it, just with a fire extinguisher nearby.  the dielectrics used in commercial ops seem to be trade secrets...so any other suggestions would be helpful.

it seems most small/micro hole spark erosion machines rotate the electrode.   I think I can make a spindle to do it, but in these tiny sizes, it gets a bit tricky.  How critical do you think that is to the process?




mtraven

#92
May 07, 2018, 06:33 am Last Edit: May 07, 2018, 06:43 am by mtraven
ps I've had an idea for a better spark sustain, but leave it till you've got the basics going.
i think I'm ready to hear about that now please?

I have a lot(like 10gallons) of lightly used veg oil...might that be worth trying?  Probably need to filter it well...but I am also thinking the circulation system should have a filter too.

allanhurst

#93
May 07, 2018, 11:33 am Last Edit: May 08, 2018, 12:50 am by allanhurst
Yes - kerosene is what I mean. I don't think anything more viscous would be a good idea - it's got to scour a very tiny gap.   It has quite a high flash point, so I wouldn't worry about fire.

Petrol ( gas to you) would be a big fire risk.  Perhaps white spirit or turpentine?

And plenty of flow- more than a fan. Get a pump. Think lathe coolant.


Rotation sounds a good idea - mechanically possible?  Better have no runout....

Enclosed spark idea. I've never done this, so you're my experimenter!  Inductor and shottky should be >> 1A rating.

The point is that when the spark ends the inductor tries to keep the current going.  Might or not do any good. Old dc carbon brazing torches and modern MIG welders do this .

Allan

mtraven

well I did try the veg oil last night...it did make a big difference.  A) I have ZERO volts all the way up to right before contact....in the water, I could detect the water and it gradually rose as I approached.  This was simpler to code, but also evidence of electrolysis.  It was actually sparking, although now the particles don't clear that tiny gap (seriously is soooo small) and it shorts on them after a few sparks...you can actually see the particle glowing.  So next on my list is a jet to clear those particles...but that's going to require some re-engineering of the setup.  I have not yet but want to try some kerosene, the thinner liquid will, as you said, do a better job of clearing a very small gap and also be much easier to move (ie a pump I have on hand would do it).

re: rotation
So i choose not to try & rotate because of runout, there is no such thing as 0 runout, so rotation will always make a bigger hole...I did a bit of studying commercial setups last night...they tend to grab the electrode (usually brass/copper wire) with a collet way high up (1-3ft above work piece) rotation is done there.  then it basically hangs down and rotates through an eyelet/guide pretty close to part.  By doing this, they minimize the effect of spindle runout.  The problem is making the collet & guide requires make the tiny holes that I am building this machine to be able to make..bit of a catch22.

I found a fun video in which two experts were challenged to make a hole through a 0.5mm pencil lead...one on a lathe and the other via edm. https://www.youtube.com/watch?v=mQAkJ-L1iVw



allanhurst

#95
May 08, 2018, 12:39 am Last Edit: May 08, 2018, 03:27 am by allanhurst
Your eyelet needn't be a hole - how about a shallow V scribed with a horizontal mill with a sharp tool against a flat surface..

Or my squeezed hole idea. Even an elastomer might do. Two in series?

Or a taper collet.

And do try a very thin fast moving clearing fluid. Your part used veg oil may also have tiny particles which could clog things.  Filter?  Also applicable to any working fluid - there will be eroded particles after a while..

You might try de-ionised water - as used for topping up car batteries. Most garages/ motorists shops stock it. Reverse osmosis H2O is nowhere near pure enough. But may degrade quickly with use - see previous post.

How about piano wire - high tensile steel - as the electrode? Very stiff and strong.

Allan

mtraven

apologize for my absence....I came down with a nasty fever that had me out of commission for nearly a week and then I've been playing catch up.

most of what I have done since the last report has been mechanical--reworking the electrode system and the flushing system.

electrode:  I had gotten the pencil lead to arc with a fair bit of consistency--didn't seem to burn much metal.  Modifying mechanical pencils was frustrating & the electrode size is still too big and not very flexible. Also these small "leads" are "super high polymer"...a little research revealed in the small "leads" the normal clay binder in replaced with some polymer for strength...not sure what the implications of this are as they are still conductive...all this lead me to give up on the leads.   So I tried to turn down the tip of 3/16" brass rod, was able to get down to 0.52 mm...it was enough proof of concept to move on to wire.  I spent some time aligning the tail stock on my lathe and was able to push some tiny carbide drills through fiberglass rods to make guides (Allan:  not familiar with your squeezed hole method, fill me in please). 

so now I have it running with 28 ga wire(actually enamel wire).  That brought up a new problem, my dielectric jet was far too powerful and displaced the tiny, flexible, wire.  So I added a second jet, symmetrically apposed to the first and a diverter valve to allow me some control over the flow rate.  This system works very well now...still using the veg oil which seems to work well.

my biggest problem right now is positioning / sensing related.  The table is working way too much pecking in & out between short & open..so I get a few sparks when the gap is right on the way down & again on the way up, but its not real stable at the moment.  With the wire electrode the  spark produced is blue (previously yellow / amber) and it cracks like a whip....both characteristics I have seen in other edm machines, but never with mine.  I take that to be a good sign.  Even with the very inconsistent sparks, it did remove a bit of metal.

I am going to spend some time playing with the control software tonight.  Allan had given me a moving average algorithm-- could you check that it is correct?  I have tried implementing it twice and it doesn't seem to work for me.

outsider

#97
May 21, 2018, 05:39 am Last Edit: May 21, 2018, 05:41 am by outsider
FWIW, the "disintegraters" I've used (for removing broken taps, drills, etc.) used an axially vibrating electrode holder that slowly lowered the electrode to the work piece, (the electrode never actually touched the work piece), the fluid (DI water) was pumped through the hollow electrode and kept the electrode cool as well as flushing the particles out of the hole.

mtraven

#98
May 21, 2018, 07:27 am Last Edit: May 21, 2018, 07:40 am by mtraven
FWIW, the "disintegraters" I've used (for removing broken taps, drills, etc.) used an axially vibrating electrode holder that slowly lowered the electrode to the work piece, (the electrode never actually touched the work piece), the fluid (DI water) was pumped through the hollow electrode and kept the electrode cool as well as flushing the particles out of the hole.
yes, from what I have seem ultrasonic vibration of the electrode is all the rage now.  But the work I need to do do is much smaller (<0.4mm) and at that level, the commercial units use wire to plunge as well.

allan--no need to check your algorithm, my mistake was in variable types (floats & ints not getting along)...its working now, gotta work on how I use that value. 

also your point on RO water was a bit eye opening(i always took pride in my RO system)...to put some numbers behind it:  I read DI water has a resistivity around 18 MegaOhms/cm.  Early on I took a resistance measurement from one end of the tank to the other, this was a baseline to compare later to determine if the water was too dirty.  Those baselines were around 160 kOhms and the tank is 30cm (probe to probe).  That's 4800 KOhm-cm || 4.8 Mohm-cm  (or   did I get that backwards? and its 5.33 KOhms-cm?).  Even in the 4.8mOhm case, thats a long way from 18.

allanhurst

Progress again....

      DI water is MUCH higher resistivity than RO - . And both have far lower viscosity than veg oil, and  would clean your work much better. As would kerosene.

I'll knock up a crude drawing of my squashed hole idea.

What about piano wire ? stiff and strong... and obtainable in very fine gauges.

Allan

mtraven

Progress again....

      DI water is MUCH higher resistivity than RO - . And both have far lower viscosity than veg oil, and  would clean your work much better. As would kerosene.

I'll knock up a crude drawing of my squashed hole idea.

What about piano wire ? stiff and strong... and obtainable in very fine gauges.

Allan
piano wire--you are right about rigidity--and a local place stocks it, I was looking at it the other day and passed...my mind was just on a different mission...perhaps I will give it a try.  I do really like the idea of the enamel wire, provides some insulation if I want to make some deep holes.

dielectric-i have spent an exorbitant amount of time on my oil pump setup...its an oil pump (intended for an engine) that I'm running with a 12vdc motor...point is, I don't know how it will handle water (no lubrication).  Now kerosene might be ok, it has a bit of lubrication value...I just couldn't deal with my shop smelling like kerosene.  Also I am beginning to think the lowest resistance DI is not what I want.  It seems dielectrics really need to be matched to the machine, specifically the voltage.  Mine being on the low end (12v), may mean I need a lesser dielectric.  I have even heard of particles being added to the DI to decrease its DI properties.  Also I am not seeing any pre mature current flow (like I was with the RO h20), I get nothing and then just about immediate short.  I was wondering what kind of mechanical forces are involved in a spark...is a tiny magnetic field produced that pulls the wire to the workpeice? or just pushes it off center?

allanhurst

There's no magnetic field, so should be no such force deflecting the electrode ( Fleming's rule ).

The spark may cause a momentary generation of vapour ( quickly quenched) which may deflect it.

Dunno.


Allan

mtraven

I was looking over the last circuit you posted (with the inductor), I would like to try it..couple of questions:  I don't have any "normal" schottkys around, but I do have some schottyk power rectifiers, would that get the job done?  And could you explain its purpose? To me, its seems like it would just bypass the gap and  short the drain to +12v?  I am sure I have that wrong....also, how critical is the size of L1? I have no idea how to measure inductance.

gap size  & voltage:
I have it tuned to a point where removing material from the gap is the limiting factors (i hear ya on using  a thinner DE, have not yet, but will be trying some stuff tonight).  But running  a higher voltage should jump a bigger gap, bigger gap = easier to clear (give same size particle removed).  That's the theory, but if I am going to try it, I need to be able to go back easily.
 I have 24v sitting right there, considering what mods would need to be made to accommodate:
-12v linear reg inline @ Vin(arduino)[drop 24v supply to 12v, let arduino Vin regulator do the rest like it is now)
-roughly double the voltage divider multiplier
-i think some of the R vals in the current control might need changing?


on dielectrics:  i remember choosing RO initially because DIY guys had done EDM in tap water...and while I now know RO is orders more conductive than distilled or DI, I wonder if it might be good enough.  I never did try RO with the jetted fluid system...remember you laughed at my fan blowing on the stagnate water?....i get it now.  But that might be worth revisiting.

allanhurst

#103
May 23, 2018, 01:26 am Last Edit: May 23, 2018, 01:46 am by allanhurst
1/ My suggestion of an inductor/ shottky is for the reasons given. Inductor value a guess - you can buy various values cheaply from assorted  suppliers. I bought 10 x 100uH toroidal inductors rated at 5A from China recently for £1.

2/ Your power shottky should be fine.

3/ My circuit should be OK at +24 with no changes except to the dividers feeding the arduino  analog inputs - ratio 4:1 giving a 5:1 division ratio should be OK. The LM358/324 will stand a +36 psu.

4/  Arduino power not > 12v - given it's not taking much power and isn't critical a series 15v 1/2 watt zener from the +24 would probably be OK. Or any regulator. Or you could use an unused section of your opamp  - just add a 3:1 divider in a classic non-inverting amplifier configuration with a NPN pass transistor from the existing 2.5v regulator   - giving  10v.

Awaiting your trials with a thinner dielectric fluid with interest.

ECM ( I presume you mean?) is a different ballgame - for that you WANT a conductive fluid. Salty water?

ECM = electrochemical ie electrolysis.
EDM = electrical discharge ie sparks.

I think my design would work with either. One advantage of ECM is that particles aren't generated.
And I suspect that the working voltages are much lower.


Allan

mtraven

no i meant EDM in tap water, but the ecm seems interesting too.

I made one attempt at a thinner fluid, I thinned the veg oil with isopropyl alcohol.  Though it made  a much thinner fuild & pumped well through my existing system, I was not able to see any sparks.  The iso made the fluid very clowdy, so its possible I just wasn't seeing it.  The meters were bouncing the way they do when its sparking in veg oil...so I am not sure what to make of that.  Any of the other DE fluids will require a different pump...I have some other stuff to try before that.

I added the inductor last night, it really seemed to make a difference...at very least it makes the voltage readings much more stable, making positioning easier.

I am going to try running this up to 24v tonight, maybe look for a new pump tomorrow, if I could get this running on distilled h20 that would be ideal.

Go Up