how can test the stator insulation and mylar coating? - Page 3 - diyAudio
Go Back   Home > Forums > Loudspeakers > Planars & Exotics

Planars & Exotics ESL's, planars, and alternative technologies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 8th April 2011, 07:01 PM   #21
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
Alex,You are absolutely correct and I do apologize for coming off a little on the offensive side.

However let me set the record straight for the third time that all was doing is to present the topology that I used to create my bias supply.

And that the schematic depicts the parts that I used as is.

I never stated that these are the parts that should be used as a final product.
And I did state that this was just a rough draft and there is much room for improvement.

I my self, the designer of the circuit, was quite amazed that the thing even worked with those particular diodes at all.

But it did, and it gets me by for now and thru the last year it has been working flawlessly.

I am sure that by replacing the diodes with the proper ones it should sustain a continuous arc at 1cm gap instead of the 4mm gap that it does do.

I do experience this drop in output voltage even when a 100 gigohm resistor divider is connect to the output to measure the voltage as observed on the scope.
I'm sure this is because of the poor recovery time of the diodes.

The easiest way to combat this is to rewind the output transformer core to operate at a much lower frequency and that will allow me to use the multiplier stage that I have already built and invested in.
As well as solving the problem of the 555 timer being barely operable at the extreme end of its frequency range.

Also being that I am experimenting with cores ripped out of old supply's they vary in sizes and material types, I don't yet have a recipe of using (Z)core and (X) amount of turns to be consistently repeated by the average DIYer.

This is why I have never post the schematic in the past and I probaly should not have done so with the circuit values displayed.

The data sheets that I have found list the recovery times as 2us for the 1N4007 and 1N5399 that I am currently using and there was one that I had found that listed it as .8us which is what inspired me to try them.
But I can't seem to locate that file now.

Although it is not an efficient way to go by any means ,It did work in a pinch.

I do have quite a few of the proper diodes but not enough to accomplish my goals at this moment.
Otherwise I would have used them and displayed their part number in the shcematic.

If one were to come out and say "I would like to use that circuit" ,Trust me I would be the first one to come out and say that those diodes are not the ones you should use and that you should use diode XXXXX for that circuit.

With Best Regards. jer
Attached Files
File Type: pdf 1N4004.pdf (183.5 KB, 20 views)
File Type: pdf 1n5391.pdf (67.3 KB, 9 views)
  Reply With Quote
Old 28th October 2011, 07:43 AM   #22
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I got a Samsung ML-1865w laser printer for $50 a couple of weeks ago and it works great for making PCBoards.

So first on my list (of course) is my variable bias supply.
I have drawn up two versions so far of the driver circuit, one with the class B output stage and one without.

I have decided to just make the 555 driver circuit so far as not to waste PCB material for now,So that I can work on lowering the frequency and work on the recipe for the transformers so that everyone can reference it.
As well as using the 1N4007 diodes more efficiently due to a lower frequency.

As it is right now the timer is set at 163khz at 50% duty cycle.

By the time I get to the final version it will have a voltage feedback circuit for too aid in regulation.
Although it has been proven very stable without it.

As well as a buffered output voltage for direct voltage measurement of the output bias voltage with a DVM or VOM at 1000v per 1v.

So here is what I have done today.

Enjoy,Jer.
Attached Images
File Type: jpg 555 DRIVER PCB.jpg (79.1 KB, 261 views)
File Type: jpg diy bias.jpg (92.3 KB, 257 views)
  Reply With Quote
Old 28th October 2011, 07:51 AM   #23
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I found an interesting page about Diode recovery times and of a tested 1N4007 as per the last time we were discussing this.
It is a great read and gives alot of insight as to why the 1N4007 happened to work in my circuit.
But as stated before the UF4007 would be better suited and cost about the same.


Diode Turn-On Time

jer

P.S. Also I found this cool spreadsheet for calculating a voltage multiplier stage as mine uses 22 stages and this spread sheet is pretty right on with measurements I got out of mine.

http://www.voltagemultipliers.com/ot...r%20Design.xls

Last edited by geraldfryjr; 28th October 2011 at 08:07 AM.
  Reply With Quote
Old 28th October 2011, 08:22 AM   #24
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
Quote:
Originally Posted by geraldfryjr View Post
I got a Samsung ML-1865w laser printer for $50 a couple of weeks ago and it works great for making PCBoards.

So first on my list (of course) is my variable bias supply.
I have drawn up two versions so far of the driver circuit, one with the class B output stage and one without.

I have decided to just make the 555 driver circuit so far as not to waste PCB material for now,So that I can work on lowering the frequency and work on the recipe for the transformers so that everyone can reference it.
As well as using the 1N4007 diodes more efficiently due to a lower frequency.

As it is right now the timer is set at 163khz at 50% duty cycle.

By the time I get to the final version it will have a voltage feedback circuit for too aid in regulation.
Although it has been proven very stable without it.

As well as a buffered output voltage for direct voltage measurement of the output bias voltage with a DVM or VOM at 1v per 1000v.

So here is what I have done today.

Enjoy,Jer.
I mean't 1v per 1000v out .he,he

jer
  Reply With Quote
Old 31st October 2011, 04:01 AM   #25
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I got my power supply back running again last night (after over a year).
With two simple mods I got it working better than the last configuration.

One was swapping the driver core to one of a higher permeability and the other was to change the 555 timer resistor connection to one resistor so that it is a 50% duty cycle although I am not sure if this is helped at all yet.

I have also been able to directly drive it from my signal generator as well.
From it, It makes no difference as to if the wave is a sine wave or a sqaure wave, either works well,but I prefer the sine wave drive better.
I will use a sine wave drive when I build the second one as I am still not getting a symmeterical output waveform using the 555 driver circuit.

Never-the-less it works fine for its intent and packs quite a wallop of a continuous arc of at least 1/4" to 3/8" and will jump nearly 1/2' when the charge is built up on the capacitors.

This is with a 14 stage multiplier and does better than when I had 24 stages hooked up when I first built it.
This is amazing for some capacitors rated at 500v from Radio Shack.

Once I get a voltage divider made and find out exactly what voltages are I may add more stages because 500v capacitors are all I have for now.
But so far no failures as of yet even with the major increase amperage performance and voltages pulling continuous arcs from it.

I have tried every different kind of core I have and I have found that most only work at about 80khz and above this is typical for powdered iron cores.
I was trying to get down to the 30khz to 40khz range but this requires far to many primary turns to make the smaller cores work properly or even at all in some cases.

I have found that the line input filters seem to work the best with little to no modification except for adding one simple winding for the gate drive and the output transformers.

I still haven't had any major issues with the diodes as regards to the reverse recovery time with my clock rates of 160khz so modifying the transformers is not an issue at this point in time.
It requires far to many turns and time to do so.

After a little research CoilCraft may have a pair of off the shelf transformers that would work that can be had for less than $5.

But if one were to try this I would go with some fast recovery diodes for good measure.

The class B amplifier stage is simple and requires two FET's and A Resistor.

It's bandwidth is from about 50khz to 300khz but it's useable performance is from about 90khz to 250khz when loaded with the multiplier stage in which this brings me to 150khz to 170khz for peak performance.

I'm getting in excess of 600v P-P to feed the multiplier stage as my scope can only display 400v P-P I will have an exact figure next time after I build the voltage dividers for measuring the high voltages.
At which time I will also try to do some power measurements as well.

I will post some more pictures after I get the rest of the thing on some PCB material and off of the breadboards as well as pictures of it in operation.
All I have left to do is the Class B power amp stage for now.


jer
  Reply With Quote
Old 2nd November 2011, 09:19 AM   #26
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I just made a 1000:1 voltage divider of 138.65 megohm resistor string and 140k ohms.
It should be accurate within A few percent or at least 1% or better.
I Did my best to verify it with my scope and a Brand new DMM which seems to be fairly good.
Although, It is not with in specs in some of its ranges but still within 2% to 3% it is a Cen-Tech P37772 that I picked up for $19.99 on sale at Harbor Freight.

I may take it back and try for a another as it is still not a bad deal as measures capacitance as well and accurate to 1pf.

So far I have measured 4.67Kv from the supply with only about 100v of ripple.
It only pulls .4 amps when it is set at full power with no arcing from a supply of 18v.
Start some arcing and the Supply current goes as high as 3 amps and pulls 4.5 amps with the multiplier stage shorted as it is about the limit for my transformer.

It chugs along quite nicely at 1 amp with some mild arcing.
So, it does use much power at all.

With a constant steady arc it still maintains about 2kv to 3kv even with the resistor string attached.
The maximum arc length is about .28" and is the longest it can do with this configuration.

This is plenty to power my ESL's but my goal is 10kv so I will have to add the extra multiplier stages that I already have made.

I will have more on this later.

jer
  Reply With Quote
Old 2nd November 2011, 08:06 PM   #27
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I was able to get more voltage out of the supply by increasing the maximum drive voltage from 10v to 13.14v.
This test was done at this level.

However increasing the drive that much causes the Fet's to run rather warm.
I did not measure the current draw from the main supply and I am sure that it was quite high from the heat that the FET's produced.
Especially for not pulling any constant arcs, But, There were no failures.

These are IRF510's and using a FET with a lower Rds such as an IRFZ42 or similar this may not be an issue.

Using the lower drive voltage the IRF510's hardly get warm at all ,and ,will require a very minimal heatsink area.

Even though I was able to get close to an extra 1KV out of it doesn't justify the higher voltage drive as two more stages of the multiplier can do this without causing excessive current and heating of the FET's.

Due to the heating of the FET's it seemed to also cut down on the available current required to maintain a constant arc as well.
Even though maintaining a constant arc is not the goal it is a good test of performance and durability nevertheless.

I also made I slight change of the FET's bias configuration and I had a slight improvement with that and I will post a revised schematic when it is finished and off of the breadboard.

The one reading of 4.85v from the meter was thru a 1megohm resistor off of the divider and seemed to coincided with the reading of the scope waveform with the 10x setting on the probe and not going thru the 1megohm resistor.......hmmmm.

This is were I question the accuracy of the meter a little,But this discrepency is with the readings on a couple of ranges not being exactly the same with just the decimal point changing place.

But after a few more tests showed that my voltage divider is exactly 1000:1 and with in about .1% error if not better.
good enough for this application any how.

Here are the pictures of today's tests.

enjoy!
jer

P.S. the scale on the ruler is in 1/10".
Attached Images
File Type: jpg BIAS SUPPLY SETUP.jpg (125.8 KB, 217 views)
File Type: jpg ARC.jpg (78.5 KB, 160 views)
File Type: jpg 5Kv OUT CROSSHAIR IS 0V.jpg (69.8 KB, 155 views)
File Type: jpg OUTPUT OF DIVIDER.jpg (52.4 KB, 19 views)
File Type: jpg OUTPUT OF DIVIDER THRU 1 MEGOHM RESISTOR.jpg (59.0 KB, 13 views)
File Type: jpg OUTPUT WAVEFORM THRU 10X PROBE.jpg (71.9 KB, 15 views)
File Type: jpg GATE DRIVE THRU 10X PROBE.jpg (459.6 KB, 15 views)
File Type: jpg 555 DRIVER OUT 1X PROBE.jpg (64.8 KB, 15 views)
File Type: jpg FET SOURCE OUTPUT 1X PROBE.jpg (70.7 KB, 18 views)

Last edited by geraldfryjr; 2nd November 2011 at 08:31 PM.
  Reply With Quote
Old 2nd November 2011, 09:28 PM   #28
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
Here are some pictures of arcing in action and a shot of the output voltage waveform while it is arcing.
Sorry for them being a little blurry, If I do anymore I will take the time to make them as sharp as possible.

Again the scope setting is 1Kv per division and the cursors are crosshair 0v and the X is 4.32Kv

Enjoy !!!

jer
Attached Images
File Type: jpg P1010011.jpg (72.9 KB, 36 views)
File Type: jpg P1010012.jpg (74.0 KB, 35 views)
File Type: jpg P1010027.jpg (47.5 KB, 38 views)
File Type: jpg P1010029.jpg (57.9 KB, 32 views)
File Type: jpg P1010030.jpg (72.2 KB, 27 views)
  Reply With Quote
Old 4th November 2011, 07:05 AM   #29
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I was able to clean up the gate drive signal today with a 100 ohm resistors in series with the gate lead today.
Doing this got rid of the ringing and cleaned it up a lot as well as allowing me to raise the gate drive supply to 15v.
I hardly get any interference lines at all on my TV now.

This is a very good thing and the FET's run much cooler than they did.

I had got a full 5kv out of it with 300v of ripple and it was running reliably and fairly cool at full tilt.

Until I got to poking around with a cap on one of the gates and popped an FET.
I ended going thru 3 or 4 Before I got it back running again.

If a gate becomes ungrounded from its pull down resistor this causes the FET to turn full on therefore blowing it, As there is only the resistance of two turns of wire as a load for each FET.
A short circuit at DC if you will.

This is probably because my solderless breadboard is old and worn out and this won't be an issue once everything is mounted and soldered together.

When I first designed this circuit I was using a single IRFZ42 and this happened all of the time only that they were so durable that all that would happen is the current would melt the solder connection to the FET and the supply wire would pop off saving the fet.
It eventualy wore out and it failed too due to heat, which is what forced me to use an IFR510 ,because I had lots.
The IRF510's work but good high current switching FET's are definitely better for this application.

Once I got back to square one every thing worked flawlessly and I ran it for several hours with a gate drive voltage of 15v and an output of 4.7KV.

Also I took the FETS of off my big heatsink and mounted them to a .080" X 24 square inch piece of aluminium plate and did some temp readings.

At full on it held at a cozy 85 degrees Celsius for 2 1/2 hours.
At 4kv about 65 to 70 degrees Celsius.
And below that 50 degrees Celsius and cooler as the FET's were just warm to the touch.

So with the second multiplier stage I will easily be able to get back to the 5kv to 7.5kv range to run my ESL without causing any extra stress or heat on the fets.

I will build another one to use for testing coatings with better and heavier duty parts.
Although this one will suffice for now.

Here is a picture of the new Gate Drive Signal full on.

jer
Attached Images
File Type: jpg CLEANED GATE DRIVE.jpg (83.3 KB, 16 views)
  Reply With Quote
Old 4th November 2011, 08:14 PM   #30
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
I am very happy with this configuration so far.
It only draws .45 amps when it is full on and no more than 1.8 to 2 amps drawing a small gaped continuous arc and around 1 amp for mild arcing.
It does draw 3.45 amps when the multiplier stage is shorted this is much better than the way it was when I first started.

I'm sure that it would do fine on a 10 to 20 watt wall transformer.
right now it is running on a 18v supply from a 25.2Vct 2amp transformer found at Radio Shack.
In fact all of the parts came from there except the switching transformers as they are just common line chokes torn out of some old PC supplies.

I don't have enough resistors at the moment to pull some power output tests.
But, I will do this on the next one as this is mainly to bias my ESL's.

Knowing me I like to push things beyond there limits and so far this is the maximum that I can get out this circuit and has proven to be very reliable so far the way it is.

Besides blowing a few FET's in the devolepment process this power supply is made compleatly of junk box parts and is quite versatile.

Being that it is variable from 0v to 4kv or more.
It has plenty of current to not drag down the output voltage if a breakdown and arc should occur.
It can also be seen on a scope exactly what the voltage is, when this breakdown happens.
This is perfect for testing high resistance coatings and such as it also has an output for direct reading of the output voltage with a common DMM at 1v per 1000v.

I will now get on with finishing the construction of this project so that I can start doing some real world tests.

jer

Last edited by geraldfryjr; 4th November 2011 at 08:17 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Role of ESL stator insulation? Few Planars & Exotics 22 8th November 2009 12:30 AM
Testing Stator Insulation bshaw147 Planars & Exotics 7 28th August 2008 09:05 AM
High reliability stator coating j beede Planars & Exotics 2 18th January 2008 09:08 PM
Stator coating Bazukaz Planars & Exotics 9 26th January 2006 08:06 AM
DIY Electrostatic Stator insulation schpeltor Planars & Exotics 7 13th February 2002 07:11 PM


New To Site? Need Help?

All times are GMT. The time now is 08:43 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2