Quote from zchiotis on April 18th, 2013, 04:16 PM

...
It is just a try.!
I saw the output on my friends oscilloscope and looked pretty much stable but had some gitter (just like the Propstick you tested with PGen), that's all.!!
When I will have the oportunity I will take a better video showing the scope's output.
...
Greetings from Greece,
Zissis

I´m wondering why people often try to solve tricky problems using components they already have ...

Assuming that Stan Meyer technology is not as simple as it looks from the pictures he made it´s unlikely that i.e. a simple mains transformer found somewhere in the boxroom will do the VIC trick at all ... But that old part will give some lessons learned by applying it to an experiment example. Agreed.

... and finally each component will have to be designed as needed. Doing it the right way right from the beginning might be a benefit if not the component costs alone but the amount of time invested is taken into account.

so exchanging experiences and load sharing can be a real benefit for a forum like ours. and of course there are different states of knowledge represented.

I´m glad that you give the uC solution a try. it´s so many years now that people all over the place have designed hardware pulser solutions with long-term requirements for change. experimental setup showed their deficiencies. software solutions are smarter because for change there is no soldering necessary and no parts must be thrown away.

for me it´s obvious that for resonance analysis we should be as precise as possible because right now we have no clue about the exact effect taking place.

the PropStick in my video doesn´t show frequency jitter. as I mentioned in place it´s a 50 Hz frequency overlay from the grid caused by bad shielding / grounding that disturbed the digital trigger of the scope.

the pulse timing itself is as constant as the hardware pulser equivalent.

But I will have to redo both videos with a better setup and better audio. lessons learned :-)

Bussi04,

After seeing what you and others have done with the uC I'm sold on it.
Leaps better than the 4046 Meyer used. That circuit has a great deal of problems from the testing I have done. And when you compare it to the uC's capabilities the Meyer PLL circuit is garbage.

I have wanted to get into uC's for a while now but I'm afraid the learning curve might take away from other projects. How long would it take someone who's never worked with uC's to get a setup like yours working?

Quote from HMS-776 on April 18th, 2013, 06:21 PM

Bussi04,

After seeing what you and others have done with the uC I'm sold on it.
Leaps better than the 4046 Meyer used. That circuit has a great deal of problems from the testing I have done. And when you compare it to the uC's capabilities the Meyer PLL circuit is garbage.

I have wanted to get into uC's for a while now but I'm afraid the learning curve might take away from other projects. How long would it take someone who's never worked with uC's to get a setup like yours working?

How long ... Difficult question.
If you have basic programming knowledge straight on programming for the propeller using some code snippets from somewhere else will take a month I would say.

after initial setup of the programming environment the second day you will have your first "hello world!" program running and that is start of a steep learning curve :-)

But running into problems as those timing issues up in this thread or reaching resource limits of the system and debugging them is just another level ...

adding knowledge for modular software design and version control and advanced interrupt techniques or parallel processing will take some more time for reading and application.

making some proper design decisions related to uC hardware architecture and stuff needs some expertise but is crucial at all.

and adding some assembler stuff to your application will make your head bring forth some scattered grey hair ...

But once you have mastered the asm level of programming you will get an eternal feeling of understanding the "Soul of a New Machine" as mentioned here http://en.wikipedia.org/wiki/Tracy_Kidder

From then it´s no longer a black box but your new playmate at day and night :-)

that kiss you won´t forget :-)



fortunately you don´t have to reinvent the wheel as some others seem to want all the time.
You can buy the tricky applications plug&play and use the same hardware for some homemade entry level applications.

talking serious - if someone would have built the 3D printers somewhere here at the forum 3 years ago I would not have bought my CNC lathe and CNC mill for fair amounts of money and used them since then.

I would have bought my designer parts from someone here. I would not have replicated the printer, I would not have begged to get some free of costs.
Cheaper and easier I would have paid for it ...

For what would I have paid? For the knowledge I didn´t need to have or aggregate, for the time I spared in building, for the material used, for the benefit to use a pro device and for the profit my vendor needs to make a living. Well that´s fair enough I think.

It´s that easy, isn´t it?

[
Yes it was in this thread, but had some mistakes, and will be uploaded again soon!
Edit: Breadboard Schematic is uploaded.

Greetings from Greece,
Zissis[/quote]Thanks Zissis, did you solved the variable gate dutty problem?...hey do you think it will work with ATmega328  cip and make it from scrach like this http://www.instructables.com/id/DIY-Arduino-or-The-DIY-Duino/


HMS Stan's pll is not garbage... it is preety smart for his time ,think about 15years ago the computers were on their debut.I built the Pll and works,locks just fine with a little adjustment,but when it loks it holds that freq.As for the arduino,you can say it's fully working because i did not see the vic conected with the wfc,i saw it locks but locks on what? if the circuit is not closed?it has no coils ...etc i know this is the first step and needs testing building coils,but you can say arduino is beter than Stan's pll untill it looks and producing gas.
The arduino is simpler to built and cheaper,it remains to see how it behaves with everything conected...Keep up the good work..cheers!

Quote from adys15 on April 19th, 2013, 02:07 AM

Thanks Zissis, did you solved the variable gate dutty problem?...hey do you think it will work with ATmega328  cip and make it from scrach like this http://www.instructables.com/id/DIY-Arduino-or-The-DIY-Duino/

No, not yet. Duty cycle for gating is 50% for now.
This code will work on any arduino programmed chip.
If you have a way to program an empty ATmega328 with a arduino Bootloader and The code next, Yes it will work.

Quote from adys15 on April 19th, 2013, 02:07 AM

HMS Stan's pll is not garbage... it is preety smart for his time ,think about 15years ago the computers were on their debut.I built the Pll and works,locks just fine with a little adjustment,but when it loks it holds that freq.As for the arduino,you can say it's fully working because i did not see the vic conected with the wfc,i saw it locks but locks on what? if the circuit is not closed?it has no coils ...etc i know this is the first step and needs testing building coils,but you can say arduino is beter than Stan's pll untill it looks and producing gas.
The arduino is simpler to built and cheaper,it remains to see how it behaves with everything conected...Keep up the good work..cheers!

If notice in the video, I am having a potentiometer set to the input instead of the pickup coil. The setup will lock in a Vmax that my screwdriver will give. This is why it defenatly locks in some frequency

Quote from adys15 on April 19th, 2013, 02:07 AM

As for the arduino,you can say it's fully working because i did not see the vic conected with the wfc,i saw it locks but locks on what? if the circuit is not closed?it has no coils ...etc i know this is the first step and needs testing building coils,but you can say arduino is beter than Stan's pll untill it looks and producing gas.

It locks on maximum voltage reached at frequencies between 1000 and 18000 hz.

that´s an interesting aspect. what´s the pattern we have to look for ... ?!?  It´s the operating condition at maximum gas production and minimum temperature we will have to identify.

thinking about the voltage detection itself ... maximum voltage is not a static condition. not to miss the maximum voltage level at 18 kHz i.e. means to take measurements 5 times during pulse on condition and maybe also at pulse off condition if there is some oscillation dynamics. or you have to integrate the signal and use hardware to identify maximum threshold.  but then there is another question whether peak voltage is important or true rms. the easiest way might be integration voltage using an LC component (and a diode). and fast voltage measurement in a high voltage environment of 2000 - 40000 V is just another level ...

lots of questions unsolved yet and lots of processing.

Quote from adys15 on April 19th, 2013, 02:07 AM

HMS Stan's pll is not garbage... it is preety smart for his time ,think about 15years ago the computers were on their debut.I built the Pll and works,locks just fine with a little adjustment,but when it loks it holds that freq.

Did you get pll lock running while using gating? In 2010 with my first board "freedom circuit" (Open Source) I realized that pll lock can´t work for 4046 when gating is applied. 4046 needs a continuous pulsing feedback pulse to create the analog voltage over a capacitor for pll lock identification or you have to store last voltage at pulsing within the gating time. Bit tricky ...

Believe me I was shocked to see all those trivial hardware solutions as stan meyer replicas popping up in 2011 and 2012 with less functionality than my pulser in 2010. bang for the buck or simple, cheap and insufficient ??? reinventing the wheel and waste of time and money ... why not buy a car instead of a cart?

Talking about development cycles Stan Meyers´, woodside, icarus, russtic etc. are all 1st generation or less, freedom circuit was 2nd generation (it had duty cycle control, PLL, output voltage regulation, frequency to voltage conversion and a uC interface for remote control) and FGen now is 3rd and 4th generation. PGen with propeller can look for whatever lock in condition you want. define it and you get it :-) what are we waiting for? what shall be the breakthru condition?

first single chip microcontrollers as intel 8048 (1976) / 8051 (1980) came up at the beginning of the 80´s. those were the days I started to apply them. since 1972 I worked on intel 8080s. those 8051s would have done Stan Meyer´s job quite well in the 90´s. but of course in assembler but not using high level language as today. 200 times more code for high level programming :-( and 200 times more processing power today :-)

btw the first integrated circuits as used in Stan Meyer´s devices later called "TTL or CMOS logic" from Texas Instruments and others were used during the Apollo program between 1967 and 1972 :-) http://en.wikipedia.org/wiki/Transistor%E2%80%93transistor_logic

good old discrete circuit technology. well done 2 generations ago!
today they are a dead end story. it becomes more and more difficult to get single parts of them.

Bussi04,

I never got the 4046 pll to lock with gating. I am now using 2 uC based frequency generators which give me accuracy to .01 Hz, along with a spectrum analyzer and oscope.

Reasons why I don't like Meyer's pll:
1. As bussi04 pointed out, it will not lock with gating on
2. It will not scan over a wide frequency range without having to flip switches and adjust the pots
3. It only works to 10kHz
4. It gives false indications and locks when you adjust the pots too far.

I spent a great deal of time and money on the Meyer pll with no success.
I've given up on Meyer's version. But I'm still working with the 4046, as it does well for some things.

[/quote]Bussy you are right but as Stan said ,''all tech was in preenginering stage the all reduced on microcip size'',for the patent grant and for simplicity/testing etc....he canot spend a long time to program cips and then to find it did not work...
The pll was without gating ,with gating i did not try ,i will and let you know...Did your propeler locks and produce gas?
btw,thanks for short hystory lesson,but Stan was not stupid to make all thouse boards without knowing what you said,also at the estate Don shows one of the microcip boards...b.r.

Quote from adys15 on April 19th, 2013, 10:48 AM

Bussy you are right but as Stan said ,''all tech was in preenginering stage the all reduced on microcip size'',for the patent grant and for simplicity/testing etc....he canot spend a long time to program cips and then to find it did not work...
The pll was without gating ,with gating i did not try ,i will and let you know...Did your propeler locks and produce gas?
btw,thanks for short hystory lesson,but Stan was not stupid to make all thouse boards without knowing what you said,also at the estate Don shows one of the microcip boards...b.r.

for sufficient gas production it needs a working VIC with those features Stan described. AFAIK nobody has rebuilt that VIC the correct way up now. Nor I have. Locking into a phase coupled system works for the propeller but that´s not necessarily the condition for maximum production and energy efficiency.
If there were a way to maximize gas production by a stand alone pulse generator without additional components needed I´m perfectly sure PGen would do the job :-) But unfortunately it depends on some major additional components.

that´s the reason why I assume that simulation for the VIC would be the easiest way to make some major progress. LTSpice or Mathcad welcome :-)

as long as those parameters are not confirmed PGen is a laboratory support tool to get access to those parameters and once they are achieved it will be reprogrammed for a process control system.
you won´t have to drop your parts into the trashcan. It will be reprogrammed and working further on.
that was it´s purpose from the very beginning. up to that point PGen supports you by exact, transparent, versatile, easy to handle pulse generation at 3 to 5 independent channels and protocol features. Soon there will be a phase shift feature added to support generation of several variable phase shifted and phase locked signals to support Bob Boyce technology and sophisticated sync lock features needed for EEC / Gas Processor coupling.  as a by product that add on will give a free 3 phase digital driver for some of Stan´s sideline applications. Of course there is some LED sync pulsing support for Gas Processor and WFC already implemented. that´s a dead end road for discrete hardware implementations.
let´s take into account that pulse generation is the easiest part of the job. process control will be the more sophisticated one. think about the huge amount of components Stan had to build for his automotive control system. minor changes necessary and all that stuff were gone ...


I saw some large boards partly assembled or not assembled at all but I didn´t identify any uC components from Stans estate.

Do you have more detailed information about that?

Bussi04

Meyer talks about using a uC in News Release # 9 & 10. He states all the circuitry (pretty much everything we have) was pre-engineering and de-bugging. The whole thing would be put into a single chip, capable of 1.6Mbps at a cost of $175 USD. This was in 1992.

Below is a picture of the chip (ref News Release # 10).

Quote from Dog-One on April 12th, 2013, 09:29 AM

Closely related is JL Naudin's pulse motor Arduino controller.  Check it out at:
http://jnaudin.free.fr/dlenz/DLE23en.htm

Having these kind of digital processors in your toolbox is invaluable for energy research and testing.

It uses the USB serial for communication with a laptop since I didn't have a display on my UNO device.  If nothing else, it will give you some code examples of how I handled pulse, duty cycle and gating.  I managed to hit a wall with the limited memory I had available to go much further, but with bigger Arduinos, you can take this out as far as you want.

AWESOME WORK HERE

GREAT STUFF a lot of Knowledge
being put in use in this thread great.
=======================
T HOUGH WOULD BE GOOD
TO HAVE FAST CODE REF from the Naudin Post
SO HERE along with other code
IS TO RAISE CODER KNOWLEDGE FASTER

====================
For other code  in text file so people can grip it faster
some coders may see it and fix it fast.

Post Text  do Same  So uptake of new people is fast when try to program ,

Dan
++++++++++++++++++++++++++

Quote from HMS-776 on April 18th, 2013, 06:21 PM

Bussi04,

After seeing what you and others have done with the uC I'm sold on it.
Leaps better than the 4046 Meyer used. That circuit has a great deal of problems from the testing I have done. And when you compare it to the uC's capabilities the Meyer PLL circuit is garbage.

I have wanted to get into uC's for a while now but I'm afraid the learning curve might take away from other projects. How long would it take someone who's never worked with uC's to get a setup like yours working?

just jumping in and getting a lightspeed fast learning curve should be the usage of 12 blocks as demonstrated here https://www.youtube.com/watch?feature=player_detailpage&v=RwQwFI1vGuE#t=922s

that´s a great demonstration video!

hanno sanders, the inventor of viewport and propscope has also implemented 12 blocks.

btw. some people here at the forum do use propscope as their oscilloscope as can be seen at their youtube videos.

another video https://www.youtube.com/watch?v=JmpPm2Di0Bo

that´s really amazing ...

I'm a propscope user!

It's great, only complaint is that it's max is 200v.
Then again,  I'll just use a voltage divider.
Btw, thanks for the videos

Hello everybody,

I am very happy to announce you that Duty cycle option on gating is now available.
You will find it at first post. It's version 0.2.10.

after having some fun with the ATmega 328 datasheet, I found that arduino uses a 16MHz clock. That means that our loss of time in each timer count can't be more than 0.0625μs. I think that is a really pretty high resolution.

If I am not right somewhere, please correct me.

So I came up (after some searching) with a library that brings those timer-interrupts out, really handy.! It is counting in μseconds. And the output is more than stable.!

In the 0.2.10 zip file i have included the library. Please install to your arduino folder to work.

Changes in the circuit.
A0 is still the voltage input for locking the high frequency.
A1 is a pot, controlling the Gating frequency.
A2 is a pot, controlling the Gating Duty Cycle from 0% to 100%

use linear potentiometers (not logarithmic) so the program can function better.

Digital pin 9, outputs the high frequency tone steady at 50% D-C.
Digital pin 10, outputs the PWM gate frequency.

Please combine these two outputs (pins 9 & 10) with a couple of transistors (e.g bc547) to test.
For some reason, I am having the output of the pin 10 inverted in my circuit. You can do whatever you want.!

If you test, please feedback!
Thank you.!!

Quote from HMS-776 on April 23rd, 2013, 07:59 AM

I'm a propscope user!

It's great, only complaint is that it's max is 200v.
Then again,  I'll just use a voltage divider.
Btw, thanks for the videos

IMO you should be able to use standard high voltage probles as 1:100 or 1:1000, correct? what does your manual tell about that?

Awesome work figuring out how to control the duty cycle! I'm still waiting for parts to try this.

This might be a dumb question but would you be able to auto-tune the duty cycle and gating frequency for best performance like you do the main frequency?

Quote from zchiotis on April 23rd, 2013, 05:36 PM

Hello everybody,

I am very happy to announce you that Duty cycle option on gating is now available.
You will find it at first post. It's version 0.2.10.

after having some fun with the ATmega 328 datasheet, I found that arduino uses a 16MHz clock. That means that our loss of time in each timer count can't be more than 0.0625μs. I think that is a really pretty high resolution.

If I am not right somewhere, please correct me.

So I came up (after some searching) with a library that brings those timer-interrupts out, really handy.! It is counting in μseconds. And the output is more than stable.!

In the 0.2.10 zip file i have included the library. Please install to your arduino folder to work.

Changes in the circuit.
A0 is still the voltage input for locking the high frequency.
A1 is a pot, controlling the Gating frequency.
A2 is a pot, controlling the Gating Duty Cycle from 0% to 100%

use linear potentiometers (not logarithmic) so the program can function better.

Digital pin 9, outputs the high frequency tone steady at 50% D-C.
Digital pin 10, outputs the PWM gate frequency.

Please combine these two outputs (pins 9 & 10) with a couple of transistors (e.g bc547) to test.
For some reason, I am having the output of the pin 10 inverted in my circuit. You can do whatever you want.!

If you test, please feedback!
Thank you.!!

Good progress! Much better than your very first version.

Unfortunately I have no Arduino for test and the next month there is some lack of time.

Some ideas:

1. If I´m correct you measure feedback voltage once a ms. your pulse frequency is between 18 kHz and 500 Hz. that means that pulsing is much faster than voltage scan. so I suggest to take feedback pulse and integrate it´s voltage by using a RC component.

2. maybe a good idea to expand the program for frequency and duty cycle scale stretching. that way you can use standard pots but have frequency changed in the Hz scale and a slight movement at the pot doesn´t change frequency by 100s of Hz.

Keep on your enriching work!

Quote from zchiotis on April 23rd, 2013, 05:36 PM

Hello everybody,

I am very happy to announce you that Duty cycle option on gating is now available.
You will find it at first post. It's version 0.2.10.

after having some fun with the ATmega 328 datasheet, I found that arduino uses a 16MHz clock. That means that our loss of time in each timer count can't be more than 0.0625μs. I think that is a really pretty high resolution.

If I am not right somewhere, please correct me.

So I came up (after some searching) with a library that brings those timer-interrupts out, really handy.! It is counting in μseconds. And the output is more than stable.!

In the 0.2.10 zip file i have included the library. Please install to your arduino folder to work.

Changes in the circuit.
A0 is still the voltage input for locking the high frequency.
A1 is a pot, controlling the Gating frequency.
A2 is a pot, controlling the Gating Duty Cycle from 0% to 100%

use linear potentiometers (not logarithmic) so the program can function better.

Digital pin 9, outputs the high frequency tone steady at 50% D-C.
Digital pin 10, outputs the PWM gate frequency.

Please combine these two outputs (pins 9 & 10) with a couple of transistors (e.g bc547) to test.
For some reason, I am having the output of the pin 10 inverted in my circuit. You can do whatever you want.!

If you test, please feedback!
Thank you.!!

Verry cool nice job,thanks veryy much,so you twicked the ardiuno to generate a secound frequency(gate frq.) and then combine the 2 ,with transistors?I sugest you can combine them with 7408/cd4093...i think you know that...For the scope you could use http://www.zeitnitz.de/Christian/scope_en  it is the best sndcard scope i'v seen so far...it is 95% acurate compared to a real one and you just plug in direct to your sndcard...Cheers!

Quote from bussi04 on April 24th, 2013, 04:15 AM

Some ideas:
1. If I´m correct you measure feedback voltage once a ms. your pulse frequency is between 18 kHz and 500 Hz. that means that pulsing is much faster than voltage scan. so I suggest to take feedback pulse and integrate it´s voltage by using a RC component.

I know. You mean that the time the code reads the voltage from the transformer,
the waveform of the wave may be at zero, and that will result in a false reading indeed.
I agree!
This might be achieved by a very high pass filter.? That for frequencies below 20kHz
will stretch the waveform at all and give a straight DC to the reading?

Quote

2. maybe a good idea to expand the program for frequency and duty cycle scale stretching. that way you can use standard pots but have frequency changed in the Hz scale and a slight movement at the pot doesn´t change frequency by 100s of Hz.

You mean for example 10.7Hz??
At the moment readings from the potentiometer are mapped to 1 - 120Hz for gating.
The onboard analog to digital converter is 10bit and returns numbers from 0 to 1023 to the code. (0 is 0volts - 1023 is 5volts).
Then the reading is mapped to the said range and the steps are 1Hz since we work with integer variables.
Same is happening for duty cycle.
We can make it with double variables so the gating frequency or duty can change with 10bit steps.

Quote from adys15

...combine the 2 ,with transistors?I sugest you can combine them with 7408/cd4093...i think you know that...For the scope you could use http://www.zeitnitz.de/Christian/scope_en it is the best sndcard scope i'v seen so far...it is 95% acurate compared to a real one and you just plug in direct to your sndcard...Cheers!

Hi, thanks for your idea I will play with gates and try it.
Also thanks for the scope. I hadn't discovered that so far...

Thank you very much for your support.
I am not very good at electronics and your help and ideas are really aprecciated.

Zissis

Hello everybody,

Thanks to adys15 I have a soundcard oscilloscope!!!!!

Sorry for the poor video and low sound...

https://www.youtube.com/watch?v=t_RMh9v-8mk
It is not so good on square waves but shows something. It is very good with no money at all.!!:cool:
I used a voltage divider for the soundcard input.

Now I have to play with Logic gates to clear the square wave at all!

Greetings.!
Zissis

I'm beginning to build a setup for testing and noticed the drawing calls for 20K pots, but the parts list calls for 10K pots.  I have few 1K, one 10K and few 100K pots handy, would either of them work just as well?  I'm not worried about the LCD pot, but the other 3.  

Thanks!

Nate

Quote from firepinto on April 27th, 2013, 10:18 AM

I'm beginning to build a setup for testing and noticed the drawing calls for 20K pots, but the parts list calls for 10K pots.  I have few 1K, one 10K and few 100K pots handy, would either of them work just as well?  I'm not worried about the LCD pot, but the other 3.  

Thanks!

Nate

Hi there Nate,

10k pots should work pretty well. You can use another pot for the input for testing, before connecting it to the VIC pickup coil.

The resistor shown between the NPN and the large POT is labeled 1K, but the color code is for 220 Ohm.  Which one is correct?

Thanks,

Nate

Quote from firepinto on April 27th, 2013, 01:02 PM

The resistor shown between the NPN and the large POT is labeled 1K, but the color code is for 220 Ohm.  Which one is correct?
Thanks,
Nate

it will work either... I used 1kΩ resistor there to ground, because it lowered the noise a bit. You can bypass this, and NOT ground this output at all:)

Quote from zchiotis on April 27th, 2013, 03:02 PM

Quote from firepinto on April 27th, 2013, 01:02 PM

The resistor shown between the NPN and the large POT is labeled 1K, but the color code is for 220 Ohm.  Which one is correct?
Thanks,
Nate

it will work either... I used 1kΩ resistor there to ground, because it lowered the noise a bit. You can bypass this, and NOT ground this output at all:)

Thanks Zchiotis,
I still need to get a 4.7V Zener diode, so I'll probably just order some new pots too.  I made a video of what I have built so far, but my internet provider (CenturyLink :@ ) decided to block youtube after my upload finished. I can't get the link to post here, but you can see it if you go to my firepinto channel.  I'll link it here when I'm out of internet detention. :dodgy:  
So far I have it scanning nothing.. but at least the LCD and LEDs are working. :cool:

Nate