Successful TPU-ECD replication !

[Feynman]

Thanks Paul

I think your plan sounds good.   If you want to go with 555s I recommend the National Semiconductor (LMC) versions from Digikey.  They are the same price and they are much more accurate chips.  Also make sure your timing capacitor is precision tolerance, because this can cause loss of accuracy. Also I wrote a program to calculate the frequency and duty cycle for the standard astable 555 circuit setup.  I posted the program and source code here: 

http://feynmanslab.blogspot.com/2008/03/555-design.html

There are other 555 circuit configurations that may be better, but I like that one for a reference design.  It's good for testing how close you are to the expected frequency, whether your 555 works, etc.

Also, if you really want to do things right, grab a used frequency counter off ebay, because the 555s tend to drift a bit depending on the tolerances of your components (the capacitor C, R1, and R2).  It's nice to know what frequency you are operating at.  This is not of course necessary, but I have found it helps me.


And last, if you find you are burning out the IRF840s,  you can try going analog by using 12AX7 (or other) tubes instead. Marco swears by the analog gear, so I must point it out as a path forward.

EDIT:

Definitely post those scope traces!  I'd be interested in seeing them.  Also if you do post, plz include the schematic and your C, R1 and R2.  Thanks!  - Feynman.

[pauldude000]

@all interested in 555 schematics...

I have two ready. Both are 50% duty cycle. However both are unique.



This  one uses a transistor and two diodes in the timing circuit to maintain 50% duty cycle. Only one pot to adjust, however the frequency range is rather limited at 50%. (around 70khz.)



This one is anything but limited. Full range of 1hz to 500khz... However duty cycle ranges from 50% to 52%. Try the transistor/diode above in this circuit maybe, for exact 50%?

I have not built this one yet, as I worked out the calculations with "555 Timer 4.01" tonight, then drew up the schematic. I'll post if I try the transistor/diode combo.

I will post the short and long duty cycle schematics as I am able, as I am not sure of the results, and want to build the circuits before posting.

Paul

[otto]

Hello all,

@Paul

500kHz is a low frequeny  (I tried this). We need a higher frequency. In the low MHz range.

Otto

[otto]

Hello all,

you all know that Im following DFROs 2 posts and saw that this 2 posts are veeeeery relevant for my work.

He mentiones to use 555 timers, mosfet drivers to make


              1 TO 4 SETS OF VERY FAST ON-TIME PULSES.....

this sets of pulses sounds familiar??? Stanly Meyer used sets of pulses. I dont remember that somewone used sets of pulses.

Otto

[giantkiller]

The www.Linear.com LT6904 goes to 68Mhz. Get the quickeval as a sample. The DC590b usb hooks to the PC and the DC726A-B. Use software to spin the freqs. It works. Ask for sample.
I have two now for free. I will get the protocol from linear and find out how to drive three from a pc.
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1010,C1784,C1096,P2416,D25439

This is, of course, if you want to play instead of work.

I got PIC and ATMEL environments coming also. $100.00 and $40.00 respectively.

--giantkiller. Gentlemen, pulse your driver stage...