View Single Post
 
Old 10-22-2009, 04:10 AM
FuzzyTomCat's Avatar
FuzzyTomCat FuzzyTomCat is offline
Silver Member
 
Join Date: Jun 2008
Posts: 620
Send a message via Skype™ to FuzzyTomCat
Quote:
Originally Posted by ashtweth View Post
Hi Glen/Harvey/Rosie/Aaron /ALL. Well time for some Action then Luc . Glen it really has helped replicators my friend. Okay we finished a few tests and built your resistor. Went with Luc's/Alex's board lay out MAKES A DIFFERENCE, Also Glens rating inductor makes a difference got better performance 37 volts back spikes Still working through it guys, want to make sure we get what glen is doing and then can start the run time tests. Okay bare with us, here is Andrew's report.

Hey all,

I've been doing some more tests on the Ainslie circuit (negative dominant August 26 2009) and done a replication of Glen's / Quantum 10 ohm resistor (32mm diameter, 48 turns spaced 1mm, 150mm long). I used resistance wire of 6 ohms per meter (3 x 5m lengths of 30 ohms in parallel = 10 ohms).
<Glen's resistor replication - windings.jpg>
Imageshack - glensresistorreplicatioy
<Glen's resistor replication.jpg>
Imageshack - glensresistorreplicatio

I changed the series gate resistance to 5.5 ohms and tried out the insertion of ferrite rods into the test resistors (the green one bought off the shelf, and Glen's).
<resistor with ferrite rod.jpg>
Imageshack - resistorwithferriterod

Sure enough, the back spikes collected off the resistor / inductors improved - I then tuned the 555 timer and the resistor tests improved again. Our replication of Glen's resistor performed better than the green one bought off the shelf.

I re oriented the mosfet so it was less likely to pick up any stray RF energy. Should I put a resistor across the gate and (source) input negative to eliminate the RF energy if there is still any left over?
<mosfet re oriented.jpg>
Imageshack - mosfetreoriented

I used two 12v lead acid batteries to power the circuit.

Results (in the order of the tests conducted):
Test 1:
Input - 24.6v, 900mA
Output - 38v across 470 ohm resistor / 4700uF 50v capacitor

Test 2:
Input - 24.4v, 1050mA
Output - 28v across 235 ohm resistor (2 x 470 ohm) / 4700uF 50v capacitor

Test 3:
Input - 24.5v, 750mA (18.375 watts of power)
Output - 37v across 235 ohm resistor (2 x 470 ohm) / 4700uF 50v capacitor (5.8 watts of power)

Temperature readings (final test):
Ambient - 26C / 80F
Mosfet - 53C / 128F
10 ohm resistor - 53C / 128F (this may be the reading because the heat has radiated out over a bigger surface area being the black silicone?)
555 timer - 68C / 155F
235 ohm 20W resistor - 105C / 222F

I did have trouble with the 555 timer as per the suggestion of adding some 4K ohms worth of resistance in series between the +12v and the power input (+) of the 555 timer. The 555 timer would not work unless the resistance was set to at or near zero (so full power was reaching the 555). I used a 25 turn 5K ohm variable resistor. The 555 chip gets really hot (see above).

Andrew
Hi Ash and Andrew,

It looks as if your getting very close just as I did at first with this small but complex circuit .... well done my friends

The 555 timer does look a little warm but if your following the revised August 12, 2009 Rosemary Ainslie COP>17 Heater Circuit there are several changes that I have made that makes a big difference or will as soon as the equipment is possibly available again for me.

The added 4k ohms at the 555 positive rail was only for the "Ainslie - Murakami Negitive Dominant Waveform Generator Circuit" and is not used for the "Rosemary Ainslie COP>17 Heater Circuit" sorry for any confusion here with two circuits ......

1) The Gate pot was a 5k ohm 10-turn and now needs to be a 10 ohm multi turn potentiometer (15-25 turn ) so that using a DMM or equivalent be able to adjust to 6.0 to 5.0 using the 32mm prototype "Quantum" load resistor.

A digital Multi Meter or bench top unit able to see 00.00 type ohms not just tenths but hundreds to get the sweet spot of 5.80 to 5.30 ohms (checking pot back lash) and using a second DMM monitoring the highest voltage reading from your 24 volt battery bank while doing the gate pot "fine adjustments"

2) The 555 off time potentiometer was a 10K ohm 10-turn needs to be a 15K ohm 10-turn to get the 3% duty cycle for the other testing at a lower duty cycle ..... right now the lowest is about 22% at full resistance at 10K

3) The 12 volt diode at the positive rail unmarked is a 1N4007

These are the only changes that has been done and are much closer to the original October 2002 "Quantum" Article and when I have time I will issue a new revised schematic with these and any other changes needed.

From the looks of things you appear to be very very close to some excellent results with your load resistor ..... these items the 32mm resistor and the 10 ohm gate pot are very important items that appear to be the key to success for the results you want.

The PDF ..... what great work strange to have my name attached to such a document but you and any helpers you have are doing a stand up job giving the "open source" community the information at there fingertips to reproduce this circuit

Best,
Glen
__________________
 

Last edited by FuzzyTomCat; 10-22-2009 at 04:58 AM. Reason: added "Quantum" article PDF