Page 12 of 16

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Fri Dec 02, 2016 10:44 pm
by Andrew Seltzman
The over voltage light should come on at the end of the test when voltage set is set to 0V and the output voltage is still decaying(due to it's RC time constant, and no load resistor). This is normal operation if the over voltage light only comes on at the end. Before that the output voltage should be tracking the set voltage on the graph.

Does it make a graph?

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Fri Dec 02, 2016 11:18 pm
by Jim Kovalchick
A single blank graph with no plot briefly appears. Then two blank graphs appear together. The graphs all disappear and an error message is displayed. "Vectors must be the same length. Error in => Spellman_DXM_control.m at line 137"

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 2:44 am
by Liam David
I ran the functioning power supply with my fusor today and had some problems. I could turn the voltage up to 27kV, but plasma would not establish at any voltage around 15mtorr. Turning the voltage above 27kV got plasma to establish for a fraction of a second, but the over current light came on immediately and switched off the power. I use a 10-turn potentiometer to control the voltage and have the current set point at the max, but the over current light comes on regardless of how finely I turn the knob. Does anyone have any suggestions? I use a 6-way 2.75 conflat chamber, so Paschen's Law gives different values for my smaller chamber than the standard 6" chamber, but I've heard that other people with smaller chambers have had few problems in establishing plasma.

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 4:09 am
by John Futter
What value is your series limiting resistor in the HV feed???

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 4:41 am
by Liam David
It's a 60k resistor, and operating it with vs. without the resistor seems to have very little effect.

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 4:41 am
by Liam David
It's a 60k resistor, and operating it with vs. without the resistor seems to have very little effect.

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 5:47 am
by John Futter
OK 60K ballast resisator is a bit on the low side
When your plasma establishes it is better than a dead short so 27kV and 60K ohms means that the plasma is trying to take 0.45amps from your supply

i think that you will find that you will need to increase that resistor to keep from tripping the Spellman internals

try doubling it first if you can 120K
if it still does it double it again 250K
also make sure to limit capacitance from your Ballast resistor to fusor ie no coax (33pf /ft)

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 9:05 am
by Richard Hull
This is what I harped on in the limiting value of only 8.56ma on this supply in a posting above in this thread. 60k ohms is fine for a linear supply that has some beef in it. Raising the resistor value might do it, but the working dissapation will go up. The voltage drop would be not important, of course at most any value of ballast, but the dissapation would be.

Lets go figure........ assume you will need virtually all of that 8ma in a running fusor situation. Let's talk 7ma for starters on a working fusor. This will run continuously through, let us say, a 150k ballast.

7ma x 150k = 1050 volts across the resistor...No big deal voltage wise. .007 X1050 = 7-10 watts a good safety margin would demand a 25 watt resistor (would just get warm)
a 300k resistor would double that to about 50 watts. Such resistors are still sold. Just be aware that even at 7ma a large working ballast will dissapate.

This is the terror of a fast safety current trip on a modern switcher that is kissing the current limit for fusor operation to start with. This problem is probably just and operational one. Where there is a will there is a way. If I owned one of these supplies I would shunt the current sense resistor with a chunk of 12 gauge wire and be done with it! You would then become the slower acting human meat-engine trip circuit if something goes horribly wrong at the fusor load end. I have no idea how inherently weak or easily damaged the HV electronics would be to a 1 second half amp gas arc short.

Richard Hull

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Sat Dec 03, 2016 4:38 pm
by Stefan Gustafsson
My supply is now fixed and up and running.
I had to replace:

RV2 - CL70 NTC Thermistor, burned
C149 - 33uf capacitor - visual leak
CR34 - MUR460 diode, shorted
Q10 - Mosfet, shorted, replaced with IPW65R110CFDA
Fuses- Two new 10A since I run it on 220V

I don't have a high voltage divider yet and can't verify it's output, but I can get it to fizzle of corona and arc if I turn it up too high.

Re: Spellman DXM70N600X3547 power supplies on ebay -70kV 8.5

Posted: Thu Dec 29, 2016 3:56 pm
by Jim Kovalchick
I was not able to get HV out of mine even when applying manual control. I disassembled and closely inspected the board again. My board was pretty clean with no apparent over heating or other visible abnormalities except the top of C149 was slightly bulged so I replaced it. I wish I had a video of the power up. The fan and lights started normally but about two seconds later there was a loud pop and a smokey mist poured out of the vent slots. The new C149 had failed. There is no other damage. Any ideas?