[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Voltage Tripler and MOT's
Original poster: "Kevin Ottalini by way of Terry Fritz <twftesla-at-qwest-dot-net>" <ottalini-at-mindspring-dot-com>
Gregg:
If you switch to a doubler or tripler your output will be DC and
although static gaps will work with DC it won't give you anywhere near the
performance that you can get with an ASRG (ASynchrous Rotary spark Gap). The
main thing you will see with a static gap is the higher the voltage the faster
the BPS rate (at least until it power arcs, argh!).
Using larger electrodes probably won't change anything, and in fact may force
you to decrease the gap even more (larger radii have higher arc voltages).
I highly recommend building an ASRG for DC TCs, they are pretty easy and really
worth the effort. Check out my web page for my ASRG pics and information.
<http://www.mindspring-dot-com/~ottalini>http://www.mindspring-dot-com/~ottalini
Since you didn't give any specific data on your MOT, let's assume it's like the
ones just posted by Godfrey (thanks Godfrey!):
Primary: 120 VAC, 60 cycles and 0.22 ohms
HV secondary: 2240 VAC and 63 ohm
Max Current: 15 amps (we'll assume 1kw).
Ok, this basically means you are running your gap at a very minimal voltage and
have to get them quite close to arc (which is exactly what you are saying).
Here are the things you can easily do:
1. Add a full-wave rectifier (4 diodes, no caps)
------------------------------------------------------------
You will get ~1.4kvDC out for every 1KVAC input so with 2.24kv in you should
get ~3.1KVDC out which will help things a little bit (don't be surprised if its
a little lower than that ... the rectifiers aren't perfect and you will
probably need a filter capacitor to capture the peaks (use filter cap for
measurement purposes only).
I wouldn't use a filter cap on the output for TC use but I would definitely use
a series resistor (5K to 20K ohms) to limit the peak current (you will blow
diodes otherwise).
Ok, there are few mysteries for the full-wave bridge, it should be able to
handle the power with no problem assuming you use a matched set of the same
kind of diodes from the microwave oven.
You can also use a simpler full wave (two diode) rectifier but there will be a
lot more ripple.
2. Build a Cockroft Doubler (8 diodes, 8 caps)
----------------------------------------------------------
Since you have a fair amount of power available we need to make sure you use
large enough caps to carry the 60Hz at the worst-case power levels.
To do this you need eight of the capacitors from microwave ovens rated for at
least 2.24KVAC.
The one I have here was 0.63uF. I'll use that for the workup so you can see
what the power limit is for this value.
What you do is build a second fullwave bridge and put it in series with the
first one, then use the capacitors (one on each AC leg) to couple the
transformer to the first bridge, then use the second set of caps to couple the
AC from the AC legs of the first bridge to the AC legs on the second bridge.
The caps are arranged this way so the max voltage seen by any cap is 2.24KV.
Please refer to the schematic on the web page for exact hookup details (look
under the DCcoil section for schematic and a picture).
The other four caps are filter caps for each stage. You use two in series for
each stage because the DC voltage from the - to the + of the bridge is 1.4X
the HVAC voltage from the MOT and you don't want to overvoltage the caps.
This entire design is known as a classical "charge pump". The first half of
the diode bridge is acting as a peak detector and the second half of the bridge
is acting as a level shifter. Each additional stage does the same thing but
has its effective DC ground reference shifted by the output DC voltage from the
pervious stage (so it really isn't really a multiplier as much as an adder).
Because the caps are now in circuit in series you have to expect some drop in
the voltage under load so expect at the most 1X the AC input per stage so
~4.5KVDC. Under no load you should still see the ~1.4x/stage or 6.3KVDC.
Again, be sure to use a limiter resistor on the output.
OK now with caps in series there is a limit to the sustained power you can draw
to charge the tank capacitor. Granted, the demand will vary dramatically
depending on the charge state of the cap but we need to make sure you don't
blow up the caps under typical heavy usage.
At 1.0kw of power, this is ~222ma at 4.5kvDC.
The simulation of the Cockroft that I use (which is quite accurate) says the
following:
Cockroft-Walton Multiplier Work Sheet
-------------------------------------
November 6,1998 K.Ottalini
-------------------------------------
Enter the number of stages:? 2
Enter the input voltage (volts):? 2240
Enter the input AC Frequency (Hz):? 60
Enter the capacitor values (Uf):? 0.63
Enter the load current (amps) or 'R':? 0.222
-------------------------------------------------------
The estimated ripple at 222 ma is: 17.6 KVolts <--Ouch!
The output voltage is: 4.48 KVolts
The estimated ripple is: 196.6% maximum <--Ouch!
The estimated Vdrop is: 41.1 KVolts <-- Ouch!
The real output voltage is: -36.6KVolts <-- No!
The percent regulation is: 459% <--No!
The power requirement for 4.48 KVolts at 222 ma is: 994.56 Watts
This is equivalent to a load resistance of 20180.18 Ohms
Well, as we can see this is not a good configuration at all. The caps at
0.63uFd are not anywhere large enough to pass these power levels reliably in a
multiplier configuration (but fine if used as a filter cap in the microwave).
It would take at least 10uFd per leg to run at these power levels and even that
gives pretty poor ripple (which probably isn't a problem at all for TC use).
Cockroft-Walton Multiplier Work Sheet
-------------------------------------
November 6,1998 K.Ottalini
-------------------------------------
Enter the number of stages:? 2
Enter the input voltage (volts):? 2240
Enter the input AC Frequency (Hz):? 60
Enter the capacitor values (Uf):? 10
Enter the load current (amps) or 'R':? 0.222
-------------------------------------------------------
The estimated ripple at 222 ma is: 1.11 KVolts <-- not good
The output voltage is: 4.48 KVolts
The estimated ripple is: 12.4% maximum <-- high
The estimated Vdrop is: 2.59 KVolts <-- Ouch Still!
The real output voltage is: 1.89 KVolts <-- Ouch Still!
The percent regulation is: 28.9% <-- not good
The power requirement for 4.48 KVolts at 222 ma is: 994.56 Watts
This is equivalent to a load resistance of 20180.18 Ohms
Since the power flow for a DCTC is very irregular, the current demand varies a
lot and what we actually see is the cap charging up to the peak voltages in
sync with the 120Hz ripple, (the higher the charge, the lower the current, the
lower the ripple, the higher the voltage).
Three stages will be even worse. In commercial designs it isn't abnormal to
use bigger caps in the first stage then smaller ones in the further stages but
here it's a moot point.
Not all types of caps can carry the forward current in this situation (the 60Hz
AC is the problem here).
Rule of Thumb:
-------------------
The rule of thumb I use is:
10ma/uFd/stage -at- 60Hz for a reasonably clean power supply (5% regulation).
With higher AC frequencies it's much much better:
2amps/uFd/stage -at- 40kHz (5% regulation)
Summary:
------------
I'm not saying that doublers won't work if you use smaller caps, what I'm
saying here is the performance and reliability is unpredictable and will depend
on the quality of the components and how much stress (and heat) they can
suffer.
I used 3ea 0.015uFd caps in parallel (so 0.045uFd) in my original DCTC 2X
multiplier that I ran at the SoCal Teslathon a couple of years ago with an
NST. Although it worked, it would eventually blow out the caps, even running
at much lower 500watt power levels.
I hope this is some help for you.
I also want to highly recommend building and adding a permanent analog HV DC
meter to the tank capacitor on your system. This serves two very important
purposes: it lets you know there is a charge on the tank capacitor (and that
the PS is working). It also acts to discharge the tank cap when the power is
removed. I use a 1 ma meter movement with a 0-20KV bezel and 20Mohms of
resistance which works great. Please refer to the web page for details on the
meter design as well.
As a caution, please be very careful with stored HV DC charges .. they can be
lethal.
Best,
Kevin
----- Original Message -----
From: "Tesla list" <<mailto:tesla-at-pupman-dot-com>tesla-at-pupman-dot-com>
To: <<mailto:tesla-at-pupman-dot-com>tesla-at-pupman-dot-com>
Sent: Wednesday, April 17, 2002 9:04 PM
Subject: Voltage Tripler and MOT's
> Original poster: "Gregg Adams
<<mailto:network-at-nexband-dot-com>network-at-nexband-dot-com>
>
> Hello all,
> Has anyone had much luck with a voltage tripler on a MOT power
> supply? If so, what size/type caps were used? The reason I ask is
> because I have to run a very narrow spark gap that tends to power arc
> quite often unlike the 15kv NST's I used to use. Opening the gap only
> causes it not to fire. I am running this gap on a fairly
> powerful "sucker" vacuum motor with brass electrodes. Would bigger
> electrodes help? I was told by a good source that I needed more caps but
> it'll be a couple of months before I receive my geek caps and I'll go
> crazy waiting that long:) So, am I on the wrong road with the voltage
> tripler or would that help my gap?
>
> Thanks much,
> Gregg Adams