[Prev][Next][Index][Thread]

Re: cap firing voltage, NST and stiff supplies



In a message dated 99-06-22 01:53:27 EDT, you write:

<< >system, such that the peak cap voltage can be made to coincide
> >with the best firing phase.  An NST system may require a 
> >compromise in the firing time.

 >John Freau
 >
 
> I tried to model an NST system with a large inductive ballast (50mH).  I
> could adjust the firing time as you say and it gives a great charging
> curve.  However, the firing voltage drops from 21kV to 8kV :-(  Looks like
> a PT or "stiff" transformer is required for this most interesting timing
> adjustment with an external ballast.

Terry, all,

That a big voltage drop you're seeing.  My reasoning was that the
cap seemed to be charging too fast in your system, so I figured more
ballast would just delay the charging, but still result in the same
peak voltage.  But I guess this only works with a stiff power supply.

Did you try other ballast values other than 50mH?  Maybe 50mH
was a little too much?  I'm only using about 26mH total, and my
cap is smaller at .0147uF.  I would think you would need only
20mH or so, including the NST leakage reactance.

> 	Of course, a stiff transformer can charge a larger cap than an NST 
anyway.
>  So there may be other things affecting it.  Could it be that you PT system
> is simply supplying higher currents and you are simply charging the cap at
> a nice looking point when, given a fixed ballast, you could get more energy
> out if you were to still fire late and charge an optimally sized primary
> cap??  By working with a fixed ballasted transformer, I get the best power
> transfer by firing late.  I would think that would hold basically true for
> any current limited transformer.  I wonder if given your ballast setting,
> you have a nice charging curve but the system is less optimal than it could 
be?

I've tried firing after the cap voltage peak with various firing phases, but
I didn't see any improvements in overall efficiency nor in power factor.
On the contrary, the overall results seem to be the best when I set the
ballast so the gap fired right at the peak of the sine wave, with the gap
phase set for a 60 to 90 degree delay relative to the input voltage phase.

I'm seeing a greater IK build up when I set the ballast so the gap fires
before the sine charging cap voltage peak is reached (still with 60 to
90 degree delay), but the efficiency and power factor are worse in 
most cases.  In one case I did seem to see a good power factor.

By *optimal* I assume you mean best charging effiency and power factor?

Or are you saying, that if I installed a huge cap, that my stiff supply 
could barely manage to charge, then I would see the best results when
firing after the cap voltage peak, the same as you are seeing?  In other
words, anytime a power supply is pushed to its max capabililties, the
best results will be obtained when the gap fires somewhat later than
the peak cap voltage.  That is possible.

Based on what you're saying, it would seem that by using an extra 
robust power supply such as I am using, this might permit static 
gaps to work better than in an NST system, since the robust
system permits the gap to fire at the peak cap voltage (with proper
ballast setting), and therefore allows the gap firing to be delayed
to the best IK realm.  Although I haven't really gotten very good
results so far with the static gaps with the PT.  The chaotic firing
of static gaps might also disrupt this plan to some degree.

Cheers,
John Freau
 
> Cheers,
 
 >	Terry >>