[Amps] ?Direct rectification of AC mains to drive the

[Jim Thomson]

Date: Mon, 23 Sep 2013 09:07:30 -0700
From: Jim Brown <jim at audiosystemsgroup.com>
To: amps at contesting.com
Subject: Re: [Amps] ?Direct rectification of AC mains to drive the
amp,

On 9/22/2013 10:10 PM, Jim Thomson wrote:
> Has anybody actually measured the PF on a transformerless supply, with 
> a big C filter ???
> I cant see it being any worse than your typ kw tube amp....with the 
> typ xfmr + FWB / Doubler +
> high C filter.

No, but the issue is not reactance, it is the fact that current is not 
sinusoidal, but rather is drawn in pulses at the peaks of the cycle to 
charge that cap. Back when most of us went to school PF was simply 
related to the phase angle between the sine wave of voltage and the sine 
wave of current. Now, the definition of PF has been expanded to include 
that impulsive current.

Also the current is impulsive,so it is rich in harmonics, and "triplen" 
harmonics add in the neutral of 3-phase systems. Because the loads on 
electrical systems are now dominated by equipment of all sorts with 
capacitor-input filters, and triplen harmonics add in the neutral of 
3-phase systems, it has become all too common for currents in the 
neutral of 3-phase systems to be equal to or greater than the phase 
currents. Remember "the towering inferno?? The cause was overheating 
neutrals, and flames were burning insulation that spread up raceways.

In EU, power factor is regulated by electrical codes. In North America 
it is not, although I think I've heard that power companies may offer 
discount rates for maintaining a better PF.

73, Jim K9YC.

##  This is all old news.  Typ ham tube type amplifiers with high C filters are the
original  pulse train power supplys.    Its easy to see these high level  pulses  every
8.3 msecs,  using ducans  Psud software.   Or with a scope across a small fixed resistance.  

##  Im well aware of the excessive high neutral current in the towering inferno movie....true story btw. 
The neutral current typ runs at 2.1 x  normal.   However, this is ONLY when you use 120 vac single
phase loads  on one leg of a 3 phase system.    IE:  your typ  208/120 v   3 phase power fed to
condos and apartment buildings..and  commercial buildings.   In one recent study, in seattle, wash,
they had hundreds of pcs..all running on 120 vac single phase, on each of 4 x floors.   From the street
it was 14.4 kv 3 phase... dropping down to 600 v  3 phase...with the usual vault xfmr.  Then a 600v
3 phase riser  cable  going to all 4 x floors.    Each floor  had its own  600 V down to  208/120  step
down xfmr... a 50 kva unit.   The load on each xfmr was only 38 kva, but the xfmrs  were ready to burn up. 

##  The normal fix  is to replace the 50 kva xfmr with a larger 75 kva unit.  The real fix these days is to use\the newer
style amphorus xfmrs..which have a 3rd winding in them..which eats the  harmonics.   Then you can easily get away with
a 50 kva xfmr. 

##  IF EU  style  PF  regulations  were ever imposed on ham gear,.. there is NO way you could ever use the typ
HV supply config, with the FWB /doubler +  high C filter.   You can easily see the ODD harmonics  on any of this
stuff, using the  HP harmonic analyzer, that’s  made for this purpose.   It  will display both even + odd harmonics  out to the
21st harmonic.   The single biggest problem in each case is the 3rd harmonic,  followed by the 5th, the 7th is way down... and the
9-13-15-17-19-21  is a non issue.   There is no easy way to implement PF correction, unless the load is steady state, like a AC motor.
With stuff like high HP motors, thel oad is inductive..and the usual huge PF correction caps are used.   Even then, the value of the cap is
only correct for a specific load.  Change the load on the motor, and the cap is the wrong value.    In pulp mills etc,  you will see
banks of caps, with dozens of contactors used to   switch em in and out.... to obtain the best PF. 

##  On a varying capacitive load.,...like what you get from a SSB /cw ham  tube type linear amp with  a xfmr + high C filter cap,  its impossible to 
implement PF correction.   With a steady load like   RTTY /FM,  an inductor can be used for PF  correction,  but it has to be sized exact for the
steady state load.   Even then, the best you can do with either a motor load + cap..... or a HV supply + inductor....is to kill the 3rd harmonic.
You cant kill the 5-7-9th harmonics..so you still wont get unity PF. 

## In NA, for commercial use,  you get penalized for bad PF.    They want to see at least .9 PF and better.   In some states in the US,  if your
PF is below .85   they are not obligated to provide power..and will read you the riot act.  Install  PF correction, or get your power shut off !   

##  The new style xfmrs  wit the 3rd winding came about after the Bush jr, energy bill came out.   It is easier to use a xfmr that handles harmonics  vs
the end user trying to implement PF correction.    This all came about b4 switchers that also had PF correction implemented.  Most switchers  don’t have
PF correction.. but that is changing..esp with EU regs.  

##  I went through all this PF  correction with the late Peter dahl  several times over a 2 year period.   The consensus was..with the typ HV xfmr  + FWB
/  doubler  +  high C filter.... don’t attempt trying to implement PF correction. Dahl told me he had tried several times....with poor  results.   I tried several
configs myself, with the help fo a ham electrician buddy...and we ended  up either making the PF worse.. or a tiny bit better.   It can be done on a  steady state load,
but not on a varying load like ssb /cw.   In the end,  I gavc up.    The good news is,  dahl sez his xfmrs will  easily handle lousy PF.   They are designed for it. 
Put in all the  C you want.   He was correct of course.  800 uf  doesnt cause any additional heating of the xfmr. 

##  having said all that,    direct  rectification of the 240 vac line is still  feasible.   The PF will be no worse than if a Xfmr + high C filter was used.  You wont
get sky high neutral currents since the neutral is not being used.  But a  50vdc  @ 65A  CCS  switching type supply, 3.3 KW  CCS,, WITH PF 
correction is  cheap at  $450.00   to $500.00      

Jim   VE7RF