Hi Jim,
You've made a good summary below; it's indeed a fruitless exercise to
attempt to improve the PF of a big SSB linear RF tube amplifier.
At least your utility Watt-hour meter is billing you for the REAL power
consumed; not the larger VAR power. The benefit of correcting PF is mainly
to the utility company, not the residential energy consumer.
Anyone running a big muscle amplifier and concerned about power factor and
how it restricts the amp's potential power output and efficiency should
instead deploy a more efficacious 3-phase HV power supply and AC mains feed;
with inherently low ripple amplitude and higher frequency ripple that can be
cleaned up with a relatively small value filter capacitor. Doing so will
mitigate the PF and load regulation problem in a substantive manner.
This approach is the norm with big broadcast transmitters and some high-end
and high power RF amplifiers. For example Emtron's flagship DX-5 10 kW PEP
4CX5000R based amplifier appropriately deploys a separate 3-phase power
supply deck.
Anyone running their single-phase / FWB / capacitor reservoir filter power
supply amplifier at the legal limit output power of their respective country
jurisdiction will not be overly concerned about the practical limitations of
a less than unity power factor. A PF of circa 0.8 has no real consequence
worth worrying about.
What's more important is having low resistance utility feed cabling from the
street to the power distribution board and through to the radio shack so the
AC mains voltage feeding the amp is stiff and doesn't sag under heavy load.
Leigh
VK5KLT
________________________________________
From: ham_amplifiers@yahoogroups.com [mailto:ham_amplifiers@yahoogroups.com]
On Behalf Of jim
Sent: Wednesday, 16 May 2012 9:00 AM
To: ham_amplifiers@yahoogroups.com
Subject: [ham_amplifiers] Re: Power factor!
--- In ham_amplifiers@yahoogroups.com, "cathar.gnostic" <cathar.gnostic@...>
wrote:
> Here is a copy of an email I sent out on another group with no reply.
> Further to this email I now know that cos = 0.78 to 0.808 under full load
> with mains voltage dropping from 230 to 226.
>
> Hi all,
>
> I am doing some tests on power consumption of this unit (DX-3sp) using the
> FLUKE 345 PQ clamp meter. More tests to come and also on the DX-3 & DX-2a.
> On idle in standby the DX-3sp consumes 257 watts, in operation no drive
> 298 watts.
> Key down 70 watts drive the inrush current is 49 amps @ 230 volts.
> SSB the peak is 38 amps @ 230 volts.
> These readings are into a 50 ohm load with X < 4 ohms with band set to 20
> meters.
> I am not impressed with the power factor as cos is less than 0.8. As yet I
> have not logged the PF but I will and I think I will play with the
> incoming supply to see if I can get the PF up to at least 0.95 under load.
> This should make quite a difference to the output but I have to figure out
> how to only switch in the correction while in TX or the correction will
> cause the power to lead under low current which is not good. Maybe PTT
> switching to the PF relay?
> Any comments?
> ZL4IV
>
## been there, done that. I could write a book on this topic. The short of
it is, with a high C filter, you are wasting your time, and esp on cw or
ssb.
## a local ham buddy, who is an electrician, suggested we use a 32 uf cap
across the 240 vac line to try and correct for lousy PF. Well, that just
makes it even worse ! We also tried several different value caps. This was
on a hb 6700 vdc supply, which used a 10 kva hypersil pole pig with a 4800
vac sec + FWB + a 50 uf oil cap.
## My buddy said the cap across the input would cancel some of the xl of the
xfmr..well it doesn't. The 240 vac mains sees either the high C filter on
the output or it sees a choke, if a resonant choke is used. A cap across the
input will not cancel the C on the output side of a high C filter. A choke
across the input would work better.
## And of course any pf correction you install, has to be sized for a full
bore load..and wont be sized correctly on RX ! The fix of course, was to
open one side of the PF correction with a small spst high speed ceramic
jennings vac relay..wired to the vac TR relays in the hb amp.
## OK, it still won't work. Problem is on ssb or cw, the load is bouncing
up-down on each syllable. You can get it to sorta work..but only on a steady
state cxr... or FM. OK, now you have a 2nd problem with high C filters.
Since the diodes in the FWB, etc, only draw current every 8.3 msecs here in
NA, the duration of the pulses is real short. Its called a pulse train..and
not much different than the touted pulse train problems in any switching
supply.
## The pulse train causes HARMONICS back towards the 240 vac..and goes back
out into the street. The harmonics, specifically all the ODD harmonics out
to the 9th... will really screw up your PF even more.
## If you just use a simple ammeter on the input side, and also a fluke dvm
to measure AC Voltage on the 240 vac side.... and multiply current x
voltage... you will get the total VA power..which as you noted, is pretty
wicked. The problem with that approach is you can not differentiate between
PF and additional lousy PF caused by harmonics. To pull that off, you at
least require a Fluke, etc, SA, designed to look at odd harmonics...and it
will measure the exact current in each of the odd harmonics. The fluke will
look at both even + odd harmonics...but what shows up is just off harmonics.
## Now if you add the current in each of the odd harmonics, and add it to
the main fundamental, your total current is now just insane. The reason the
numerically added up fundamental + all odd harmonics does not equal your
ammeter is cuz the odd harmonics are partially out of phase with each
other...and the fundamental.
## The biggest offender is the 3rd + 5th harmonic. The 7th is way down...
and the 9th is miniscule. The fluke SA will look at both even and odd
harmonics out to the 20th.
## as you can see, its a real bitch to improve PF on any high C B+
supply...and impossible on modes like ssb-cw. Looking at commercial installs
like compressors, high hp motors and other stuff, the best I see is when
they install a choke in series with the typ PF correction cap. On any 3
phase set up, the typ deal is 3 x pf caps are used. In some cases, 3 x
chokes are added, one in series with each cap. This is to cancel the 3rd
harmonic...which is the main offender as far as harmonics goes. And that
scheme only works on steady state loads...not ssb-cw.
## at the telco I worked at, all the -52 vdc rectifiers used for the UPS
were typ ferro-resonant types, that put out typ 100-200-400-800 Amps CCS....
and as many as a doz of em in parallel. Those would typ have a PF of .9
## They are being replaced with the newer switching types. The typ switcher
is tiny, with 6 x 3.3 kw ccs supplies in one 25 inch wide rack..and only 8
inchs tall..and 13 inch deep. One shelf does 400A CCS. The old style type in
the same 400A is the size of a fridge. Another huge advantage of the
switcher besides its tiny weight and size is it has PERFECT PF at all phase
angles..under any load condition. The PF is always greater than .99 and typ
.998 That alone saves a ton of $$$.
## I find it amusing that in EUrope, any supply greater than 75 watts must
have electronic PF correction built into it..and thats for residential stuff
too...like PC's, tvs etc, etc. here in NA, PF and esp harmonics is a real
pita to the power co. It causes just over double the normal current flow in
the neutrals of 3 phase systems. In one case in Seattle, wash.... 3 phase
600 vac xfmr on grnd floor... then a separate 208-120 vac xfmr on each floor
of a typ office building. TONs and tons of pcs used on each of the 4 x
floors. And all of em run on 120 vac. The harmonics are so bad that the 50
kva xfmr used on each floor was ready to explode, yet the load on each one
was only 37 kva! The normal fix was to install a 75-100 kva xfmr. There is a
new type of xfmr out..with a 3rd winding... used to kill the harmonics. When
the new type is used... in a 50 kva format, the 50 kva xfmr runs cool. That
new type is more eff, and cheaper, vs the old style 50 kva.
## I went through all this PF stuff, several times with the late peter dahl,
and he said he had been down the same road b4. Told me to forget it...and
that I was pissing in the wind trying to implement PF correction, on TX...
with a SSB-CW load..and a high C filter cap. One thing though, his hipersil
xfmrs laugh at harmonics. You can use hundreds of uf of filter cap on the B+
supply, and you will not cook his plate xfmrs when using high C. You will
not cook this xfmrs, but you will get lousy PF..as you noted. The saving
grace is its an int load while talking on SSB... with a low duty cycle to
boot..... Plus its only on TX..not on rx.
## On RX, you have a different problem...called magnetizing current. I
worked on the PF thing, on and off for at least 7 yrs..and finally threw in
the towel. Now if this was a simple 2-5 kw SS amp with a mating switching
supply, with built in PF correction, like the telcos and cell sites use, we
would not be having this conversation. Even those things are fast enough to
handle ssb-cw.
## You could always send a note to Rudi at Emtron..and whine about the lousy
PF.... it would be interesting to see what his response is. On a similar
note, the PF on my yaesu FT-1000D is real lousy. Sucks 1000 va from the 120
vac wall plug.... just to get 200w into the ant ! I have not measured the
load on my switcher used in my newer yaesu FT-1000 MK-V. It has no PF
correction, so its probably lousy too.
end of rant.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|