[Amps] Shorted turns experiment

Thu Aug 5 19:19:13 PDT 2010

Here is W8JI's response about tapped toriod coils: It was over 3 years ago,
I thought it was only a year.

http://lists.contesting.com/archives//html/Amps/2007-03/msg00361.html

This was my original question on the subject and the start of the thread.
http://lists.contesting.com/archives//html/Amps/2007-03/msg00348.html

Here is another link to a past discussion on tank circuit Q, part of the
ongoing discussion on my TMC amp's inefficient tank circuit. More toroid
losses are discussed. 
http://lists.contesting.com/archives//html/Amps/2007-04/msg00155.html

73
Gary  K4FMX

> -----Original Message-----
> From: Carl [mailto:km1h at jeremy.mv.com]
> Sent: Thursday, August 05, 2010 8:06 PM
> To: garyschafer at comcast.net; 'Larry Benko'
> Cc: amps at contesting.com
> Subject: Re: [Amps] Shorted turns experiment
> 
> Gary, Larry, I could especially understand that if the toroid was in the
> Pi
> and used/tapped  up thru 30 MHz, I wouldnt use one that way either. I
> tried
> that decades ago with poor results above about 14 MHz using a 2 Mix.
> 
> Ive been discussing 160/80 taps in a pi and all band taps or maybe 10/15
> in
> an air coil and 20-160 on the toroid in the L section in existing
> commercial
> amps that show no apparent difficulty in making good efficiency on any
> band.
> If the output isnt drifting there cant be excessive
> losses/heat.....agreed?
> Tuning drift is almost always due to the fixed padder caps used on 80/160
> in
> my experience....thats where the manufacturers cut corners on cost.
> 
> An air coil is highly influenced by its surroundings. In any steel cover
> amp
> do the cover on and off test on any band. A toroid is much less influenced
> resulting in a smaller amp volume.
> 
> If the efficiency at 1500W is 1-2% less in the toroid.....who cares? Just
> size it so that it stays well within its temperature specs.
> 
> Carl
> KM1H
> 
> 
> ----- Original Message -----
> From: "Gary Schafer" <garyschafer at comcast.net>
> To: "'Larry Benko'" <xxw0qe at comcast.net>; "'Carl'" <km1h at jeremy.mv.com>
> Cc: <amps at contesting.com>; "'Dennis OConnor'" <ad4hk2004 at yahoo.com>
> Sent: Thursday, August 05, 2010 7:56 PM
> Subject: RE: [Amps] Shorted turns experiment
> 
> 
> >I have an old TMC (technical materials company) amplifier that uses a
> pair
> > of 3-500s that has an auto (preset) tune tank circuit. The tank coil has
> a
> > tapped toroid core in it. I replaced part of that coil with a copper
> tube
> > air wound coil and the output power increased by 200-300 watts. Off hand
> I
> > don't remember exactly.  The efficiency of the amps was quite low. I
> > posted
> > about it on hear around a year or so ago and had discussions about the
> > tapped core with W8JI as well. He also suggested that it is a bad idea
> to
> > short turns on a toroid.
> >
> > 73
> > Gary  K4FMX
> >
> >> -----Original Message-----
> >> From: amps-bounces at contesting.com [mailto:amps-bounces at contesting.com]
> On
> >> Behalf Of Larry Benko
> >> Sent: Thursday, August 05, 2010 4:41 PM
> >> To: Carl
> >> Cc: amps at contesting.com; Dennis OConnor
> >> Subject: [Amps] Shorted turns experiment
> >>
> >> Carl and Denny,
> >>
> >> Hope you don't mind me butting in here.  I had a private exchange with
> >> another guy and will post an experiment I did here as it might help.
> >>
> >> This following reply was in reference to a particular amp that I was
> >> mistaken on concerning the output network which had some shorted turns
> >> in the output inductor which was wound on a powdered iron core:
> >>
> >> Shorted turns must not cause enough heat to be a problem since this amp
> >> design has been around and is well respected.  As I indicated before
> the
> >> "designer" made a decision here given some set of constraints.
> However,
> >> shorted turns ANYWHERE that are cut by flux lines will produce a
> current
> >> and result in loss.  An amplifier designer is in a predicament given
> the
> >> fact that band switches are becoming one of the biggest cost items in
> an
> >> amp after the tube and the transformer.  Some of the bizarre custom
> >> switches that we saw 40+ years ago will never be seen again unless you
> >> had switch volumes big enough to get the attention of the switch
> >> manufacturer so we use what is available.
> >>
> >> Just for grins I just did a little experiment.  I found a #2 powdered
> >> iron core which happened to be already wound and center-tapped with
> >> about #18 wire.  I did a quick measurement and the inductance of 1/2
> the
> >> winding was ~3uH.  I put this winding in parallel with a dummy load and
> >> preceded it with a large "L" network tuner to give the amp a 1:1 SWR.
> I
> >> turned up the power until I got some moderate heating.  At about 600W
> on
> >> 1.8MHz  the core temperature rise was about 10 deg. C in 2 minutes.
> >> Then I let it cool down a and shorted the other half of the coil (it
> was
> >> open in the 1st test).  The SWR was no longer 1:1 (as expected) so I
> >> re-touched the tuner and again applied 600W. This time the temperature
> >> rise was about 22 deg. C in 2 minutes.  Clearly the shorted turns
> caused
> >> the additional temperature rise.  However the 22 deg. C rise is still
> >> very much within the allowable temperature rise for the core.  So one
> >> person might say the shorted turns are bad and someone else say they
> are
> >> acceptable.  It is very possible that an amp designer might increase
> the
> >> core size a little and live with the shorted turns as the best
> >> compromise.  However if I was designing an amp I would try to avoid
> this
> >> scenario unless I did some REAL testing to understand what the effects
> >> will be.
> >>
> >> Good designs are always compromises and if "my" legal limit amp was 20W
> >> more efficient than "your" amp but "your" amp cost $300 less than "my"
> >> amp you might sell a lot more amps than I would. :)  Thanks for the
> >> discussion.
> >>
> >> Just a FYI data point guys and there is no reason to believe that
> >> another design would have the same relative temperature rise as my
> >> experiment.
> >>
> >> 73,
> >> Larry, W0QE
> >>
> >>
> >> _______________________________________________
> >> Amps mailing list
> >> Amps at contesting.com
> >> http://lists.contesting.com/mailman/listinfo/amps
> >