[Amps] Price per Watt Conversation

[Jim Thomson]

##  I have given up on the quest for good eff +  good imd.   On some of 
these SS amps, they provide for a lower voltage, to increase eff, when using 
low power.
The AL-600  has this feature with its manually selected   42V and also 50 V 
position. Other SS amps  will auto switch the Voltage down, when PO is 
reduced below a set threshold.
Problem is, the eff increases back to its normal 50%.... but the IMD also 
gets worse.  None of these SS amps currently in production, including the 
latest and greatest have good imd and good eff.  Its always one
or the other.   The pre-distortion scheme, like used in the ADAT etc will 
improve the imd, typ by 10 db.   Problem is, they all have this  20-30 msecs 
delay, so there is no way you can monitor your ssb signal, it doesnt work.
Anything much over 6 msecs  will sound horrid, when listening on headphones 
with the monitor on.  IMO,  50% eff is pretty bad..and 35% is abysmal.   For 
now, I will stick to GG triodes.  At least good IMD and 65-70% eff,
and zero cooling issues..all in the same sentence.   Im not about to buy an 
ADAT xcvr anytime soon, so will have to stick with what yaesu has given me. 
Class A on the MK-V provides for superb imd,  but the eff is beyond
useless.  Other than brief tests, it will cook the heatsinks. On class A, it 
sucks 10A CCS  at 30 vdc.   300w CCS of diss, when u hit the footswitch..and 
say...nothing.   When actually driven, its then 300w input.... and a max of
75 w output..so then diss is reduced to 225w.  So although class A works, 
the eff is so bad, it has been rendered ineffective.   I know one fellow who 
burned out 3 sets of transistors on his MK-V.   It doesnt help that there
is no air going down past heatsinks 1+2.... only down 3+4.

Jim  VE7RF



-----Original Message----- 
From: Catherine James
Sent: Tuesday, April 25, 2017 6:24 AM
To: amps at contesting.com ; Jim Thomson
Subject: Re: [Amps] Price per Watt Conversation

Thanks, Jim.  I am learning something new here every day.

I've been clear for some time on the fundamental conflict between good 
linearity (indirectly creating good IMD) and efficiency. With the last 
couple of days of discussion added, I now see that there's also a 
fundamental challenge in getting good efficiency and good thermal handling, 
and that the negative feedback solution isn't necessarily practical for some 
designs.

The ideal amp would be highly efficient for lower power draw and minimal 
need for heat-sinking, have excellent linearity and IMD, use cheap devices, 
and support ALC, voice compression, and 100% duty cycle modes.  Looks like a 
square circle. :-)

73,
Cathy
N5WVR

Jim Thomson <jim.thom at telus.net> wrote:

###  NFB =  Negative feedback.  3 db of NFB means you also reduce the gain
of the amp by 3 db.  On paper, to achieve say a 10 db improvement in IMD, 
you require 11 db of NFB.

##  6 db of NFB on a SB-220  means you now require quadruple the drive
power.   NFB will work on a SS amp, since the gain is sky high  to begin 
with.