How have you measured the noise from the AR7030's soap on a rope
supply?

The power dissipation in a linear regular is something that you
engineer the product to handle. Running hot isn't good, but hot to the
touch isn't nearly as bad as you think. It gets impractical to keep
adding heat sink after heat sink to a product, or worse yet a fan that
untimately will fail.

Switching power supplies, AKA switch mode power supplies, suck.
The active switch device is certain to fail at some point. I have
lenear
supplies that are 30+ years old, and except for replacing the filter
caps
caps at 15~20 year intervals will work until lightning gets them.

I had to attend a school on repairing switchmodes and the Sony techs
warned us that they are very sensitive to even minor changes in device
parameters. Gain,leakage everthing changes wth age and that is not
good.
As to repair, the replacement active devices must match exactly. Their
advice was repair was difficult and the repaired supply would never
be as trustworthy as a new one. They also showed us how the switching
transitor junction acts radiates ultrasonics that will eventually cause
the
junction to fail. They had a nifty ultrasonic transducer that
downconverted "audio" up to several MHz to down to 50~10,000Hz audio. I
owned a JBL professional audio amp with a switchmode that would eat
swithcing power transistors every couple of months. I got a friend to
mill the head off the
transisotr case and the junction was shattered like glass. We
sacrifcied a new transistor and the junction looked like those in a
text book.

Management insisted that we repair the switchmode supplies and sure
enough, no matter what parts we changed, they all failed in less then a

year. Part of the problem was we had no way to pick truely matched
parts.
Thjose little dabs of color on the parts show oddites like turn on/turn
off time etc. Diode turn on had to match transistor turn on etc.
A big goat rope and lots of fun.

Switchmodes are dirty as hell, and any attempt to round those nice
harmonic generating square waves only made them run hotter and fail
sooner. Some of the low noise equipemnt has the switchers in double
shielded cases with fancy feed throughs for power.
And they are still dirtier then a good linear supply.

Terry

Thanks for that Terry.
I was recently in the market for a new power supply and straight off thought
of one of the lightweight portable switchmode power supplies.
Thank goodness my good friend Guy Atkins warned me off them, as you do now,
so I bought instead a high quality linear (old fashioned) power supply that
is doing sterling service here in this QTH right now.

Powered up with this I got a nice catch yesterday morning at sunrise
(0500) - WTOP, Washington DC, 1500
Khz, 8,000 miles from here.
Have not logged a U.S. station on 1500 Khz before (from any location), so
was quite thrilled with this new catch. The U.S. stations come in so rarely
here over the mountains that it is quite an occasion. Had the opportunity to
fiddle with the Icom 756 a bit, and found that I probably would not have
been
able to ID and record this weak catch without the 756's outstanding NR
facility.

--
John Plimmer, Montagu, Western Cape Province, South Africa
South 33 d 47 m 32 s, East 20 d 07 m 32 s
Icom IC-756 PRO III with MW mods
RX Drake R8B, SW8 & ERGO software
Sony 7600D GE SRIII
BW XCR 30, Braun T1000, Sangean 818 & 803A.
Hallicrafters SX-100, Eddystone 940
GE circa 50's radiogram
Antenna's RF Systems DX 1 Pro, Datong AD-270
Kiwa MW Loop
http://www.dxing.info/about/dxers/plimmer.dx

wrote in message
oups.com...
Switching power supplies, AKA switch mode power supplies, suck.
The active switch device is certain to fail at some point. I have
lenear
supplies that are 30+ years old, and except for replacing the filter
caps
caps at 15~20 year intervals will work until lightning gets them.

I had to attend a school on repairing switchmodes and the Sony techs
warned us that they are very sensitive to even minor changes in device
parameters. Gain,leakage everthing changes wth age and that is not
good.
As to repair, the replacement active devices must match exactly. Their
advice was repair was difficult and the repaired supply would never
be as trustworthy as a new one. They also showed us how the switching
transitor junction acts radiates ultrasonics that will eventually cause
the
junction to fail. They had a nifty ultrasonic transducer that
downconverted "audio" up to several MHz to down to 50~10,000Hz audio. I
owned a JBL professional audio amp with a switchmode that would eat
swithcing power transistors every couple of months. I got a friend to
mill the head off the
transisotr case and the junction was shattered like glass. We
sacrifcied a new transistor and the junction looked like those in a
text book.

Management insisted that we repair the switchmode supplies and sure
enough, no matter what parts we changed, they all failed in less then a

year. Part of the problem was we had no way to pick truely matched
parts.
Thjose little dabs of color on the parts show oddites like turn on/turn
off time etc. Diode turn on had to match transistor turn on etc.
A big goat rope and lots of fun.

Switchmodes are dirty as hell, and any attempt to round those nice
harmonic generating square waves only made them run hotter and fail
sooner. Some of the low noise equipemnt has the switchers in double
shielded cases with fancy feed throughs for power.
And they are still dirtier then a good linear supply.

Terry

My preference is for linear supplies, but today's switchers are not all
that bad. I'm baffled about your comment on the ultrasonics causing the
transistor junction to fail. Most power devices fail due to hot spot
(really the weak points in a reverse bias breakdown) or
electromigration. Modern switchers use power fets, so there is no
junction in the bipolar sense, unless you count the fast recovery
diodes.

The real problem with switchers was the poor level of protection
circuitry. [I'm lumping start up and short circuit protection together,
though start up can really be a problem.] It takes a great deal of
overhead to make a rugged switcher. Discrete designs couldn't add all
that "just in case" hardware than is common in good switcher controller
IC made today.

For my own use, I still build/buy linear supplies. I don't care if they
are big and heavy. But think how big a 400 watt supply would be in a
PC!

One problem with switchers is the noise they generate on the line
itself. You need to filter both the output and the input on a switcher.
For radio use, the line noise can radiate and cause QRM.

I realize that the inquiry was over six years ago; I do not like to be negative about anything, but I thought I could post my experience with the Racal 6790; I have owned the rig for about three years; on-the-air time has been no more than an hour or two (literally); during that short tenture, I was impressed with the receiver; yes, no modern bells and whistes, but it seemed not to get in the way of the signal; that being said, I would not buy the receiver again; the first time I fired the rcvr up, it worked for about an hour and then I heard a pop and a little smoke came from the chassis; off to New England it was sent; it cost me a few hundred dollars for the repair (and not much less for the shipping as the rig is about 2' xs 2' square); got the receiver back and it worked for another hour and then I kept getting an odd error message on the screen that the manual didn't identify; make sure you download the first few pages of the manual as you will need it for the receiver's self-diagnosis system (push the AM and LOCK membrane switches simultaneously for such); my Harris 590A, Icom R-9000, Racal 17, and R-390 are comparable in performance and have run for years without so much of a hick-up; It is surprising the number of boat anchor people I talk to that include the 6970 in their collections that adds, "with issues". I see two listed today on QTH.COM for sale "as is". IMHO, I don't mean to bash anyone or anything, I just thought my two cents was worth giving to those still mulling this receiver. Thank you for the bandwidth, Phil