SIERRA MODS

This is a collection of emails I've collected over time which
indicate how to put some mods into your Sierra transceiver.
Most of the mods have to do with improving the power output.

72, George N2APB

=================================================
from Bob Follett, WA7FCU (now: AB7ST)

All of the mods, save one, have been blessed by Wayne, even if they
don't fit his design intent of a "2 watt packpacking radio".  The
one mod he hasn't blessed is the reduction of inductance on L5 as a
means of creating a lower matching impedance to the PA collector.
One reason is that no one (that I know of) has tested what the mod
does to the third harmonic, in terms of -dB relative to total power
out.  Having said that, here we go:

BAND   Original Sierra   All mods except PA        MRF237 PA
                         13.7V IN   15.3V IN   13.7 IN  15.3V IN
-------------------------------------------------------------------
80M       n/c             4.2W                  4W       4.3W
40M     2.5W              4.0W        4.5W      4.9W     5.8W
30M     2.1W              3.0W                  4.0W     4.9W
20M     1.8W              2.6W                  3.2W     4.0W
18M     1.0W              2.0W                  2.9W     3.5W


Other mods include:

Q5 changed to a J310, all SM caps in modules except 80M, Meacham
"optimized LO Values", L5 reduced inductance, by removal of 1 T on
30M - 18M, 2 T on 40M & 80M (Wayne recommends the 2:1 matching
xformer in place of the PA Collector choke instead, but I had
already tweaked all my L5's), ferrite bead on the 2N2222A.

As you can see, adding the MRF237 did make a sizable difference,
except on 80M.  I haven't measured the total xmit current draw yet,
but will.  You will have to decide on how you each evaluate the
formula of PO vs. current draw vs. cost.  BTW, I got the final for
$11 from MCM....Oh, and you need to bend the base lead to the other
side of the case to reverse the lead layout of this device.

The biggest changes came from reducing inductance on the L5 toroid
on each module, but of course, I already had the SM caps on them to
reduce losses once power was increased.  Whats missing are the
Meacham "optimized LO values".  Here they are, inbedded in my tune
up instructions.

You don't need any calibrated wattmeter, all you want to do is
optimize the radio. So use what you have.

1.  Replace Q5 which is a J309 with a J310.  They are pin
compatable, and Wayne will start using the 310 in the next
Wilderness release.  Don't expect much from this change, maybe .1
to .2 Watts, but it helps.  This is not a priority, just something
to do on the mother board.

2.  Put a ferrite bead on the base lead of the 2N2222A driver to
eliminate VHF parasitics that show up on the higher bands.  If you
can't find one, Bob at Wilderness will sell you one...Call him.
(He also has the J310)

MODULES:

First things first.  All those trimmers on the top of each module
will, if within the proper range on their respective toroids, have
TWO peaks as you rotate them, not counting the frequency trimmer,
C70.  Re-peak the power output into your dummy load starting with
the two PMO/BPFs, C66 & C64.  Have two peaks? pick one, and move
on.  If not, you will have to modify the associated toroid.  Start
by compressing/expanding the existing turns.  If that doesn't do
it, try removing a turn. You may have to re-wind to add a turn, but
not likely.

Then move to the XMIT BPF, C36 & C33.  Same deal as above.  Now you
have as much power out as peaking those stages will allow.

That leaves the biggie, the two PI filters, L5 & L6.  You want to
remove each L5, and remove one or two turns.  Remove one T for 30M
and above, 2 turns below.  What you are doing is presenting the
final PA collector with a load of around 38 ohms rather than 50
ohms.  (I might mention Wayne does not like us doing this mod as it
MAY increase the harmonics, but no one has ever tested).  This mod
will produce your biggest power gain of all the tweaks.  Leave the
L6's alone, or slightly compress them to see if you get more power
output.  REMEMBER not to key down for for than a few seconds at a
time.

The hard mod:  on 40M and above, replace C47 C48 & C49 with silver
mica caps, preferably < 500V ones, or where you can't find the
proper SM, use polystryrene.  The problem here is finding small
SM's in large pF values. You can make any of them fit with some
bending, but its ugly!

Dave Meachem, in the Sept 95 QRPp came up with values for
"Optimized Output Filters" for the Sierra.  He did this for 50ohms
to the PA, and as I said, you will get more power by lowering the
inductance on L5.  (Wayne also suggests he would rather see you
wind a 2:1 xmformer toroid to replace the choke on the PA collector
(This would mean you leave L5 at original value, and the xformer
would cut the impedance to the PA in half).  He is probably right,
but I haven't tried it since I already re-did all my L5's.

Anyway, the SM's are less lossy than what's in there, and you will
pick up some free power.

BAND   C47   C48   C49
----------------------
80M    910   1800  910  (These guys are big.  you may try the
                         Polys)
40M    430   910   430
30M    300   620   300
20M    220   470   220
17M    180   360   180
15M    150   300   150

I ordered mine from Mouser.  You don't, of course, have to do all
of them, but the higher the freqency, the more they are needed.  I
left the 80M module stock, other than the L5, removing 2 T, and I
get 4.2 W out on a 13.7V supply.

For 160, 12 and 10 meters, the optimized output filter values are: (Per
Bob
Follet e-mail of 02/28/97)

BAND   C47,C49    C48      L??
--------------------------------------------------------------
160M   1600       3000     4.10uH, 32T on a T37-2 core
12M     120        250     0.324uH, 9T on a T37-2 core
10M     120        220     0.275uH, 8T on a T30-2 core (Note: I think
this is
                                            a typo, and should actually
be a
                                            T37-2 core.)

If you get through all of this, and want some more fine-level
tweaking, I have the instructions from Wayne to insure that all the
xmit stages are performing as they should.  I did this, but only
picked up .1W, so I don't recommend it unless, like me, you have
too much time on your hands.  :-)

One last thing.  The easiest way to increase power out is to
increase the supply voltage.  Wayne posted an E-mail about this
about 6 mo.  ago.  Just make sure you replace D7, the zener diode
rated at 36V with one rated at 43V, or even the next step up.  It's
all in the the front pages of the Sierra manual.  I have a AA pack
of 12 cells, which produces almost 16V, and I can get 4.5W out on
40M.  It really is the easiest way, so don't forget about it.
OTOH, I wanted to optimize my rig, regardless of voltage in, and I
think the mods were worth it.

In summary, order parts, then while waiting, start with the
modules, looking for db. peak, and re-winding the L5's.  This will
show the largest power increase.

If you need some clarification, don't hesitate to shout.

73, Bob

==================================================
Sierra Mods ... Clark Fishman, WA2UNN

My friend, KF2EW just finished building a Sierra and I had given him
some mods I found on the WEB.   The 2 we did was to change the final
transistor to a MRF237 and to change the choke feeding the +12 volts to
the final collector into a 4 to 1 impedance transformer.  Before we
changed the choke the rig put out a reading of 1 watt (actually the
wattmeter was not calibrated for use on 3.5 MHz and the rig was probably
putting out over 2 watts) .  I had a core that they used for the RF
choke so I wound up some #26 wire on an electric drill and wound 10
bifilar turns on the core.  The center tap of the windings(formed by the
start of one winding and the end of the other winding)  goes to the
collector and one feelead goes to +12 and the other to the output
lowpass filter.  The output doubled.  I took an accurate power output
measurement and the modified Sierra was putting out 5.4 watts on 80
meters.

Have fun

Clark Fishman  WA2UNN

======================================================
I modified my Sierra for RIT/XIT and wrote it up in the Dec. '1996 QRPp
(page 60).  I set it up for -1 / +6 khz XIT/RIT for split operation in
DX pile-ups.

Also, the K2 has two VFO's which can run wide split RX/TX. (
http://www.elecraft.com )

73,  Eric
=======================================================
Hi all:

A quiet weekend in Glen Rock (thankfully... no pun intended). Thursday at
the in-laws in Hunterdon as usual and the rest of the weekend at home. No
shopping for me, thank you very much. Call me Scrooge.

Not much operating time at N2SMH. I chose to spend my radio time doing some
Sierra tweaking. On one of my recent West Coast trips, I picked up a bag
full of silver mica capacitors for the express purpose of replacing the
ceramic and mono caps in the band-module output filters. Less loss and all
that good stuff. Some people blanch at the prospect of removing components
from plated-through boards. Not me. A good hot iron, a roll of braid and a
decent-quality solder sucker and off I go. Sometimes I like doing this
stuff more than operating. Depends on what mood I'm in, I suppose.

After replacing all the caps, I rechecked my power output. Unscientific as
I am, I can't really report hard figures on what I gained in terms of
output. But I did pick up at least something on all bands. Oh, I also
replaced Q5 (buffer between transmit mixer and driver), which originally
was a J309, with a J310 for a little more oomph into the driver. That
seemed to help a little, too. And seeing as how I'd replaced the stock 36V
zener at D7 with a 43 V diode, I thought I'd try running with a higher
input voltage to see what that would do for me. At 16 V input, I'm now
getting 3.3 W output on 40 meters. At about 13.5 V input (fully charged gel
cell), I'm seeing 2.5 W out on 40. Didn't try 16 V input with other bands
yet but I will sometime just for grins.

The best part was gaining a new band on the Sierra. I'd built a 10-meter
band module almost a year ago and had never been able to get any output on
10. I knew going in that the tuned circuits would be a bear to peak up
(extremely high Q) and try as I might, I was never able to get even a
whisper of movement on the WM-1. I tried, oh, has to be a dozen different
times with no luck. "Must be the toroids," I figured at one point. Removed
*all* of them in a fit of exasperation, rewound them, and *very carefully*
prepared the leads and reinstalled them. That wasn't it.

So this weekend, with the idea of replacing the output-filter capacitors in
mind, I went at the 10-meter module again. This time was different. I have
output on 10 meters! Yay! And just in time too, with solar flux heading in
the right direction. I measured 1.1 W out on 10 once I got it peaked. It
was simply a matter of being extremely careful and methodical. George
(N2APB), I think you'd said that you were in the same boat with your
Sierra. Keep trying, guy... it can be done. I think I can get a little more
output, too. I remembered after the fact that I should also repeak the
pre-mix trimmers. Gonna have to go back and do that on all the modules.

Last night I had my one and only QSO of the weekend. Heard WA4CND calling
CQ on 40. Maury was pretty weak, about 339. He answered me on the first
call and it turned out that he was running 4 W on an MFJ rig. His QTH is
Louisville, KY. Goes to show what a little 10.7-cm flux can do for you.
Typically, I'd never expect to get an answer from so weak a station. Gotta
love it.

72 David N2SMH
Glen Rock, NJ
=============================================

George:

I must admit that, to my own disappointment, there was no magic involved in
coaxing 10-meter output. Like you, I'd been thinking about sending the damn
thing to QRP Bob in the hopes that he could figure out what was wrong.
Well, nothing was wrong, and there probably is nothing wrong with yours.
Bob has told me that these filter circuits have a very sharp peak. I don't
remember having too much trouble with my 12-meter module but I know that
it, too, has a pretty sharp peak. As you go down in frequency, the peaks
get broader and broader (as the tuned circuits' Qs get lower).

So it wasn't the SM caps or toroids that got it peaked, but rather a very
methodical approach. The two trimmers involved are C33 and C36. What I did
was to move C33 just a couple of degrees clockwise. Then, I v-e-r-y slowly
rotated C36 over 360 degrees while intently watching the WM-1's needle (on
100-mW scale). If I saw no movement, I moved C33 a couple more degrees and
repeated the process with C36. Eventually, on one of these cycles, the
wattmeter kicked up ever so slightly. Bingo! Now I knew I was in the
ballpark and the rest was fine tuning. VERY fine tuning, mind you. It is
*so* touchy. If you so much as breathe on those trimmers, they go out of
whack.

I'm a pretty impatient person so for whatever reason, I'd never before
worked up the patience to go at the process in this manner. But this seems
to be the way it must be done.

Before you go about replacing the caps in your modules, you might want to
consider something that I was remiss in myself. The September 1995 QRPp has
an article by Dave Meacham, W6EMD, on optimum filter-component values for
the Sierra. If getting the absolute correct impedance match for your filter
is an issue for you, you probably want to use Dave's values and not the
stock values as you rebuild your modules. Me, I'm lazy and didn't do it.
Someday I'll probably end up doing this all over again as a result. If you
don't have the 1995 issues, I can make you a copy and send it out. Oh, and
the December 1996 issue has a great article by Bob Follett on boosting
Sierra output. That article is kind of what got me started on this little
exercise. It pretty much covers the waterfront on the possibilities for
mods to the output stages, including some previously unpublished tweaks
from Wayne himself.

Clark and I talked about his output-stage mods at the last meeting. I'm not
that ambitious yet. I tend to do this stuff in stages anyway. I don't know
if I'd ever go that far, but then, I resisted putting in a KC-2 for a year
or so and now I'm very happy that I did.

The Rainbow bridge sounds like it would be a good addition. Yet another
rainy-day project!

Pete Hoover's bandspread article was quite interesting. It doesn't look
like too difficult of a mod, either. Something else for the brain to
cogitate on, right? And as far as the homebrewed Sierra, thanks for the
pointer on more pictures. I will check them out later today. That was a
very impressive article and I told Doug that it was a highlight of the
issue for me. When I saw that Wayne had gotten the ARRL to put his article
in the Handbook, I figured it was a nice feather in his and NorCal's caps.
I never *dreamed* that someone would actually try to build one from
scratch! Nick did a jaw-dropping job. That is real homebrewing. I'd love to
see that rig someday.

OK... lunch time. Talk to you later.

72 David N2SMH
Glen Rock, NJ
===============================================

Well, this weekend saw yet another round of playing with Sierra band
modules, in particular the transmit bandpass filter. I'm not a power
monger. I don't necessarily want (or need) my rig to have a full gallon of
output. But I am interested in "optimizing" Wayne's design. An article in
the September 1996 QRPp by Bob Follett got me headed in this direction and
I think I'm getting close to the end... maybe.

Last time around, I replaced all the capacitors in the output filter with
silver micas. This bought me a modest increase in output on some bands but
not all. Recently, I was reading (and re-reading, as I often do) old QRPp
issues and was going through Wayne's original article on the Sierra.
December 1994 issue, I think. In that article, Wayne mentions compressing
the winding on L6 and spreading L5 to increase output. Follett mentions
taking a turn off of L5 as well.

So I opened up each module and starting compressing L6. Well, the results
were quite rewarding. For example, I *doubled* my output on 10 meters from
a scant 1 W to a solid 2 W. Don't mean to make you jealous here, George,
but the ol' WM-1 don't lie :-). I'm up to 2.3 W output on 12 meters. I saw
improvements on all bands. In my rig, 40 meters has always had the most
output and I'm up to 3 W there with 13.8 V input. All these measurements,
by the way, were from 13.8 V input with a 50-ohm HB load. When I use a 16 V
supply, I'm seeing at least 0.6 W more output on all bands. Almost 4 W on
40. Someday, I'd love to look at the waveform on a spectrum analyzer. I
wonder what, if anything, I've done bad wrong to my harmonic suppression.

The mystery band is 15 meters. That one is giving me fits. I can't get
output higher than 1.9 W no matter what I do. Compress L6, take a turn off
L5, tweak L3 and L4... doesn't matter. I even thought I lost the module
completely at one point. After one unsolder/resolder cycle of a toroid, I
had no output at all. Seems I had just completely altered the tuned
circuit's peak and couldn't find it nohow. That had me worried for a while.
But it lives... albeit at 1.9 W. It's weird when you can get more output on
10 and 12 meters than you can on 15. Oh, and with the 16 V supply, I get
*no* output on 15 meters. If I throttle back to 13.8V, I get my 1.9 W
again. Not sure why that is. I might have to go to the gurus on this one
(N6KR and/or W6EMD).

In between all the tweaking, I managed to make a few contacts in the MI-QRP
test. Not enough to write home about, maybe 15 or so. Yesterday, 17 meters
was way open. Worked K7QD in Idaho for a new QRP state (makes 40). And I
spent a good 45 minutes in a pileup for a KH8 station (that's American
Samoa if you don't have a chart handy). No, I didn't get him. But he was
*loud* in north Jersey, at least 569 on my W3EDP wire. You gotta love that.
The high bands are back! Anybody work him?

72 David N2SMH
Glen Rock, NJ
============================================

As you can see, the following goes back a-ways. But I had alluded to this
Sierra modification a couple of weeks ago. It's the one that enables me to
use my Wilderness Sierra to check into the Knightlites net (occasionally)
on 3686.4 MHz and to listen to 20 or 15 meters SSB while building or
modifying other gear.

This will hopefully get you started. There was some followup stuff I had
from off-line discussions with Wayne... as usual, I needed a little bit of
hand-holding. I'm pretty sure I have that in hard copy at home and I just
have to dig it up. But this is Wayne's original note to "Rich" (not sure
who he is) on QRP-L and I took my cues from this description. It does work
very well. The additional material I have is basically Wayne confirming
that I had interpreted his instructions correctly and providing a smidge
more detail, as I recall.

I would point out that output power up in the Novice band on 80 meters is
somewhat reduced. I get about 2W or maybe a bit more in the low end of 80.
Up there at 3.868 MHz, I think I'm looking at closer to 1W. We are
operating somewhere outside the design parameters here...

72, David N2SMH
Glen Rock, NJ

============================================
>From burdick@interval.comSun Feb 25 11:42:04 1996
Date: Sun, 25 Feb 1996 05:25:26 EST
From: Wayne Burdick <
burdick@interval.com>
To: Multiple recipients of list <
qrp-l@lehigh.edu>
Subject: Re: Sierra ABX, plus: VFO range switching, and receiving USB/LSB

Hi Rich,

Glad you like the ABX mod.  I use it all the time now and really love it.

I made another mod today that you might like to try: a VFO range switch
that lets me get up into the SSB parts of the band.  It's fun to listen to
LSB on 160 through 40, and you can listen to USB as well if you add a BFO
switch.

To add VFO range switching:

Put a small (very small) SPDT-center off toggle switch under the PCB, at
the far bottom right-hand corner of the front panel.  Connect the center
terminal of the switch to where the VFO tuning cap and the 180pF cap meet.
Next, put a 33pF silver mica or NPO ceramic disc from one side of the
switch to ground, and a 75pF from the other side to ground.

Result:  about 0-150kHz in the OFF position (center), 150-250kHz w/ the
33pF cap, and 300-400kHz with the 75pF cap.  You can use different caps if
you want complete overlap. Note that the tuning is not as linear in the
higher ranges and doesn't cover exactly 150pF.  This is because the
inductance has stayed fixed.

Also note that you may have to reduce the size of the 180pF cap to take
account of the capacitance added by the switch, even in the center-off
position.  In my case, I had to add a 1200pF polystyrene cap in series with
the 180pF cap to get it down to about 160pF.  You may not have to do this
if you have enough range left in the VFO setting trimmer.

To add the LSB/USB switch:

Mount another small toggle switch (DPDT, *not* center off) as close as
possible to the BFO crystal (possibly between the RF and AF gain controls).
The crystal must then be rewired from the switch common to C16, the
original trimmer.  The "LSB" side of the switch then goes through the 39uH
choke to ground, and the "USB" side can go through an additional 50pF
trimmer to ground.

To align the BFO, adjust the original trimmer in the "LSB" position, then
adjust the new USB trimmer in the "USB" position.

Note that LSB/USB are reversed on 10 and 12 meters.


Have fun!  I still hope to do an SSB adapter for the Sierra converter
sometime, and this is the first stage of the modification.

73,
Wayne
N6KR

=========================================
See the G3YCC website for some additional mods

[END]

rimmer in the "USB" position.

Note that LSB/USB are reversed on 10 and 12 meters.


Have fun!  I still hope to do an SSB adapter for the Sierra converter
sometime, and this is the first stage of the modification.

73,
Wayne
N6KR

=========================================
See the G3YCC website for some additional mods

[END]