EMRFD Message Archive 1348

Message Date From Subject
1348 2008-01-22 13:11:40 bkopski SOME DIODE RING MIXER EXPERIENCES
As part of a recent project, I built a double balanced mixer
incorporating the usual trifilar wound FT37-43 core and 1N4148
diodes. The latter were carefully, repeatedly matched for forward
drop at 1 mA to the nearest 1 mV.

Upon application of this DBM I seemed to have more signal loss than
reasonable, so I dug deeper. This "digging" led me to more fully
evaluate this new DBM and three other mixers as well. The latter
included an old HB single balanced structure (built like the DBM), an
SBL-1, and an ADE-1. These four are shown in an accompanying photo.

For purposes of this test/comparison, the RF input is 5 mHz at -10
dBm, and the LO is 15 mHz at +7 dBm – both sources being well
padded. The SA display in the photos has the zero spur at the
leftmost grid, the vertical is 10 dB per, and the horizontal is 5 mHz
per. Vertical input signal reference is one grid down from the top.

The SBL-1 photo clearly shows the 10 and 20 mHz outputs at just about
6 dB down – as expected. The ADE-1 operation was indistinguishable
in this test so no photo is included.

Both the HB SBM and DBM performance photos show why (I think) I'm not
building any more HB mixers!! Aside from loss values of about 10 and
13 dBm respectively for the two main outputs, the rest of the
displayed stuff is – in my view – a horror show. Since both HB
mixers were carefully built (I don't think I can do any better), I'm
now in a quandary because I've often read how HB mixers "are just as
good" (at HF) if the diodes are well matched and the cores carefully
wound. So - does it take hyper-precision diode matching and robotic
core-winding – or what? :-) :-(

At this point I feel this is the end of my mixer-making UNLESS some
kind reader can suggest something I might investigate to get to the
bottom of this. On the other hand, if this really is typical
performance for HB mixers, then surely I can't be the first to have
experienced / documented such. Or am I …? (In any case – just in
case – I've ordered 6 more ADE-1's!!)

Cordially and 73,
Bob K3NHI
1349 2008-01-22 13:46:28 jr_dakota Re: SOME DIODE RING MIXER EXPERIENCES
It's the diodes ... commercial DBM's use hot carrier diodes and you
will see a difference in performance using them instead of switching
diodes...

That being said by the time you buy 2 ferrite cores and 4 hot carrier
diodes (Dan's has a kit for $2.50) you aren't saving that much
compared to an ADE-1 for $4.50 at Dan's .... even if you figure
minimum wage for your labor it's a no-brainer which is the better deal
....

1350 2008-01-22 13:56:54 John Levreault Re: SOME DIODE RING MIXER EXPERIENCES
What happens when you increase the LO drive?

de John NB1I

bkopski wrote:
> > For purposes of this test/comparison, the RF input is 5 mHz at -10
> dBm, and the LO is 15 mHz at +7 dBm – both sources being well
> padded. The SA display in the photos has the zero spur at the
> leftmost grid, the vertical is 10 dB per, and the horizontal is 5 mHz
> per. Vertical input signal reference is one grid down from the top.
>
>
> Both the HB SBM and DBM performance photos show why (I think) I'm not
> building any more HB mixers!! Aside from loss values of about 10 and
> 13 dBm respectively for the two main outputs, the rest of the
> displayed stuff is – in my view – a horror show.
1351 2008-01-22 14:00:34 Robert Cerreto Re: SOME DIODE RING MIXER EXPERIENCES
Bob,

Can't see your pictures.

Are you sure the cores are FT37-43? The reason I ask is that avender sold me some FT37-43's once and I used them in L-Section matching networks to a XTAL filter. My filter was a disaster. Then I took a close look at my cores, made some measurements, and it turned out that the vender sent me type -71 (or maybe -77) material cores!!!! They looked almost identical to the -43 cores!!! I don't want to insult your careful building, but make sure (again) that your windings are polarized correctly. Also, try a different set of 1N4148 diodes. Hot carrier diodes are better....but not that much better.

I have built many 1N4148 DBM's (and SBM's) with great success Their loss was always about 6Db and the output waveforms were as expected. The commercial ones are better but, there have been many times when I needed something in a hurry and I would build my DBM's in about 45 mins.

73, Bob WA!FXT

jr_dakota <SG2112@yahoo.com> wrote:
It's the diodes ... commercial DBM's use hot carrier diodes and you
will see a difference in performance using them instead of switching
diodes...

That being said by the time you buy 2 ferrite cores and 4 hot carrier
diodes (Dan's has a kit for $2.50) you aren't saving that much
compared to an ADE-1 for $4.50 at Dan's .... even if you figure
minimum wage for your labor it's a no-brainer which is the better deal
....

1352 2008-01-22 15:01:10 Johan H. Bodin Re: SOME DIODE RING MIXER EXPERIENCES
jr_dakota wrote:
> On a related topic (kinda) I've been experimenting with low noise
> power supplies for active antennas and am finding (as I suspected)
> that power supplies using 1N4007 rectifier diodes with their PIN
> structure are quieter (For VLF and LF) than a 1N4001 and 1N4002

Yes, those high voltage diodes do not switch as "hard" as lower voltage diodes
in the 1N400x series. It has something to do with the charge carrier lifetime, I
think. Another trick to make (linear) PSUs quiet is to put a 0u1 ceramic cap'
across each rectifier diode. I learned this the hard way when playing with
direct conversion receivers many years ago... This prevents LO leakage from
beeing "chopped" by the rectifier diodes - creating lots of hum sidebands! ;-)


72
Johan
1353 2008-01-22 15:09:19 Robert Cerreto Re: SOME DIODE RING MIXER EXPERIENCES
Bob,

I just got home and checked my lab notes. My LO drive was +14DB.

73, Bob WA!FXT

bkopski <kopskirl@cavtel.net> wrote: As part of a recent project, I built a double balanced mixer
incorporating the usual trifilar wound FT37-43 core and 1N4148
diodes. The latter were carefully, repeatedly matched for forward
drop at 1 mA to the nearest 1 mV.

Upon application of this DBM I seemed to have more signal loss than
reasonable, so I dug deeper. This "digging" led me to more fully
evaluate this new DBM and three other mixers as well. The latter
included an old HB single balanced structure (built like the DBM), an
SBL-1, and an ADE-1. These four are shown in an accompanying photo.

For purposes of this test/comparison, the RF input is 5 mHz at -10
dBm, and the LO is 15 mHz at +7 dBm – both sources being well
padded. The SA display in the photos has the zero spur at the
leftmost grid, the vertical is 10 dB per, and the horizontal is 5 mHz
per. Vertical input signal reference is one grid down from the top.

The SBL-1 photo clearly shows the 10 and 20 mHz outputs at just about
6 dB down – as expected. The ADE-1 operation was indistinguishable
in this test so no photo is included.

Both the HB SBM and DBM performance photos show why (I think) I'm not
building any more HB mixers!! Aside from loss values of about 10 and
13 dBm respectively for the two main outputs, the rest of the
displayed stuff is – in my view – a horror show. Since both HB
mixers were carefully built (I don't think I can do any better), I'm
now in a quandary because I've often read how HB mixers "are just as
good" (at HF) if the diodes are well matched and the cores carefully
wound. So - does it take hyper-precision diode matching and robotic
core-winding – or what? :-) :-(

At this point I feel this is the end of my mixer-making UNLESS some
kind reader can suggest something I might investigate to get to the
bottom of this. On the other hand, if this really is typical
performance for HB mixers, then surely I can't be the first to have
experienced / documented such. Or am I …? (In any case – just in
case – I've ordered 6 more ADE-1's!!)

Cordially and 73,
Bob K3NHI






---------------------------------
Looking for last minute shopping deals? Find them fast with Yahoo! Search.

[Non-text portions of this message have been removed]
1355 2008-01-23 06:54:12 Allison Parent Re: SOME DIODE RING MIXER EXPERIENCES
Briefly,

HB diode mixers using silic
1358 2008-01-23 13:15:53 dixonglennb Re: SOME DIODE RING MIXER EXPERIENCES
Bob said,

> At this point I feel this is the end of my mixer-making UNLESS some
> kind reader can suggest something I might investigate to get to the
> bottom of this.
>

Hi,

I'm Glenn, AC7ZN and this is my first post. I joined about six
month's ago at W7ZOI's invitation but have been lurking until now.
EMRFD is one of my favorite books.

Bob, You might enjoy seeing the picture of the TAK5 mixer I put
together about three months ago. Great minds etc.

I was very puzzled at your DBM diode mixer spectrum. It has always
been my experience that even poorly matched diodes attenuate the LO
much more than shown (of course, a 5 MHz third harmonic is at that
freq, but still...).

So I duplicated your experiment last night. The 1n4148 diodes were
matched with the diode test function of a multimeter. The
transformers are 8t trifilar on FT37-43. As the spectrum shows, I did
obtain a 6dB loss on the two main products. I shifted to a 14 MHz LO
so the spectrum could be distinguished from the analyzer grid, and so
that mixer products did not fall on top of each other. The LO
attenuation was more in line with what I would expect---about 61 dB
(actually, that was more than I expected--lucky this time). I didn't
pad any I/Os in this particular experiment.

I tried to wire the mixer and take the picture so the connections
could be discerned. Copper-colored wire is the RF and LO primaries;
the red and green are the secondaries. Note that the LO port is on
the right...I didn't put it on the bottom for some reason.

I've been squinting hard at the DBM picture you posted and I imagine I
see bifilar connections to two of the ports...if that is true I don't
quite understand that connection scheme.

73,
Glenn AC7ZN
1385 2008-02-01 07:04:27 wa7mlh Re: SOME DIODE RING MIXER EXPERIENCES
Bob:
I just got signed in to this group this evening as suggested by W7ZOI

I recently purchased 10 or 20 of the ADE-1 mixers.
They are amazingly cheap for what they are.

Plastic casing and you can see the guts of them from the bottom view.
They are used in a new multiband cw tx topology that I am almost done
with. (More on that later).

I used one ADE-1 as a true surface mount on some Cu clad FR4 ground
plane board building ugly. I Dremelled out some pads and that worked
well, yet was a bit of a nuisance for what it offered.

The draw back of the dedicated pads is that the mixer can not be moved
without grinding out more pads in a new location.

The second application of an ADE-1 was done upside down.
It fell in to place just great with a couple of resistor lead wires as
grounds. They lend themselves well to what we do in the HB scene.
They were easily half the price of any of the metal can DBMs from
MiniCircuits.

The data sheet in my opinion is very misleading.
They are not 5dB IL (typical) as stated, based on the data presented
in the actual data sheet pdf. Not that that matters, relative to a 6dB
or maybe a little less IL otherwise.

The HB diode mixers I have built in the past have been either single
or two-in-series (level 17 or so) and they did have a few dB more
conversion loss than a typical levle 7 SBL-1 mixer.
I did not know about driving them with 10dBm until recently. But they
still worked just fine, aside form the extra loss. But as a tx mixer
it was not an issue.

The one thing I did notice here in the Portland OR area was that a HB
diode ring mixer using true Schottky diodes was rather robust as a
front end for a direct conversion rx without any filtering between
antenna and mixer. I was amazed.

The diode ring mixer I built using 1N4152 diodes was severely affected
by one of the local FM stations (I was line of sight to the tower
maybe 5 miles away)and it took a bit of tuning with an extra external
FM radio to figure out what I was hearing. I could tell it was a pop
music station from "the rythym" but that was all. Once I found the
station, then it was obvious. It would have been a show stopper in
terms of trying to use it. A double tuned circuit helped a lot, but
not enough. An extra 5th order low pass solved the problem completely.

I am presently using 17dBm (two series Schottky diodes) mixers in both
rx and tx paths in my 40m/75m SSB/CW txvr and they seem to be working
just fine. I have not done any IM measurements on the rig yet, easily
10 years after a complete rebuild and redesign. Yet many times when
appliance ops on the air are complaining about QRM via poor dynamic
range and strong adjacent sigs, my rx can't even hear what they are
complaining about.

Interesting that you have brought up this issue since I was close to
doing some new evaluation of HB diode ring mixers since I was annoyed
at the expense of SBL1s and TUF1/TUF3 mixers in small qty, yet
weighing in what my time is worth. I even bought a few dozen new -43
toroids last month for this effort, only to decide to try out the
ADE-1 mixers and came to the conclusion that HB mixers were not worth
the bother unless I was out to make a point, more along the lines of
"fundamentalist home brew", as a theme. The real small foot print and
goof proof nature of the ADE-1 is hard to beat. Yet I am still drawn
to the idea of being able to make my own mixers. I think they look
cool too.

Hope this is some help.

73
jd
WA7MLH
1389 2008-02-02 09:01:24 john lawson Re: SOME DIODE RING MIXER EXPERIENCES
Jeff, good morning.

I read with interest your post on the ADE-1's. I
logged on to Mini Circuit's website and saw that in
quantities of 100, they are about $1.99 each. However,
in very small quantities (1-4)they want around $14.95
for each of them...Obviously they don't want to hurt
their sales of the SBL-1 and the TUF-1's. I'm
wondering if you know of a vendor that you can
purchase small quantities from, at a reasonable price,
thereby saving some money vs. the SBL-1/TUF-1 prices.
Many thanks, John Lawson, K5IRK

1390 2008-02-02 13:56:57 Pat Bunn Re: SOME DIODE RING MIXER EXPERIENCES
I would be willing to buy 50 or 100 of these and sell them through my mail order tube audio business at a much . I almost bought 50 yesterday after reading the posts.reduced cost. I can use 10-20 or so but don't need 50.

Pat Bunn
STF Electronics
----- Original Message -----
From: john laws
1391 2008-02-02 20:02:43 Leon Re: SOME DIODE RING MIXER EXPERIENCES
----- Original Message -----
1392 2008-02-03 05:09:32 Brad Hedges Re: SOME DIODE RING MIXER EXPERIENCES
Pat,



If you go ahead with this purchase, I'll buy 10 of them from you at
something closer to $2-3 /ea than $15/ea. Can you take PayPal, would you
like a check, or??? I'm good in QRZ.Com



Thanks,

Brad

AL4T




<http://groups.yahoo.com/group/emrfd/message/1390;_ylc=X3oDMTJydmpwMnQ3BF9TA
zk3MzU5NzE1BGdycElkAzE4MTc3Njc5BGdycHNwSWQDMTcwNTA2MzEwOARtc2dJZAMxMzkwBHNlY
wNkbXNnBHNsawN2bXNnBHN0aW1lAzEyMDIwNDIyMzc-> Re: SOME DIODE RING MIXER
EXPERIENCES

Posted by: "Pat Bunn"
pbunn@patbunn.com?Subject=%20Re%3A%20SOME%20DIODE%20RING%20MIXER%20E
XPERIENCES> pbunn@patbunn.com <http://profiles.yahoo.com/pbunnstf>
pbunnstf

Sat Feb 2, 2008 1:56 pm (PST)

I would be willing to buy 50 or 100 of these and sell them through my mail
order tube audio business at a much . I almost bought 50 yesterday after
reading the posts.reduced cost. I can use 10-20 or so but don't need 50.

Pat Bunn
STF Electronics





Bradley D. Hedges

Amateur Radio AL4T

Sergeant First Class, Infantry

82d Airborne Division





[Non-text portions of this message have been removed]
1393 2008-02-03 08:18:35 Pat Bunn Re: SOME DIODE RING MIXER EXPERIENCES
I purchased 100 off the Mini Circuits site last night. When I get the shipping notice, I'll let you know what the total costs are and I can take payment via Paypal or with a credit card via my business. In fact - I'll post a shopping cart button on the site.

Pat

----- Original Message -----
1406 2008-02-05 03:42:36 wa7mlh Re: DIODE RING MIXER Sources
John:
Sorry for the late response.
Man, this group moves fast!
Greg Daly WB7RSG and I put in an order direct to minicircuits for the
ADE-1s and I recall them coming in at maybe a little more than $2 each.
The $15 price would have certainly motivated us to go with TUF-3s.
I recall that each of got 20 ADE-1s
So I am not sure what happened on the pricing.
I do not know of any small qty vendors that offer anything other than
maybe 20% extra for singles.
But I have not looked around lately.
I think the strategy of the group order was a good way to approach
this issue.