Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: expected bandwith of Norton rf amplifiers.  (Read 12961 times)
ZL1BDA
Member

Posts: 33




Ignore
« on: December 22, 2012, 02:13:00 PM »

Good day all!
I've run into an unexpected problem in finishing my MK XII home brew receiver.
This is an up converting circuit to 45 MHz with a 9MHz 2nd IF.
Now I use a Norton rf stage before a high level mixer. This works fine way up to
over 40 MHz. However the mixer needs a +17 dBM//50 Ohms local oscillator signal.
The synthesiser has an output of 1 volt//50 Ohms so I need to put in a gain of about
5x to get all the LO signal I need.
SO --- I built another Norton circuit using a 2SC908, just like the Rf stage using a binocular
balun core. It works fine up to ~57 MHz (high side injection), but by 70 MHz the gain is
down to about 1.5x.
Anyone have any experience with the Norton at up to 100 MHz?
Is it (as I suspect) the core material, or is the 2SC908 just not wide band enough to cope?

Hoping there is someone out there who has had the same problem and found a fix.
73 Cliff Wright
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #1 on: December 22, 2012, 02:54:37 PM »

The 2SC908 may be the problem.

A rule-of-thumb for amplifiers is that Ft of the transistor should be ten times the highest frequency required of the amplifier.

The Mitsubishi data sheet doesn't give Ft but the device seems to run out of steam at 600MHz or so.

The 2N5109 is the traditional device for small amplifiers; Ft is about 1.2GHz (depending on who wrote the data sheet) and it can sustain 50mA of current at 12v.

I used MRF581s (Ft 5GHz) in this two-stage Norton amplifier;

http://i49.tinypic.com/jrr3eq.jpg

http://i50.tinypic.com/95omu1.jpg

The transformers are 43 material/2302-style cores.
Logged
KB1GMX
Member

Posts: 815




Ignore
« Reply #2 on: December 22, 2012, 06:00:51 PM »

Be careful with the ferrites used as well.  Around 50-150 mhz is a break point
for lots of common ferrites and may be part of the problem as well.

Also the FT needs to be in the over 1ghz range and all my references for the
2sc908 say not much more than maybe 500-600mhz.


Allison
Logged
ZL1BDA
Member

Posts: 33




Ignore
« Reply #3 on: December 22, 2012, 06:37:32 PM »

Thanks folks. It looks as though my concern about the 2SC908 was probably right.
FYI I've tried several balun cores including UHF rated ones with no change in performance
and the same frequency roll off. One of my big problems out here in NZ is the supply of
 unusual components. Some of the bits already came direct from the US and Australia.
The 2SC908's were the best devices available locally at the time.
 Now I will research the MRF581. It looks like the identical balun cores in your amp that
goes up to 220 Mhz so that looks like the way to go.
Many thanks for the very fast reply and good advice.
73 Cliff wright
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #4 on: December 22, 2012, 06:56:31 PM »

Your component-supply problem in NZ is perhaps a little worse than ours in VK but not much.

You have, as I have, Jaycar and thank goodness for them; they are great for basic parts like resistors, cases etc but nothing terribly exotic.

I bought my "stash" of the MRFs from ebay which is where I get quite a lot of "exotic" things that our US brethren can get from Mouser etc.

W8DIZ is a great source of ferrites and far cheaper than I can get them here; in fact, I'm about to check my stock and order some from him.  He charges minimal postage.

Mini Kits in VK5 is a good source of some things, good connectors for one.  Not cheap but Mark VK5EME sells good products at a modest profit and you can't ask more than that.

Allison is correct (as always  Smiley  ) about the ferrite type; 43 is probably the all-purpose material for HF/low VHF with 61 useful to get up a bit higher.

I suspect that the roll-off in my amplifier is due to the transformers running out of steam; the peak may also be due to resonance in the transformers.

PM me and I will give you my e-mail address if you want to have a "QSO".

Kerry.
Logged
TANAKASAN
Member

Posts: 933




Ignore
« Reply #5 on: December 23, 2012, 02:27:31 AM »

17 dBm is about 4.5 volts peak to peak and that's a hell of a large signal to have around a sensitive thing like a receiver. Here's some ideas that might work:

1) Your required signal level is just below standard TTL voltages and most of the modern chips will operate to 200 MHz without breaking into a sweat. So, why not operate your synth at 90-150 MHz then have a divide by two as the final stage? You can then feed the Q and -Q outputs into a balanced line which will cut down on the internal radiation. Use a balanced attenuator at the other end which will reduce the level to exactly 17 dBm and give the mixer a better match.

2) If you really want to operate 45-75 MHz then consider using a short length of optical fibre to carry the signal. If you use a 5V device the receiver will output a signal at 17 dBm.

Both of the above ideas mean that you're working with square waves so simpler amplifiers than a Norton can be used to increase the synth output before distribution. Mixers (especially H-Mode) like to be driven by square wave signals.

Tanakasan
Logged
G3RZP
Member

Posts: 4847




Ignore
« Reply #6 on: December 23, 2012, 04:39:53 AM »

With those sort of signal levels, make sure the filtering is adequate to not radiate into the antenna input.  You really want rx radiation to be below -57dBm.
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #7 on: December 23, 2012, 12:38:14 PM »

If "strong" mixer performance is required, high LO power is required.  Mixers are identified by their "level", the LO power in dBm, so the OP's mixer is a Level 17 device.

Here is the Mini Circuits range; Level 3 is the "weakest";

http://i45.tinypic.com/2s80l1t.jpg

whilst Level 27 is the "strongest"; with half-a-watt LO power, that's a mixer!

The often-used TUF 1H+ Level 17 mixer has an LO/RF isolation of about 60dB at low frequencies; isolation is about 50 dB in the HF range and degrades to about 36 dB at 600 MHz.

A pre-amp will have an S2,1 of perhaps 10 - 20 dB; this, added to the mixer isolation, will keep antenna radiation low.

LO/IF isolation for this device is only a few dB less than the RF/LO isolation; a diplexer will add to this isolation by "filtering-out" the LO frequency as well as providing a good match for the mixer.
Logged
WB6BYU
Member

Posts: 13487




Ignore
« Reply #8 on: December 23, 2012, 01:10:58 PM »

And, of course, any front end tuned circuits or band-pass filters will provide additional rejection
of the LO signal because it is so far from the receive frequency.
Logged
ZL1BDA
Member

Posts: 33




Ignore
« Reply #9 on: December 23, 2012, 01:53:36 PM »

Well Thanks to all. After Xmas I will be trying to follow up some of those suggestions.
However, call me old fashioned!!!! But I really don't like the idea of driving a mixer with
square waves. I'm 72 and I remember too much trouble with harmonics!
I do actually use the TUF1H+ mixer and that's why I need +17 dBm of drive power.
The receiver uses a switched input attenuator followed by a tunable bandpass filter
from 1 to 30 MHz then the Norton RF stage (this one works fine with the 2SC908)
then the mixer followed by a mimic post mixer amp @ 45 Hmz then a 4 pole 15 kHz BW
xtal filter.

Since I ALWAYS use an ATU then hopefully most strong unwanted signals will be well down.

If I have to order from the US then I will also get some FT50-43 and 50-61 toroids to help
keep my bandwidth up.
Happy holiday to all and a better 2013. At least the Mayan apocalypse never came hi!
73 Cliff Wright.
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #10 on: December 23, 2012, 02:42:59 PM »

I'm not a great fan of MMICs; they certainly have their place and I do use them but I think a better post-mixer amp can be made with discrete components.

This is the mixer section of a receiver I'm still building (it's intended as a measurement receiver for detection of very small RF power, not as an "off-air" receiver so the IF is 5 MHz);

http://i47.tinypic.com/2vjeyx0.jpg

The design was "stolen" from W7ZOI (EMRFD I think) but I added a diplexer;

http://i50.tinypic.com/347trpl.gif

The 2N5109 is a very good device for making "strong" amplifiers; I think DIZ sells them at a reasonable price, check when you order the ferrites.  He also has other useful things like J310s at low prices.

Merry Christmas to all; it will be hot here in VK, it will be +30C where I am today but Perth in WA is looking at several days of close to 40C.

Logged
WB6BYU
Member

Posts: 13487




Ignore
« Reply #11 on: December 23, 2012, 04:42:52 PM »

Between 30 and 40 degrees here in Oregon also, but F rather than C.  Probably won't
be quite cold enough for snow, but lots of rain.

The 2N5943 is another substitute for the 2N5109.  And any of the power transistors
such as the 2N4417 or 2N3866 intended for 1+W output at VHF should work, but
watch the fT vs. Ic curve.  (The fT varies with the operating conditions.)  I've also
seen designs using several J310 FETs in parallel.  Sometimes that's  good approach
when the individual parts can't handle the standing current.


Diode mixers such as the TUF-1H actually work better with square waves instead of
sine waves:  that allows it to switch states more quickly, rather than hanging out
in mid switch during the longer zero transition period of the sine wave.  And one
advantage of using a double balanced mixer is that it reduces the strength of the
LO in the antenna port.
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #12 on: December 23, 2012, 05:18:57 PM »

Just noticed a typo in my earlier post; I should have said S1,2.

A diode mixer is, in itself, a square-wave device; as 'BYU said, it operates by rapid switching so a square-wave LO is beneficial.

But it will work fine with a sine-wave LO.

Apropos transistors; the BFW16A is a good substitute for the 2N5109 and the BFY90 is a good substitute for the '5109's little brother the 2N5179.

BFR91A & BFR90A are also good; they are in a "helicopter" package which is handy for "ugly" construction.

I have a stash of all the above; many came from a VK supplier of TV components such as were used before ICs; others came from ebay or DIZ.

You have to look around and grab a bargain if you see one; many ebay sellers are dreamers on price & postage but bargains can be found.

Here is a Christmas wish for all;

http://www.icq.com/img/friendship/static/card_7944_rs.swf

You may get a warning when you try to open it but it's perfectly safe.

It's just so corny that it's brilliant; it always makes me smile.
Logged
TANAKASAN
Member

Posts: 933




Ignore
« Reply #13 on: December 24, 2012, 11:54:21 PM »

To VK2TIL

Are you sure about that 6400 pf capacitor in the diplexer? Seems awful big.

Tanakasan
Logged
VK2TIL
Member

Posts: 342




Ignore
« Reply #14 on: December 25, 2012, 12:38:46 AM »

Dunno; what component values do you get for a Q=10 / 5MHz design?
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!