EMRFD Message Archive 13566

Message Date From Subject
13566 2017-02-08 16:31:56 Jim Leslie Local oscillators for a high level mixer
For a 20M receiver I have underway I had planned to use a Minicircuits SBL-1 for the mixer. But I just happened across a couple of SAY-1 mixers and after reading an article in QEX May/June 2001 on the use of high level mixers, thought briefly about switching gears and incorporating one of those in place of the SBL-1. Getting 7dBM for the SBL-1 is no problem but the 23dBM LO requirement for the SAY-1 made me think a bit. For a quick test I breadboarded the general purpose class AB amp on page 2-32 of EMRFD and was able to get 22.5 dBM out of it which is fine, but for continuous duty it does run quite warm. After much searching I have not found many LO references or circuits using high level  mixers. Has anyone used a level 23 mixer and has any LO guidelines/advice/stories/circuit designs to share? I suppose I could heat sink my 2N3866 better or better still up the ante a bit O/P transistor-wise, but I'm very curious & would really prefer to know I am on the right track and that I'm not overlooking a better path to take.
73 - Jim
13567 2017-02-08 21:36:07 John Marshall Re: Local oscillators for a high level mixer

I breadboarded a receiver strip years ago using a RAY-3 mixer with 23 dBm LO. Pretty sure I used a resistive feedback amp driving a class C 2N2219 with a several-section low pass filter. Class A or AB is probably a better way to go. Ideally you would like to get 26 dBm out of the low pass filter, then use a -3 dB pad between that and the mixer. That's going to take a watt or so DC input and plenty of heatsink. Use an insulated heatsink that you can mount to a chassis or diecast box. You could also consider using 2 of devices in parallel as shown in EMRFD Fig. 6.98. That circuit shows 2N3553's but lower Ft devices should do fine at HF.

23 dBm mixers seem like a brute force approach in these days of H-mode FET mixers, but "build with what ya got" is the ham way. Good luck!

John, KU4AF
Pittsboro, NC

13570 2017-02-09 07:07:27 Rod KM6SN Re: Local oscillators for a high level mixer


H mode mixers are introduced in EMRFD.

Recently Bob  N7SUR kindly pointed me to some amazing original research on them, saying:

Have you looked here?

H-Mode mixer



The H mode mixers can have +40-dBm IP3 intercept, and they do not require high power LOs.

Well worth taking a look.

The .1 inch center transformers used in


are available on ebay very cheap. Contact me via private email if you cannot find them.

Best of luck with all,


13573 2017-02-09 09:12:18 Jim Leslie Re: Local oscillators for a high level mixer
Thanks! Also just found a class A with apx 800 mW o/p in QEX Mar/Apr 2001 based on an SSD design which also uses the 2N3553's. The H mode is great but this idea was just because I came across a couple of SAY-1's.
73 - Jim

13574 2017-02-09 09:34:52 Pat Byers Re: Local oscillators for a high level mixer
Hi Jim,

Have you considered using a Class E amplifier for the vfo? The heat generation should be far less than that of a Class AB amp and you can use cheap MOSFETs for the output. The NorCal NC2030 achieves about 5W with three BS170 MOSFETs and the output can be reduced by lowering the supply voltage. The manual suggests 7v will get you 1.5W. I don't know how this translates into dBm but we could look it up. A description of the NC2030 transmitter is on page 121 of this document:

I have some +17 dBm mixers that I want to use some day so that was the approach I was going to take.



13575 2017-02-09 10:03:15 Jim Leslie Re: Local oscillators for a high level mixer
No, actually never thought about class E. But after looking at the NC2030 pdf and re-reading P2-32 in EMRFD I also just downloaded the QST article May/June 1997 by the Cal Tech group that discusses class E amps, I will look at that now.
thx - Jim

13576 2017-02-09 10:31:01 Pat Byers Re: Local oscillators for a high level mixer
There are some presentations on the design of the NC2030 and Class E PA by Dan Tayloe kicking around the Internet. I don't have the URLs handy but I think the NorCal Web site would be a good start.



13582 2017-02-11 03:30:06 Bill Carver Re: Local oscillators for a high level mixer
class E has VERY HIGH harmonic output. Just need a really good LPF ahead of the mixer: even harmonic content (2nd probably being strongest)  will blow the third order intercept of the mixer. Rohde used to use a 7-pole elliptic filter with a parallel-tuned notch at the second harmonic.

13583 2017-02-11 04:30:06 Bill Carver Re: Local oscillators for a high level mixer
I drive a +23 mixer with a square wave LO from a 74AC540. I'm putting 6.0V on it for power, four sections in parallel at input and output, with a 43 ohm resistor to the mixer so it's a broadband source of 50 ohms. This delivers a few dB less LO power to the mixer, but being a square wave it has faster rise/fall times at the switching point and that increases the mixer intercept over a sine wave.

(the only rationale for only using four in parallel is I had several +23 mixers so the other four are driving an identical transmit mixer so I don't have to switch inputs and outputs ports of the two mixers when going from TX to RX).


13584 2017-02-11 08:01:36 winston376 Re: Local oscillators for a high level mixer
Here's an extract from post # 12622 (Apr 9, 2016) that I posted appropriate to this topic:  "....Using an si570 based Vfo, my solution to increasing the power level AND provide 3 additional Vfo output ports was to use a 4 port HCPL ic isolator.  I can simultaneously connect a freq counter, scope, SA as well as being connected to the mixer with minimal interaction.  The HCPL provides a high impedance to the Vfo and doesn't, load it down at all.  This architecture can provide an isolated second Vfo output ...."   The output provided sufficient power to drive the high level mini-circuits DBM that I was using.  Works very nicely.
13589 2017-02-12 11:20:29 Jim Leslie Re: Local oscillators for a high level mixer
Thanks to everyone for the replies. In the end I started down the road John KU4AF suggested - the amp in fig 6-98. Sadly I don't have 3553's but I do have some mystery transistor pulls from an old surplus transmitter UHF board. Don't know anything about these things other than the form factor is like the MRF300 series and they were configured push pull. They're marked CD2403C but can't find any data on them anywhere. Guessing 10W+ range so 23dBM should be easy. :) Have a few more so if results look promising, I may just rethink the PA stage!
Getting there is most of the fun...
73 Jim