Homebrew ring diode mixer balance

[W4AMV]

So, the little mixer home made by our new member works quite well. Without posting to the end of a 7 page run, here are the results. He matched the diodes quite well and after getting the transformers corrected the capability is quite good. Better than 30 dB rejection RF-IF and LO-IF through HF (30 MHz).

[WA1GFZ]

GOOD MAN!!!

Now try to build a type 3A DBM with 1/2 watt LO

[ssbothwell SWL]

thanks so much for doing all these tests on the mixer. what is this latest screenshot showing?

WA1GFZ, what would 1/2 watt mixer be used for? AM modulation?

[W4AMV]

I'll leave that task to our new designer! There are TWO screen shots overlaid, one is stored in trace memory. The brighter screen shot is swept LO signal on the input to the LO port of the mixer and the output of the mixer is the IF. The lighter second trace is the swept RF input port again with the output port the IF. In both cases the other port, LO or RF,  was always terminated into 50 ohms. The sweep is from a few kHz to 200 MHz. You see that as the frequency increases the balance degrades and certainly as we approach 200 MHz, the balance is quite poor. Commercial mixers will do much better in that regard, however additional circuitry is added. As you stated, for HF work, the IF filter could be removed now and the IF signal would be clear to see with the much improved signal balance (at HF).

[WA1GFZ]

A 1/2 watt LO mixer would be used in a high performance receiver.
Cubic used them in their R3030, R2411 type receivers. Racal used FETs rather than diodes in their HP mixers in many models. They used SD5000s in the RA6793 and RA6830.

[ssbothwell SWL]

do you have any documentation or articles that discuss how to make a 1/2 watt LO mixer? do you need more amplification of the RF input to match the 1/2 watt LO input?

also, i am curious about how mixers are used for AM modulation. are diode ring mixers practical for this purpose or are other types of mixers more typical? my homemade mixer uses 1n5817 diodes a ft37-43 transformers. hypothetically, would i be able to use it as high level modulator?

[WA1GFZ]

WOW you got 5817s to play. That is a power supply diode with real nads.
very cool idea.
Find some old watkins johnson app notes on mixers. They are the best.
They used to be on line. search on double balanced mixers.
Higher lo= better imd till you blow the diodes.
My homebrew RX I had 16 1N5711s. that ran warm. I hit it too hard once and took out 1/2 of them. I think i was around a watt lo using a 2n3375 as a LO driver

[ssbothwell SWL]

honestly, i wasn't being clever when i picked 5187 diodes. they were the cheapest schottky diodes available. what about using them is a cool idea?

i found some of the watkins technotes here: http://www.rfcafe.com/references/articles/wj-tech-notes/watkins_johnson_tech-notes.htm

[WA1GFZ]

You have the right time frame but the mixer note isn't there.

Type 2 mixer: 8 diodes 2 per leg, 13 dBM LO
Type 3: 8 diodes 2 per leg with a series resistor in each leg. 25 to 50 ohms.
              23 dBm LO
Type 3A: type 3 with a cap across each resistor. 27 dBM LO

The type 3A sets up a DC voltage across the caps to help reducing diode switch times

[w3jn]

AMV - looks like a HP 3577A you got there.  One of the most useful pieces of equipment on the bench, IMHO....

[W4AMV]

Yes JN it is. Very helpful. On the modulator, the simple ring you have is configured for DSB. See photo below. The carrier suppression is reasonable and matches up closely with the prior posting. The LO port supports the carrier, the IF port supports the audio, the RF port is now the UP CONVERTED output port, DSB. In this test, the carrier at the RF port is 7 MHz, and the audio IF at 20 kHz. The carrier level is reduced to almost 40 dB below each of the sidebands. 

[WA1GFZ]

here is an AM modulator

[ssbothwell SWL]

oh very cool article WA1GFZ. i really wish i had a spectrum analyzer.

so the basic idea is that you input the carrier wave and the audio source into the mixer and then you add carrier back into the modulated signal? i was looking in the arrl handbook and they mention modifying the diode mixer to make it slightly unbalanced. the article you sent me mentions is somewhat vague but Figure10 shows the step attenuator from the carrier and the modulated signal going into a "+" symbol. what does that represent?

i also found this website on modulators: http://www.eie.polyu.edu.hk/~ensurya/lect_notes/commun_cir/Ch8/Chapter8.htm
they mention that there are two main ways of doing AM modulation: low level and high level. heres their two diagrams: http://www.eie.polyu.edu.hk/~ensurya/lect_notes/commun_cir/Ch8/Fig8-18.jpg

low level uses a mixer and high level uses the audio source to modulate the power supply input to a power amp stage. here is their diagram of a high level mixer on a power amp stage: http://www.eie.polyu.edu.hk/~ensurya/lect_notes/commun_cir/Ch8/Fig8-19.jpg

at what dBm (or dBw?) do you need to switch over from a low level modulator to a high level modulator? now that i have my license, i would love to try to assemble a 5-10w AM transmitter. could i use a diode ring mixer for such a transmitter or would i need to use a high level modulator.

[WA1GFZ]

A low level modulator would require linear amplifiers after the mixer.
High level can be class b,c,d,e etc

[ssbothwell SWL]

right. would it be reasonable to have a class-A amp running at 5W or at that power level should i be switching to some other class of amp?

[WA1GFZ]

push pull you could get away with AB in a broadband amp and still get a good IMD.
AB works single ended as long as there is a resonant tank circuit.

[ssbothwell SWL]

very cool. once i get my receiver project completed i will try to build a push/pull amp. i had already been wanting to try building one and now i have a reason.

[WA1GFZ]

Push pull norton amp is the best you can use in an RX project

[ssbothwell SWL]

i am getting ahead of myself but.. i was asking about amplifying a low level modulated signal because i am imagining a possible transmitter project using my diode ring mixer to produce AM signal (if i figure out how to mix carrier back into the DSB signal) and then run it into a linear amp to get it up to a decent power output. would an amp like this be suitable for amplifying a low level modulated AM signal? http://golddredgervideo.com/kc0wox/wa2ebyamp/amppart1.pdf

[WA1GFZ]

I would put a low value resistance in series with each IRF510 source. This will make it more stable over temperature. 1 to 5 ohms should work.  You will need driver stages. A DBM output is around 0dBM so you will need about 30 dB of gain to drive that amplifier.