Homebrew Double Balanced Mixer Part 2 of 3 

Hi Trevor, Thanks for watching and for your post. Be sure to look at the other two parts. This DBM was part of a project and article that appeared in QRP Quarterly called Let's Build Something and that something was a homebrew SSB Transceiver. Let me not fool you and let me caution you. Building a DBM is not a casual thing and your results unless you are extremely lucky will not match what you can commercially purchase. The way the windings are spaced on the core, the tightness of coupling, using for real matched diodes and the drive level are crucial to the success. I found that the homebrew units required more LO drive. Also be sure to add the balance pot as that will arrest some of the mechanical problems with windings and diodes. Building a homebrew DBM is like having a date with the girl who is known to be "loose!" You do it once to check the block and then you find a nice girl as a regular date! Packaged DBMs used to be expensive but now can be had for more reasonable prices. So go ahead and check the block but if you are building a converter for an SDR Dongle then look to units like the ADE-1, SBL-1 or the TUF-1. Look me up in QRZ.com under my call N6QW and you can email me direct. The loose girl was lots of fun, but not someone you take home to meet Mom! 73's Pete N6QW

Hi Peter, Thanks for your comment -- I actually have received several inputs that criticize the method I used . So your input is much appreciated. I just reason as in so many insttances -- those who made negative comments never used it but simply opened their mouths and made comments. Pete

Thanks for taking a look and posting a comment. 73's Pete N6QW

Thanks for your post. The videos accompanied a published article where these DBM's were part of a SSB transceiver project. My co-author and myself received several emails --your DBM doesn't work. When asked if they watched the videos -- the reply was well no. I do not know how we could have explained it any different. Three windings that could be easily identified and follow what was said about the hook up --well you know the rest. Glad it was useful to you.73'sPete N6QW

Pete, the windings could arguably be buzzed out, as long as you which end is the start and which is the finish, I was unaware that the videos accompanied a published article, but it is always good to see what people are building. FWIW I have seen a commercial HF rig with one of the diodes in the ring reversed, it still mixed... 73s G0FVT

Thanks Tony, Almost forgot I made those videos. The trio of videos was for the LBS project, 73's Pete

Hi John, Thanks for the feedback. I did find that because the DBM was not made on an assembly line that it took a bit more drive (like 10dBm) but other than that I too found it works FB. If you like to experiment you might try some matched 1N5070 Shottky Diodes --versus the Silicon. Have fun. 73's Pete

@@N6QW I’m driving the LO with 1v pk pk which does equate to +10dBm at 50 ohms. I’m also using 1N5817 Schottky diodes. My transverter will be 432 kHz to top band using a 1.5 MHz I.F. Thanks again for the video.

@@NorfolkPilot Hi John. You might want to check your calculations 1 V PTP = 4 dBm , 1.414 v PTP = 7 dBm and 2 V PTP = 10 dBm. --so you are I think driving it with 4 dBm. The formula I use = VPTP^2 and that times 2.5 (conversion factor including 50 Ohms) the the log of that times 10. Check my math --but you may be OK because of the Shottky diodes. Pete

@@N6QW you are correct Pete. I calculated 0.707 of 1v when I should have calculated 0.707 of 0.5 volts. Even so, increasing drive above this results in little extra output, but more noise.

That certainly is an excellent solution. Bravo Thanks for the post. 73's Pete N6QW

Hard to respond on you tube. Send me an email to n6qwham@gmail.com. I will be happy to share the LT Spice simulations -- with a DBM you have 6 dB of conversion loss --actually maybe 10 dB with a homebrew mixer so you might have to also have a strong post mixer amplifiers to make up the gain deficit. The J310 DGM is a neat trick as I have used them as mixers, Amplifiers and product detectors Pete N6QW

Hi Lee, You have put your finger on the issue = MATCHED. It really doesn't matter too much -- the 914's or 41448 or 4125 -- The problem is very closely matched diodes. In critical applications I would forego the homebrew mixers. the only reason they were used was in connection with the Let's Build Something project. The homebrew DBM's were lossy and needed more drive. power. Save yourself grief and buy an SBL-1. ADE-1 or TUF-1 and just move on. 73's Pete N6QW

Thanks for looking at the video. 1st it is well documented that the type 43 material is well suited for use in RF transformers in the HF range. The Al value is 350 for a type FT-37-43 core. The number of turns was a bit of a 1st try but a single winding of 10 turns looks like an impedance of about 395 Ohms at 1.6 MHz and around 2K at 9 MHz and 6.6K at 30 MHz.So anything pumped into the ring at 50 Ohms would not be loaded by the winding. At 1.6 its 8:1 and at 9 MHz it is 40:1 and 30MHz 132:1. The type FT-37-43 will easily handle a triple 10 turn winding using #26 wire. Now just for fun if you use a Type 82 core (FT-82-43) with a winding of 70 turns that looks like a 2.5 Milli-Henry coke 73's Pete N6QW

@@N6QW So a lot of interesting information! Thank you for the answer! Now I try figure out how it works and my knowledge is so small. At current time I try assemble super heterodine using tubes so I collect information as much as possible. 73! Oleg UR7WKH

Yes! Lead Marked #1 should be at the same point as the connection to lead #2. The wires are all wound in the same direction. 73's Pete N6QW

@@N6QW Thanks for the quick reply Pete, I need to go back and check some of my builds now. It's interesting because SSDRA (fig 3.30) shows a phasing dot on the primary just like yours whereas EMRFD (fig 5.19d) does not. But I know your view of the relative merits of these two books. Much appreciated, 73s.

@@mars667 I keep telling everyone -- EMRD makes a great door stop. Think in reverse -- if the wiring approach is unimportant then why ever were phasing dots created.

Depends on the core you use and the usable frequency range of that core and the operating frequency of the DBM. The inductance of 10 turns is 35 microhenry and at 9 MHz the impedance is 1979 Ohms. Turns = 1000*(square root of inductance in millihenry/ Al Value). For current FT-37-43 production the Al value is 350. Reverse the math and you get 0.035 Millihenry which is 35 microhenry. Pete N6QW

@@N6QW Thank you very much for your reply , Mr Juliano Since I do not have access to the main components, I use a low-energy lamp core, and 10 rounds of this core have a value of about 160 micro-Henry. Is a value higher than 35 micro-Henry due to signal loss? In general, what is the role of inductance and winding resistance in this type of mixer?

@@fariborzsadeghivafa4473 The role of inductance is complex as is its value. I suggest you do an internet search on double balanced mixers as I am sure the Inet would provide you the answer that is hidden in your questions. One issue is that the match to 50 Ohms typically the in/out of a commercial DBM.. The rule of thumb is 10 times the IMPEDANCE (not resistance) should be the core winding versus 50 Ohms. So if you are at 9 MHz then the impedance is just under 2K --which is 40 times the 50 Ohms. But should you move the DBM to operate at 3 MHz then the Z across the 10 turn winding = 650 Ohms -- so the low end would be likely at 160M which is only 6X . You need to conduct some research beyond just building a DBM but exactly how do they work, By the way the 160uHy may NOT be signal loss but purely related to the core Al Value which would suggest for 0.160 Millihenry the Al would have to be 1600. So I don't know what know of any ferrite with that large of an Al value. If you want to homebrew this is where you need to add to your knowledge base. Pete 73's Pete N6QW

This is a passive mixer and part of the magic of the mixing process occurs in the transformer windings -- that is why the phasing is so important. Active mixers use a different process for mixing and thus may not require the transformers. The NE602 is a Gilbert Cell mixer and except for impedance matching or coupling to external circuits does not require the trifilar windings. That is about all I know so you may want to do some additional research on the subject.

Pete Juliano Thank you so much, If I want to buy a commercial type of Transformer, do you know what I should order? What is the specific commercial name for impedance matching transformer? there are a variety of transformer.

hossein jiryaei Take a short break and take another look at the 3 videos on Double Balanced Mixers. The videos tell you how to construct the transformers and the type of core material. I usually wind my own transformers and those videos give you all of the information that you need. I simply cannot recommend a commercial product to you and quite possibly you need to do more study of the subject. I have no idea of what is your application and you will either need to consult others or become more knowledgeable about the subject. There is nothing more I can provide you other than what you see in the videos. I am sure there are commercial products but that is for you to investigate. Pete

Very cool circuit. I thought making the 2 transformers would be hard, but compared to a lm1496 gilbert cell mixer using 13 resistors and 7 capacitors, its a whole lot easier!

The Title is DBM Part 3 of 3 and it is there

Well you are starting in a bit of a hole based on the following. If you go back over part I of 3 of the DBM I mention that the double balanced mixer we are using is a GAIN loss device with as much as a 5 to 6 dB conversion loss so an RF amplifier stage is very likely required. The RF amp is un-tuned and contains but a few parts. There will be an upcoming video on that piece. Not complicated and as I said a few parts. While you could bypass the Band Pass Filter --the DBM used as a Product Detector will be as a broad as a barn door --so for serious listening beyond just getting to work, the BPF should be in the mix. The knowing it if works is pretty much is a go no go --you will either hear signals or you won't. Now that is just getting it to work -- getting it to work at peak performance may be an entirely different matter that would entail being able to make measurements. So hook it up and if you are lucky you may hear signals. But I would strongly suggest you go back over all of the videos and get a really good feel what is happening in each block before you just start soldering parts. There is a lot of info on the internet about DCR's and you should read up on those as well.

Pete Juliano It seems that I might should start with an already designed kit. I've built a Softrock Ensemble, but find that I desire a robust SSB receiver. Is it too much to ask from a rockmite (or similar ne602 based cw txr) to widen its receive bandwidth to pass SSB? all the kits that I can find that are SSB ready are $80+ -- too high for my experiments at this time. thank you in advance for your time, i'm still enjoying your videos ! frank

Hi Frank, Thank you for your comment and input. You ask an interesting question. If the basic radios you mention DO NOT contain any sort of a narrow crystal or audio filtering then it should be fairly easy to copy SSB on these radios. The difficulty with the rockmite is that it uses a crystal to set the receive frequency and there is a very narrow Delta tuning with that crystal. AND the crystal is in the CW portion of the band. One thing you could do is to build a VFO and substitute that for the crystal in the rockmite and then you can tune the CW as well as the SSB portion of the band. The NE602 offers more possibilities in that with pins 6 and 7 you can hook up a tank circuit and you now have a VFO -install a tuned network on the front end (Pin 1 I think) and an audio amp off of either pins 4 or 5 and you have a direct conversion receiver not unlike you have seen in the videos. So I think what you are asking can be done very inexpensively! Do an internet search on the NE602 as a direct conversion receiver and I think you will find many radios that can be built for less than a $20 bill. I am delighted that you are taking up the iron and building something as that is how one learns. If you have more questions you can contact me directly at radioguy90 at hotmail dot com. 73’s Pete N6QW