UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 9th Apr 2021, 2:52 pm   #1
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default DIY FM tuner with 6CW4 nuvistor and ECC85

This is my first attempt to build a FM tube tuner and is a work in progress. It uses a 6CW4 and a pair of ECC85. The 6CW4 is not really essential but it will make the receiver more selective;and the tiny nuvistor looks "cute". The tuner head is supposed to be mounted on a bigger chassis made out of aluminium.

At the start, I knew it would be significantly more challenging than building MW tube receivers. Since there is no book or web site telling me how to build an FM tuner from scratch, I had to take a deep dive, reading every textbook on FM receivers design theory and studying the schematics of many existing tuners. I don't think many people are interested in DIY FM tube receivers. It seems the majority of hobbyists are interested in homebrewing HAM CW and SSB circuits.

Initially, I was going to use a pair of ECC85 as RF, mixer, oscillator and AFC using a twin gang air variable capacitor. But I have a triple gang variable cap from Ebay of unknown origin. So I have settled with 6CW4 as ground cathode RF tuned amp, and ECC85 as the untuned ground grid RF amp.

I cannot use the three-point tracking solutions for MW receivers because FM's tuning capacitors have a much small Cmax/Cmin capacitance ratio, and a padder is often not used. I derived a simple compatibility equation to allow me to find the trimmer values for my triple gang air variable capacitor so that it tracks perfect perfectly across the whole FM band for a DIY set of RF and oscillator coils.

At the moment, I am still debugging the tuner head. At first, the oscillator is not oscillating when I wired it in Hartley configuration. So I re-wired it in self-excited configuration (no feedback coil; it is magic) The oscillator is now oscillating exactly 10.7MHz below the RF. I have been running into problems with the plate currents for both ECC85 and 6CW4 are abnormally high; so a lot of head-scratching wondering if I got the values of cathode bias and plate resistors correct.

My biggest problem is to find the motivation to build the remaining IF amp, limiter, detector mounting on an aluminium chassis. It is a difficult project and it is very easy to get discouraged. Lately, I got distracted and I put this project aside for a month. I have been building shortwave and FM solid-state kit receivers. More worrying is that I start to get interested in Ham radio technology..I need to get myself a kick in the butt to finish this FM project.

P.S. I was toying with the idea of replacing the conventional ratio or discriminator with the 6BN6 gate beam tube as a quadrature detector. The tune 10.7mHz coil can be made form modifying an 10.7MHz IF.
Attached Thumbnails
Click image for larger version

Name:	FM1.jpg
Views:	746
Size:	101.3 KB
ID:	231261   Click image for larger version

Name:	FM2.jpg
Views:	577
Size:	129.7 KB
ID:	231262   Click image for larger version

Name:	FM3.jpg
Views:	471
Size:	84.3 KB
ID:	231263  

Last edited by regenfreak; 9th Apr 2021 at 3:00 pm.
regenfreak is offline  
Old 9th Apr 2021, 3:19 pm   #2
dragonser
Heptode
 
dragonser's Avatar
 
Join Date: Aug 2009
Location: Carshalton, Surrey, UK.
Posts: 734
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Hi,
that looks like a really impressive project.
__________________
Regards Peter B
dragonser is offline  
Old 9th Apr 2021, 3:48 pm   #3
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

With a triode RF amp, the amount of gain you can get is a matter of how well you neutralise Cag. This is why cascodes are a common solution. The low value of Cag in the nuvistor will help, but not get you quite as far as neutralisation or a cascode will. Neutralisation is a relatively narrow band technique in many implementations and ratiometrically, the FM band is rather wide.

There are two sorts of ultimate FM tuner: One with lots of RF selectivity to survive when there are multiple strong broadcasters thundering away at you, and one with very low noise and the gain to bring up distant signals. You can't be really good at both of these at once, so some sort of compromise has to be chosen.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 9th Apr 2021, 5:53 pm   #4
agardiner
Octode
 
agardiner's Avatar
 
Join Date: Sep 2020
Location: Thetford, Norfolk, UK.
Posts: 1,731
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Well done so far and keep at it! Great project.
agardiner is offline  
Old 9th Apr 2021, 6:31 pm   #5
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Quote:
that looks like a really impressive project.
Thanks Peter. It is still a long way to go. However i have built a simple solid-state IC chip FM stereo decoder for it and it is working. I have plan to make it to stereo FM with MPX output once i get it working. The idea is to start with simple, two IF stages with the 2nd IF stage acting as both a IF amplifier and limiter. The 2nd IF stage acts an bandpass amplifier for weak signals but becomes an AM limiter with strong signals. This can be achieved by using low plate voltage and proper grid leak bias so it operates near cut-off.

Once I overcome all the hurdles, I can upgrade the tuner from mono to stereo by adding more IF stages. This explain why it is in module design that can be transferred to another chassis. For good audio fidelity, I will have to replace the mono IF transformers by dedicated FM stereo IF transfomers with broader bandwidth. It is not a problem as i have many chinese FM stereo IF lying around. Usually stereo FM IF have 3 to 4 stages at least where the last two IF stages work like progressive limiters and IF amps. I was studying the schematics of FM stereo tuners made by Kenwood, Fisher, Heathkit, Sherwood and HH Scott ectc.

Quote:
With a triode RF amp, the amount of gain you can get is a matter of how well you neutralise Cag. This is why cascodes are a common solution. The low value of Cag in the nuvistor will help, but not get you quite as far as neutralisation or a cascode will. Neutralisation is a relatively narrow band technique in many implementations and ratiometrically, the FM band is rather wide.

There are two sorts of ultimate FM tuner: One with lots of RF selectivity to survive when there are multiple strong broadcasters thundering away at you, and one with very low noise and the gain to bring up distant signals. You can't be really good at both of these at once, so some sort of compromise has to be chosen.
Yes David, it was you who mentioned me about nuvistor and cascodes in another post last year. So it is credit to you! Afterwards i was digging around and figuring how to build a pair of 6CW4 in cascodes. But i cannot find an example of schematic with component values. I do have a pair of 6BS8 and they are commonly used in cascode configuration.

I am trying to scan my hand drawn schematic in an A3 paper somehow.

For the ECC85, untuned RF amplifier, the grid is grounded so that the cathode is shielded by the grid, minmising Miller capacitance. It has self biasing resistor and a RF choke at the cathode. It has broader bandwidth, lower gain and lower noise ...thats what i read anyway.

The tuned RF stage with the 6CW4 employs grounded cathode amplifier that has much higher input impedance than the grounded grid amplifier. Therefore the LC tank has to be tapped to match the input impedance.

I often ponder why some high-end FM tuners have 4 to 6 gang of air variable capacitors, they must have killers selectivity in separating strong broadcast stations but I think each tuned RF stage introduce insertion losses so it compromises with lower sensitivity. I suppose when ceramic filters came along, they could introduce high selectivity in IF stages more economically. I mean it is expensive to manufacture 4 to 6 tuning gang variable capacitors.

Last edited by regenfreak; 9th Apr 2021 at 6:44 pm.
regenfreak is offline  
Old 9th Apr 2021, 6:33 pm   #6
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Quote:
Well done so far and keep at it! Great project
Thanks man. I need some motivation and need to get my act togethor
regenfreak is offline  
Old 9th Apr 2021, 7:13 pm   #7
Cruisin Marine
Heptode
 
Join Date: Jul 2018
Location: Worthing, West Sussex, UK.
Posts: 983
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Looks really cool- well done!
Cruisin Marine is offline  
Old 9th Apr 2021, 8:19 pm   #8
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Heavy RF fltering, with multiple stages serves to limit the total number of signals hitting the mixer, and on their peaks, driving it into non-linearity.

Ceramic IF filters live downstream of the mixer, so they can't reduce whatever sigals are tenderising the mixer. Ceramic filters in the IF provide the ultimate selectivity of the receiver, they shape the channel which is being received. Their job is to reject adjacent channels.

Because the RF filtering is on a higher frequency, they are inevitably harder to make narrow for a given technology. Moreover the RF filters have to tune to the frequencies of different stations so they can't use high-Q resonators like ceramic elements and quartz crystals.

That's the power of the superhet idea. You don't need to do channel defining selectivity up at RF frequencies, you just need to reject the mixer image, and reduce as much as possible the influx to the mixer. The real channel defining filter is at a lower and therefore easier frequency, and it's fixed. The superhet idea was so good that the man who'd invented and patented it was hounded to suicide by a wealthy industrialist who wanted it and had lots of company lawyers. The same inventor also came up with FM.

DAvid
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 9th Apr 2021, 9:47 pm   #9
Aub
Nonode
 
Aub's Avatar
 
Join Date: Mar 2005
Location: Nuneaton, Warwickshire, UK.
Posts: 2,034
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Looks great. Good luck with it.

Aub
__________________
Life's a long song, but the tune ends too soon for us all.
Aub is offline  
Old 9th Apr 2021, 10:18 pm   #10
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Quote:
Well done so far and keep at it!
Great projectLooks great. Good luck with it.
thanks both

Quote:
Heavy RF fltering, with multiple stages serves to limit the total number of signals hitting the mixer, and on their peaks, driving it into non-linearity.
This is very interesting point about non-linearity. I have been learning to build solid state AM and FM superhets with discrete components ( I am not good at the theory of transistors). Only a few days ago, I learnt that mixer is non-linear by nature ( i dont understand the complex mathematical theory behind mixing). In the mixer, the bipolar transistor has to be biased near the non-linear cut-off region of the Ic vs VCE characteristic curve. I thought this is wow, it has to be non-linear!


Quote:
Because the RF filtering is on a higher frequency, they are inevitably harder to make narrow for a given technology. Moreover the RF filters have to tune to the frequencies of different stations so they can't use high-Q resonators like ceramic elements and quartz crystals.
Emmm...this brings me a question: How come CW/SSB HF bandpass are used at RF frequencies but not in the IF frequencies?

For example, this is a two-stage Butterworth HF filter:

http://zl2ctm.blogspot.com/2019/

who showed how to do the calculations this in his youtube channel:

https://www.youtube.com/watch?v=beIUzEGuUTg

(I do not know much about Ham radio but I find myself reading about the design of HF bandpass filters..i find them interesting that i am going to build one. even I dont own a QRP or Ham transceiver.)
I guess at UHF frequencies, these bandpass filters become lossy with the toroid magnetic core losses and therefore cannot be used in FM.

From time to time, i see a high-Q ceramic BPF is used in the RF front end of solid-state FM receivers though.

Quote:
That's the power of the superhet idea. You don't need to do channel defining selectivity up at RF frequencies, you just need to reject the mixer image, and reduce as much as possible the influx to the mixer. The real channel defining filter is at a lower and therefore easier frequency, and it's fixed. The superhet idea was so good that the man who'd invented and patented it was hounded to suicide by a wealthy industrialist who wanted it and had lots of company lawyers. The same inventor also came up with FM.
David Sarnoff had blood in his hands. VNA and spectrum analyzers also use superhet topology. In dual conversion superhet, the low IF frequency is used to reject image and the higher IF to reject adjacent channel. Modern DSP radios often uses triple conversion of three IF frequencies; high, middle and low.
regenfreak is offline  
Old 10th Apr 2021, 3:23 am   #11
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

SSB and CW filters are sometimes made in the 'RF' frequency range, but they are placed in the IF section of receivers.

I have an Eddystone EA12, rather a glorious set, but an early example of a radio for SSB. It uses L-C filtering in the shape of a lot of IF transformers, but to be able to get LC filters narrow enough for SSB, the centre frequency of the filter has to be low. So the set is a double superhet with the final IF at 100kHz, chosen to make that LC SSB filter possible. The IF bandwidth is actually variable. A front panel control turns a shaft that via levers and linkages moves one coil in each IFT closer or further from its neighbour, thus varying the coupling. When you move it to the very narrowest extreme, a microswitch closes and brings a quartz crystal into play for the very narrow bandwidth wanted for CW. 100kHz was a common IF frequency for higher performance communications receivers.

So the mode you want to work sets the bandwidth you need. Then looking at the different filter technologies, you find there is a happy region of IF frequencies you can make good filters over for each mode and for each filter technology.

Quartz crystals have very, very high Q compared to the best we can do with LC circuits. This is both a blessing and a curse. They can make some of the narrowest filters, but they are very difficult to use for making wider filters. Narrower and wider are terms that depend on the centre frequency.

You can make quartz filters on 455kHz for CW and SSB bandwidths, but the crystals are large and therefore somewhat expensive. The big-three Japanese makers of amateur radio opted for a 9MHz IF for many of their sets (though the Americans did it first). SSB crystal filters are cheaper there, CW are a bit more difficult but not too bad. The big saving was that for an amateur band only receiver, it could be a single superhet, and their accountants could celebrate (probably with a new set of golf-clubs).

But their customers wanted general coverage receive in their HF sets. Drat! There went the single superhet architecture. So they mixed up to a high IF somewhere in the 35 to 70MHz region where new sorts of overtone crystal filters were becoming affordable, then they mixed back down to 9MHz and it was business as usual from there onwards and they could use their existing crystal filter parts. You may find these receivers called 'upconverters' because their first IF is higher than the highest input frequency they can tune to.

You should now be beginning to see that filters rule receiver design. You need filters to do various things and the available and affordable technologies limit what sort of frequency you can do them at. Consequently signals in receivers are mixed up and down willy-nilly to suit the filtering edicts. Oh, and those mixers require additional filters to block their images.

Enough for tonight. Give me a nudge and the next time I'm sleepless, I'll go over mixer images and the whole business of whether they are linear or non-linear.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 10th Apr 2021, 1:15 pm   #12
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Thanks David. I always learn something new in your posts
Quote:
have an Eddystone EA12, rather a glorious set,
I have downloaded the Eddystone EA12 manual. It is really nice manufacturers took the trouble to explain how each parts of the circuits work. I can see it uses a pair of variable Mu double triode EC189 wired in cascode RF amp configuration. It is neat that it has variable gain 10k pot for the first triode in grounded grid amp mode (C11 to RF ground). I dont own any vintage communication receiver or FM stereo tuner, but I have learnt a lot by reading manuals only.

Quote:
nough for tonight. Give me a nudge and the next time I'm sleepless, I'll go over mixer images and the whole business of whether they are linear or non-linear.
Now, the idea of non-linearity is good for mixing seems counter-intuitive. More interesting is that distortion means to be good for mixer. It got me head-scratching for a while. My understanding (please correct me if i am wrong) is that mixing is multiplication of two sine waves of different frequencies and amplitudes.
So what is wrong with a linear mixer?
In my homebrew FM tuner, one half of the double triode EEC85 is used as a mixer. In order to make it "non-linear", the cathode self-biasing resistor must be adjusted so that it operates in class C. The oscillator triode section creates a large oscillator voltage supplied to the grid of the mixer to turn the it on and off, creating clipping action. This chopping effect creates higher order harmonics with intermodulation frequencies of:

(integer m x oscillator frequency ) plus and minus( integer n x IF frequency)

At the moment, I am trying to debug my oscillator as I have suspicion is that the oscillator output voltage is too low in my homebrew.

Now this is hard part about non-linearity that blows my mind away: intermodulation distortion. It is a concept that is difficult to grasp as most textbooks threw in lines and lines of equations.

Quote:
ou can make quartz filters on 455kHz for CW and SSB bandwidths, but the crystals are large and therefore somewhat expensive. The big-three Japanese makers of amateur radio opted for a 9MHz IF for many of their sets (though the Americans did it first). SSB crystal filters are cheaper there, CW are a bit more difficult but not too bad. The big saving was that for an amateur band only receiver, it could be a single superhet, and their accountants could celebrate (probably with a new set of golf-clubs).
To fund the golf course membership of the Toko executives, most ham operators must be well-heeled individuals who likely paid off their mortgage, having big final salary pensions or alternatively they spend loads of money behind their wives' backs.. Lately. I am horrified to find how how expensive new Ham HF transceivers costing thousands of pounds.
regenfreak is offline  
Old 10th Apr 2021, 4:36 pm   #13
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

OK, mixing.... and you get to understand intermodulation as a free bonus. Intermodulation IS mixing, just when signals you want to sty separate get mixed together.

A 'Frequency Mixer' to give it its rarely used full name, is a machine for changing the frequency of a signal. It produces a simple linear shift in frequency. It needs a locally-generated signal to control the amount of offset.... the local oscillator.

The biggest user (but not only) user of mixers is a radio receiver, so a lot of mixer terminology is receiver oriented and can be confusing in other applications.

The typical mixer has three 'ports'. The incoming signal(s) is applied to the 'RF port" the frequency offset creating signal is applied to the 'LO port' (LO, Local oscillator) and the output comes out of the 'IF port'

A mixer is a machine which performs multiplication of one signal (RF) by another (LO)

Why this should do what we want does take some mild maths. You don't really need to understand it, but it helps if you see where it comes from.

In school maths you may have come across some equations usually referred to as trig identities. Most people are scared of trigonometry, but they're manageable and useful.

one of them reads something like sin(a+b) = sin(a).cos(b)+cos(a).sin(b)
and another is sin(a-b) = sin(a).cos(b)-cos(a).sin(b)

a and b are the instantaneous phase angles of the signals... which are continuously increasing at a rate set by the frequency, so we can say that a and b are the frequencies of the signals. cosine waves are like sinewaves, just shifted 90 degrees in phase.

Playing with these equations with bog standard school algebra (people are scared of that, too) and we can arrive at a statement that if we multiply two sinusoidal signals together, one at frequency a, one at frequency b, we get two new sinusoidal signals one at frequency (a+b) and the other at frequency (a-b). ONE of these is the frequency shift we bought the mixer to do. The other is called the image and can be a problem unless it's handled properly.

WHICH one you want and which is the image is your own choice. All mixers always make both. You thus have to sort them out. Mixers tend to hang out with filters needed to block images.

You can view a mixer as a mathematical multiplier of the voltages of two inputs.
You can view a mixer as a variable gain amplifier where the voltage on one port controls the gain the signal passing from the other port to the output experiences.

Wild, eh?

Just remember that when two signals are mixed together, you get both the sum and the difference frequencies made from them. One you want, the other you don't.

The sum and difference thing works another way. If I have an IF at 1MHz and I want to mix a 10MHz signal into it, I can use a 9MHz local oscillator (10MHz-9MHz = 1MHz) so difference mixing does the job. It will also make 9+10 = 19MHz, so I need a filter to remove that. But wait, if any 8 MHz enters the mixer, that's also 1MHz difference form the LO, and the mixer turns it into 1MHZ.

So our receiver tries to receive on two frequencies at once. We rarely want this. So hello to another filter to block that image.

In a good mixer, if there are multiple signal components going into the RF port, we want each to be mixed individually, and each to come out on the appropriate, shifted frequency at the IF port. Some people say that the multiplication activity is non linear. It isn't necessarily, but it often is implemented using the non-linearity of devices so most folk think mixers are non-linear.

However, with multiple signals coming in, we most definitely do not want them to start mixing amongst themselves. This would be due to say a non-linear effect in the path from RF to IF ports. Of course a non-linearity there would also generate harmonics of all the signal components AND allow all of them to mix with all of the others. This is a mess.

Unfortunately some of these harmonic mixing products lie very close to the wanted output frequency.

This is intermodulation at work. It's just uncontrolled mixing on the path you wanted to be nise and linear. Amplifiers, mixers, mag cores, crystals, they all do intermodulation to some extent.

The only solution is to make things linear enough that the mess of intermod products are small enough to be lost in the noise.

This takes very artful RF design.

enough for now

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 10th Apr 2021, 5:50 pm   #14
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,951
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Quote:
Originally Posted by regenfreak View Post
Lately. I am horrified to find how how expensive new Ham HF transceivers costing thousands of pounds.
Yet do the calculations equating those "thousands of pounds" as a percentage of current average income, and then repeat the calculations to compare the cost of something like a KW-2000, an Eddystone 888/EA12 receiver or a Yaesu FT101 as percentage of the average-income-of-the-day when said radios were introduced, and you'll find not much has changed.

Apart, that is, for the sensitivity/selectivity/frequency-accuracy-and-stability of the equipment!

In the 70s I built a neutralised Nuvistor preamp for 144MHz using a 6CW4 - the design was in the RSGB Handbook. It worked rather well, though it was still significantly noisier than a solid-state equivalent using an AF239 Germanium UHF-TV-tuner transistor or a cascode using 2N3819 FETs.

Neutralising the Nuvistor preamp was interesting: the recommended method-of-the-day was to set it up backwards: Output-socket to a local signal-source, input-socket to the input of your receiver. You then adjusted the neutralising capacitor for minimum received signal. After which you turned it right-way-round and off you went - always noting that maximum gain and lowest noise-figure did not coincide - it was often worth sacrificing a few dB of pure gain in exchange for a reduced noise-figure. Adjusting the Nuvistor's HT voltage [from memory the 'default' voltage being around 80V] up and down could help with optimisation too.
G6Tanuki is offline  
Old 10th Apr 2021, 7:04 pm   #15
Trevor
Octode
 
Join Date: Aug 2006
Location: Stoke-on-Trent, Staffordshire, UK.
Posts: 1,082
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

your tuner looks great any chance of a schematic

Trev
Trevor is offline  
Old 10th Apr 2021, 9:29 pm   #16
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Thanks David for informative post as usual. You have a gift for explaining complex concept in the most understandable language. You should be a university professor.

Quote:
laying with these equations with bog standard school algebra (people are scared of that, too) and we can arrive at a statement that if we multiply two sinusoidal signals together, one at frequency a, one at frequency b, we get two new sinusoidal signals one at frequency (a+b) and the other at frequency (a-b). ONE of these is the frequency shift we bought the mixer to do. The other is called the image and can be a problem unless it's handled properly. Wild, eh?

Just remember that when two signals are mixed together, you get both the sum and the difference frequencies made from them. One you want, the other you don't.

The sum and difference thing works another way. If I have an IF at 1MHz and I want to mix a 10MHz signal into it, I can use a 9MHz local oscillator (10MHz-9MHz = 1MHz) so difference mixing does the job. It will also make 9+10 = 19MHz, so I need a filter to remove that. But wait, if any 8 MHz enters the mixer, that's also 1MHz difference form the LO, and the mixer turns it into 1MHZ.
I have drawn the 2nd order intermodulation products and undesirable harmonics in a spectrum plot (attached) using your example. This helps me to visualize the concept better, otherwise it is a bit like like decoding occult numerology. I have left out 3rd order harmonics...
Mixer seems a hard subject. I dont want to think about ring diode mixer used in SSB receivers.

Quote:
In the 70s I built a neutralised Nuvistor preamp for 144MHz using a 6CW4 - the design was in the RSGB Handbook. It worked rather well, though it was still significantly noisier than a solid-state equivalent using an AF239 Germanium UHF-TV-tuner transistor or a cascode using 2N3819 FETs
This is interesting. The circuit i saw used an inductor in series with the neutralising capacitor bridge. The noise was adjusted by changing the inductance. I do not know how a triode could be operated in reverse with signal goes to the anode and output from the grid. For FM the 6CW4 is happy without neutralisation.

Quote:
your tuner looks great any chance of a schematic
Thanks. I have attached my scruffy schematic with scribbles (i did not expect anyone would be interested to look at it). I have trouble of uploading the schematics. My image is very long and the forum web software automatically scales it down to very small. Every time I upload a readable file, it scales down to very small even i have chopped it into two parts.. It is out of my control.
The design is very simple and conventional using a ratio detector. The 6CW4 can be left out completely. The de-emphasis RC network can be changed either UK (50 mirco seconds)or US (75 micro seconds) standard easily. I use time constant of 2.35 micro second for the 2nd If/limiter stage.

At the start I studied the German FM tuner designs and it killed me because they were fond of complex permeability tuners..i could not understand the nuances of the highly elaborated designs. I can DIY something only if I can understand how the circuits work.


The second picture is two possible oscillator configurations: 1. Hartley, 2. self-excitation mode. Initially I used Hartley with 1 turn tapping but it did not oscillate. I could try 2 turns tapping but I switched to self-excitation configuration. At the moment, the oscillator output is a bit feeble. I think the RFC at the cathode contributes to the positive feedback of the oscillator somewhat like the RFC in FM one-valve super-regenerative receiver (It has to made out of air core high Q RFC promoting oscillation of the quenching frequency. The working principle of super-regenerative receiver is very complex). I think my mistakes was that I used ferrite choke instead of air core choke. So I am going to replace them.
Murphy law rules in debugging phase.
Attached Thumbnails
Click image for larger version

Name:	intermodulation.jpg
Views:	195
Size:	44.6 KB
ID:	231415   Click image for larger version

Name:	oscillators.jpg
Views:	224
Size:	44.9 KB
ID:	231417   Click image for larger version

Name:	DIY FM part 1.jpg
Views:	242
Size:	53.3 KB
ID:	231419   Click image for larger version

Name:	ursine FM part 2.jpg
Views:	354
Size:	138.1 KB
ID:	231420  

Last edited by regenfreak; 10th Apr 2021 at 9:56 pm.
regenfreak is offline  
Old 11th Apr 2021, 3:08 am   #17
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Just a few odd comments.

The diode ring mixer is the nice one, much more easy to understand than the others. The others have all sorts of foibles and difficulties that you have to get your head around. It's one of a class called 'double balanced' mixers. An unbalanced mixer leaks plenty of the RF and LO inputs out with the IF products. A single balanced mixer acts to null one of them, and you get some compensation of non-linearity. A double balanced mixer acts to null both leakage paths, and gives a bit better compensation of non-linearity.

It is routine to drive diode ring mixers very hard with lots of LO signal, and then you can view, for a first understanding, the diodes as being turned hard on and hard off by the domineering LO. So the mixer becomes an inverting switch to the RF to IF path, controlled by the LO. Mathematically, the RF signal gets multiplied by +1 or -1 and you have a switching mixer.

Intermods, don't stop at 2nd order harmonics. The excrement hits the fan when you consider the 3rd harmonic of one signal mixing with the second of the other and the two spurs created flank the original signals at the same spacing which puts them in the right place for maximum trouble. 4th and 3rd put spurs one more spacing out on either side of those.... and so on. It all gets very messy very quickly and the only escape is to try to make things as linear as possible and to run the signal levels low enough for the spurs to be lost in the noise.

The diode ring mixer is rare in domestic stuff because you get about 6 dB loss in them, while transistor and valve mixers can give a bit of gain. So a diode ring usually means you need an extra stage of gain along the line, but they do tend to out-perform active mixers in aareas harder to fix than mere lack of gain. It all comes down to pennies. A couple of American FM tuners had diode ring mixers.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 11th Apr 2021, 11:54 am   #18
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Quote:
The diode ring mixer is the nice one, much more easy to understand than the others. The others have all sorts of foibles and difficulties that you have to get your head around. It's one of a class called 'double balanced' mixers. An unbalanced mixer leaks plenty of the RF and LO inputs out with the IF products. A single balanced mixer acts to null one of them, and you get some compensation of non-linearity. A double balanced mixer acts to null both leakage paths, and gives a bit better compensation of non-linearity.

It is routine to drive diode ring mixers very hard with lots of LO signal, and then you can view, for a first understanding, the diodes as being turned hard on and hard off by the domineering LO. So the mixer becomes an inverting switch to the RF to IF path, controlled by the LO. Mathematically, the RF signal gets multiplied by +1 or -1 and you have a switching mixer.
So the goal is suppress both the strong LO and RF as much as possible, leaving mostly IF in the output. So linearity is not a bad thing right? You want more linearity in a non-linear mixing device. I guess the input and output baluns are simple broadband impedance matching transformers optimised for minimum core losses. It looks so much like a full wave rectifier circuit. (Lately I am interested in the design of bandboard balun or unun because i do not know if it is possible to have a 50 ohms input matched to a high impedance LC resonance (10k-100k). I only see 64:1 balun and never any higher ratio. I guess any higher ratio will introduce unacceptable insertion losses.

Quote:
Intermods, don't stop at 2nd order harmonics. The excrement hits the fan when you consider the 3rd harmonic of one signal mixing with the second of the other and the two spurs created flank the original signals at the same spacing which puts them in the right place for maximum trouble. 4th and 3rd put spurs one more spacing out on either side of those.... and so on. It all gets very messy very quickly and the only escape is to try to make things as linear as possible and to run the signal levels low enough for the spurs to be lost in the noise.

The diode ring mixer is rare in domestic stuff because you get about 6 dB loss in them, while transistor and valve mixers can give a bit of gain. So a diode ring usually means you need an extra stage of gain along the line, but they do tend to out-perform active mixers in aareas harder to fix than mere lack of gain. It all comes down to pennies. A couple of American FM tuners had diode ring mixers.
I like your figurative language. D. There seem to be differences in "nastiness" between even- and odd-order intermodulation products. Did folklore said odd numbers are dodgy, evens are good?

The reason I thought about ring diode mixer is because i saw his SSB homebrew rig a few days ago and I was wondering what advantages it offers:

https://www.youtube.com/watch?v=w0baCsMlZEA&t=67s
regenfreak is offline  
Old 11th Apr 2021, 4:45 pm   #19
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

If you pick two loud signals, say 100 and 101 MHz

Mix them together in a non-linear amp or mixer and you get some 1MHz
Mix the 2nd harm of 101 (=202) with the fundamental of 100 and you get some 102 MHz which is in band and therefore a lot morr trouble.
Mix the 2nD harm of 100 (=200) with 101 and you get some 99 which is also trouble.

Mix the 2nd of 101 with the 2nd of100 and you get some 2MHz and some 402, both well out of band.

Short cut: take the two harmonic numbers (like 1 and 2 in the example above) add them together. If the result is odd, it's an odd order intermod and trouble. If the result is even, it's an even order intermod and likely far enough away to be out of trouble except for oddball mixer applications.

So the 101 +100 intermodding to 99 and 102 is called the third order intermod products. As a rule of thumb, if you reduce both your strong signals by 1dB, the third order products will drop 3dB, the 5the order by 5dB, the 7th by 7dB. You see the pattern. This is NOT a hard and fast ruling but it's pretty close on where the prroducts are small. When the intermods get close to the top of your range the cosy relationship can get broken.

It comes out of breaking down the non-linear curvature of a device mathematically into the sum of a power series. (Mathematical powers like square, cube, etc)

This illustrates how distortion products shrink much better than pro-rata as you reduce signal levels.

The folklore-y bit is purely an audio thing about harmonic distortion. People in general tend to find even order distortion products not as distracting as odd order ones. However, higher order distortions sound bad no matter whether even or odd.

But as your ears don't hear RF frequencies, the issue with mixers doesn't have to follow this. It's more of a coincidence that the odd order intermods turn up in the part of the spectrum where your signals reside. Imagine truing to listen to a distant station on 102 MHz with strong nearby ones on 100 and 101.

This is far more serious than is obvious.

Some of the intermod products of FM channels in use fall into the aviation band.... and the worst-hit part of it is the lower end, 108-118MHz Used for instrument landing systems and auto-land as needed in fog and low cloud. Could you think of anything worse to interfere with? I want a new word inthe English language "Pessimised" being the opposite of optimised.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 11th Apr 2021, 6:11 pm   #20
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: DIY FM tuner with 6CW4 nuvistor and ECC85

Thanks David. I think i understand it after translating your sentences into a spectrum sketch (see attached). I am a visual thinker. I understand better by drawing.

Quote:
It comes out of breaking down the non-linear curvature of a device mathematically into the sum of a power series. (Mathematical powers like square, cube, etc)
It almost feel like hearing Einstein theory of relativity saying gravity is the result of the curvature between space and time.

Quote:
Some of the intermod products of FM channels in use fall into the aviation band.... and the worst-hit part of it is the lower end, 108-118MHz Used for instrument landing systems and auto-land as needed in fog and low cloud. Could you think of anything worse to interfere with? I want a new word inthe English language "Pessimised" being the opposite of optimised.
"Pessimised" is an anagrams of semi-pixxed I must confess I am incredibly ignorant of Instrument Landing System (ILS) frequencies. Did I learn or remember a single thing in my Avionics lectures in the 80s?
Attached Thumbnails
Click image for larger version

Name:	intermod-spectrum.jpg
Views:	102
Size:	27.5 KB
ID:	231498  
regenfreak is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 8:15 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.