It is currently Fri Dec 15, 2017 7:35 am

Minimalist Superhet Radio.

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sat Nov 09, 2013 1:13 pm

Hi Rich
That's strange, it was difficult to make the instability happen today. But I peaked the aerial trimmer a little and made it happen. It is worse with the AGC disconnected.
I have had a degree of success, I think it is associated with the AGC wire, I have moved it and it is even more difficult to make happen. It seems best with the AGC wire positioned so that it passes close by the MW aerial coil.
In the photo the red line shows the path of the the AGC wire, it connects from R14 which is on the component board to pin 2 on V1. The Green line shows the path I have left it on.

Do you think it would be better if I mounted R14 on the other end of the wire so that it is soldered direct to pin 2 of V1. If it is picking up stray signals from the oscillator maybe the 1M resistor being on the valve pin will reduce the effect.

Mike
Attachments
AGC wire.jpg

 
Posts: 3712
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: Minimalist Superhet Radio.

Post by Terrykc » Sat Nov 09, 2013 1:53 pm

Crackle wrote:... Do you think it would be better if I mounted R14 on the other end of the wire so that it is soldered direct to pin 2 of V1 ...

Yes, Mike!

Funnily enough, in my previous post I meant to suggest investigating the anode and grid circuitry layout to minimise any possibility of unwanted feedback ...

I hadn't watched the video then ...

... all that gleaming ali! Doesn't it really look good?

You've done a splendid job so far - I hope you are close to a final solution - it is really something to be proud of!

One thing that worries me, though, is the hand capacity effect which doesn't seem to matter where you touch anything that is supposed to be connected to ground ...

Pending any other solution, have you considered putting a mains earth on it to see what happens?

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sat Nov 09, 2013 2:18 pm

Hi Terry
Thanks for your kind comments.
There was a mains earth fitted, by the time we took the movie, to the single earth on the chassis and from there to the mains TX screen. The capacitive effect on the controls may be because the RF section earth is not attached to chassis it goes back to the AF section and joins the earth and chassis there close to where the smoothing capacitors are.
Do you think connecting the RF section earth to chassis will help the effect on the control spindles.
I will move R14 and see if that helps.
Thanks
Mike

 
Posts: 3712
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: Minimalist Superhet Radio.

Post by Terrykc » Sat Nov 09, 2013 3:32 pm

Crackle wrote:.. The capacitive effect on the controls may be because the RF section earth is not attached to chassis it goes back to the AF section and joins the earth and chassis there close to where the smoothing capacitors are ...

Ah! Too much experience of audio amplifier layout with all those guitar amplifiers, methinks!

RF is different. You want all earth paths to be as short and direct as possible.

The ideal is a star ground like this ...

Star_Ground.PNG


... although it is normally used for RF equipment working at much higher frequencies. Put some solder tags under the valve-holder fixings and take your wires to the nearest one. Any length of wire is an inductor and the longer the wire, the higher the inductance. Think of it as putting an RF choke in series with all your components ...

The clincher, in your case, is the biggest inductor which will be common to everything else - your common bus - which is effectively connecting all your RF decoupling together and isolating it from true earth!

Crackle wrote:... I will move R14 and see if that helps ...

Yes! Within reason, you want components mounted as close to the valve pins as possible to prevent unwanted radiation and pick-up.

Using my marvellous powers of hindsight, it is a pity due didn't have antbody looking over your shoulder to see what you were doing! It might have saved you a lot of work, some of which will doubtless be tricky in the confined space of a fully built chassis ...

By the way, as drawn, your aerial coils are still transposed ...

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sat Nov 09, 2013 5:45 pm

hi all
I have made some further improvement.
Moving R14 helped with the instability.
I cut about 2 inches of the AGC wire from V1 grid 1 to the component board.
I cut about 2 inches of the wires to the tuning capacitor fixed vanes.
All the above has helped, and after re-aligning on MW is far less noticeable that there is higher sensitivity in one part of the dial, with more sensitivity than before, all across the MW band.
The instability is now only occurring at the very bottom of the band (below MW) where I wont normally go.
The high end of MW is very good but it seems the tuning is even more sensitive now and a little tricky to centre. I wonder if this will help separate some of the stations which before tended to overlap at night.
Studying the Ferranti circuit again, I found I had left of a fixed LW oscillator trimmer capacitor, now all the long wave stations are spread across the dial as they should be, and are not all crammed into about an inch of the dial.

Terry,
From the manufacturers service sheets, Ferranti has listed L4 (the aerial coupling coil primary) as 30 ohms, the secondary, L5 is 2.5 ohms.
I had it the other way around to start with and could not get any "peak" out of the aerial coil.
The radio is really performing well now.

It seems like it has been a long process, but I now think the circuit is as good as it is going to be and it really performs well.
Tomorrow I will strip the cabinet and see what it looks (how many bad scratches), I haven't yet decided whether to oil it or varnish it.

Thanks guys for all your helpful advice and encouragement, this has been a good learning experience.
Mike

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sat Nov 09, 2013 6:30 pm

I forgot to mention I connected the RF earth point to chassis and that has helped the sensitivity when touching the chassis and controls.

 
Posts: 2853
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: Minimalist Superhet Radio.

Post by sideband » Sat Nov 09, 2013 10:08 pm

With regard to earthing RF, the centre spigot of the valveholders is used for this very purpose. If you look at some of the photos of mine, you will see how I used the centre spigots for earthing.


Rich.

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sun Nov 10, 2013 7:54 am

Is the ratio of turns important for tuning purposes or is it just going to effect the impedance matching of each winding.

If it does not effect the tuning so much, could it be that the smaller winding is in actual fact as big as a "normal" coil, to match with the condenser, but the Ferranti has an even bigger aerial winding on the coil.

I agree it doesn't make sense because the Ferranti coils would tend to "step down" the signal.

Mike

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sun Nov 10, 2013 10:48 am

Hi Trevor
I posted the circuit of the Ferranti 345 in this post here. Its just below my circuit with the red component values.
The coil data is in this post
Thanks
Mike

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Minimalist Superhet Radio.

Post by Michael Watterson » Sun Nov 10, 2013 3:26 pm

It's best to either design coils or use new coils of known characteristic for "home brew" Radio.

A) Some radio sets are slightly strange.
B) Not all sets used "sensible" coils either.
C) The RF tuning coil can be for a particular loop aerial, short wire or long wire depending on period of time.

Reusing old coils:
Also to consider is what series and parallel capacitors (and/or valve capacitances) originally used, not just the band and tuning capacitor range and circuit topology of original.

Also to reuse old coils you really need a good Dip Meter and RF inductance meter as well as measuring resistance of windings. The winding design also affects the Q and self capacitance as well as the actual type of wire. Actually, Litz vs solid isn't such a huge issue as is often made out. Sometimes wire is fabric covered and NOT Litz because it's more stable to wave wind a large narrow coil or a LW coil.

Maplin has 40 SWG reels of wire, 250g, which I calculate is about 2.2km length and about 1.4 Ohm per meter. That's 0.125mm (i.e. 8 turns per cm length max) and about 36 AWG. It's a good general purpose for small mains or audio transformer primaries and IF, MW & LW coils. For higher Q on longer coils simply use 4 or 8 strands in parallel (much cheaper and same effect as Litz). For longer bobbins of mains/audio primaries or high voltage secondaries then spray each layer with polyurethane (it's polyurethane based enamel) and soak bobbin in warm polyurethane varnish before mounting on core.

A single sub-miniature AM Radio Aerial coil (£2 to £5 radios) on rod makes a transistor IFT (small winding to base via cap) and pair 1cm apart make a good valve IFT (ignore small winding). Use 330pF to 560pF in parallel depending on desired IF frequency. Tune either with 50pF presets or by moving core. Used with 1/2 to 2/3rd out core they will usually work as a LO transformer (or fully in for modulator). Small winding coupled to g2 or anode via cap large winding to grid, or small winding to cathode and larger one tuned on g2 or anode. Tune with 10 to 275pF.

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sun Nov 10, 2013 3:52 pm

Hi
Just out of curiosity I have measured the MW aerial coil again.
The primary (aerial) measures 5.25mH & 32.3ohms.
The secondary (tuning cap) measures 0.19mH & 2.6ohms.

Mike

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Minimalist Superhet Radio.

Post by Michael Watterson » Sun Nov 10, 2013 5:14 pm

It's a LW coil the wrong way round by those figures!

 
Posts: 3712
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: Minimalist Superhet Radio.

Post by Terrykc » Sun Nov 10, 2013 5:27 pm

Ah! Now that could explain something ...!

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Sun Nov 10, 2013 5:33 pm

Crackle wrote:Hi Trevor
I posted the circuit of the Ferranti 345 in this post here. Its just below my circuit with the red component values.
The coil data is in this post
Thanks
Mike


Hi Michael
Ferranti listed this coil as a MW coil on the manufacturers service sheet, see links in quote above. The LW coil is even bigger.
Would having the 100pf cap in series with the tuning capacitor and coil account for this appearing to be odd.

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Minimalist Superhet Radio.

Post by Michael Watterson » Sun Nov 10, 2013 9:07 pm

I know you did. I'm echoing!

If it's a MW coil and correct way round then it's for a 500pF capacitor (or more!) and for an odd aerial arrangement.

I use about 0.5mH for MW and 2mH for LW

 
Posts: 2839
Joined: Tue Jul 26, 2011 5:30 pm
Location: Gateshead

Re: Minimalist Superhet Radio.

Post by Till Eulenspiegel » Thu Nov 21, 2013 12:27 pm

The circuit diagram of the Champion Mini-ette. The triode section of the UCL83 is operating as a leaky grid detector.

Till Eulenspiegel.
Attachments
Champion851-1.jpg
Champion851-2.jpg

 
Posts: 2839
Joined: Tue Jul 26, 2011 5:30 pm
Location: Gateshead

Re: Minimalist Superhet Radio.

Post by Till Eulenspiegel » Sun Dec 08, 2013 9:22 pm

Hi Trevor,
That'll be an infinite impedance detector. It resembles a cathode follower and like a cathode follower it has no gain. In fact gain is less than unity. However, like the anode bend detector it imposes hardly any loading on the preceding circuits. The cathode resistor has a high value > 50Kohms.
Low distortion, good for HiFi AM radios

From this discussion group: http://amfone.net/Amforum/index.php?topic=26132.0

Worth noting that a FET can be used instead of a valve.

Till Eulenspiegel.

 
Posts: 2839
Joined: Tue Jul 26, 2011 5:30 pm
Location: Gateshead

Re: Minimalist Superhet Radio.

Post by Till Eulenspiegel » Sun Dec 08, 2013 9:33 pm

A follow up to my last post on this subject. Those AMfone guys talk about full wave detectors for AM radio. It works for TV video detectors so I guess a FW detector will fine for AM radio although I'm not sure if there will be any advantages.

Till Eulenspiegel.

 
Posts: 4092
Joined: Wed May 02, 2012 3:35 pm
Location: Worksop

Re: Minimalist Superhet Radio.

Post by Refugee » Sun Dec 08, 2013 11:56 pm

It would double the carrier ripple frequency and require two coupling capacitors. This would make it better for a low carrier and high modulation frequency.
It would have a higher frequency response and that would help explain why it was tried with video detection.
Perhaps worth a try for a LW set.

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Thu Dec 19, 2013 9:57 am

Hi
Following some advice from Rich regarding the problems I was experiencing trying to calibrate my dial, I turned the amplitude from the signal generator right down to minimum and went through calibrating the dial again. I found that the higher frequency "image" was the strongest all the way down through the SW frequency band to the last 3 MHz, where it became impossible to tell any difference between the 2 images and on the lowest frequency, 6.5MHz, the lower image was slightly stronger at this point I also had to turn up the amplitude a little, but I continued to use the upper image.
This is an image of the result of my calibrations, I know its not perfectly aligned with established LW and MW bands but this seems to be the best compromise for gain and the unwanted instability, which now only occurs at the very bottom of the MW band. SW reception is pretty good on a long wire. LW is receiving quite a number of French sounding stations as well as R4. Reception on MW is now quite good with fairly even sensivity across the band. The only problem is my favourite station "Gold" sounds a little distorted, the rest all sound fine, I wonder why this should be, it is not the strongest signal, and I have noticed this on some other valve radios.
dial.png

The tuning drive was rearranged to get rid of the slow mo drive which I originally fitted by mistake on the capacitor instead of the control spindle. I made a drawing for the new drive drum and asked a friend to make it in his workshop. The result is I now have a 7:1 turns ratio drive, (3.5 turns of the knob for full 180 deg rotation of the capacitor. This gives me a scale length of about 10cm, just slightly more than the aperture size in the cabinet.
tuning drive.png
tuning drive1.jpg
tuning drive2.jpg
tuning drive3.jpg
refurb cabinet.jpg



There was a strange anomaly around 240kHz on the LW scale where the marks were unequally spaced, but I ignored this on the final drawing.
dial2.png
dial2.png (5.71 KiB) Viewed 2240 times

Chris
I wonder why this image is shown in a window, when the others are not?


Mike

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Minimalist Superhet Radio.

Post by Michael Watterson » Thu Dec 19, 2013 10:21 am

If an image is narrow enough to fit the SW seems to ignore height sometimes and puts it actual size in window rather than clickable preview. This seems to be more frequent with PNG than JPG

 
Posts: 2853
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: Minimalist Superhet Radio.

Post by sideband » Thu Dec 19, 2013 11:11 am

Crackle wrote:The only problem is my favourite station "Gold" sounds a little distorted, the rest all sound fine, I wonder why this should be, it is not the strongest signal, and I have noticed this on some other valve radios.


This could well be due to the fact that Gold and similar stations use a high level of modulation. I've also noticed distortion on this station as well. I think it used to be convention that AM modulation was 30% of carrier amplitude...certainly that figure sticks in my mind. Now a lot of local stations like gold use lower powers but tend to use the same frequency so several stations will broadcast on 1548 for example but with low power, interference is kept to a minimum. However the modulation levels are increased to give more 'punch' to an otherwise low-power station. I was told that the modulation levels are as high as 80%. Some detectors don't operate so well with the increased modulation and tend to distort and it seems that the anode bend is one of those detectors. Perhaps some adjustment of the bias on the EF80 will improve things so may have to look into that.

Basically if a radio station is using 30% modulation and the detector is biased to remove 70% of the carrier, leaving only 30% with the modulation in tact then that's fine. However if the radio station is using 50% modulation, the detector is still going to remove 70% of the carrier, but some of the modulated part will go as well leaving a distorted signal...well that's how I understand it!

The same thing happens with my Philips Superinductance set which also uses anode bend detection but I don't want to alter anything on a 75 year old radio!


Rich

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Thu Dec 19, 2013 12:11 pm

On a good transistor radio I can pick up about 4 or 5 Gold signals across the dial and they all seem to have a different tonal quality. I also believe some change after dark and sound thinner. Is it possible that they change the modulation to reduce co-channel interference with other stations at night.
Mike

 
Posts: 6972
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: Minimalist Superhet Radio.

Post by Michael Watterson » Thu Dec 19, 2013 1:00 pm

sideband wrote:
Crackle wrote:The only problem is my favourite station "Gold" sounds a little distorted ...


This could well be due to the fact that Gold and similar stations use a high level of modulation. I've also noticed distortion on this station as well. I think it used to be convention that AM modulation was 30% of carrier amplitude...


It's even more complicated. They vary the carrier power dynamically to save electricity. This causes extra distortion on all but synchronous (or carefully tuned SSB) demodulators. I can hear the difference on my ICF2001D very clearly when this is done too aggressively as it has diode AM, Synchronous AM and SSB, with narrow or wide filter.

An anode bend detector is distorting anyway, so it becomes progressively worse with this kind of modulation. There are diode detectors that are more advanced than simple Diode + R C filter that give lower distortion with traditional AM and the adaptive carrier systems. You don't want too fast AGC either with modern adaptive carrier AM.

The conventional diode detector is better than Anode Bend and doesn't need to be "loading" the IFT. But even with ordinary AM it has two kinds of distortion. It's best as a detector for AGC!
http://www.st-andrews.ac.uk/~www_pa/Sco ... page2.html

A synchronous detector for AM is usually by ICs, but in theory possible with valves. Opposite of Minimalist!
http://www.radio-electronics.com/info/r ... tector.php

A problem is the simple detector charges the capacitor and the resistor discharges it. So adding series resistance and an inductor helps
http://www.ham-radio.com/k6sti/ma.htm

There is a circuit some place for a diode detector that copes better with dynamic carrier power.

 
Posts: 2513
Joined: Sun Jun 10, 2012 7:23 pm
Location: Basildon

Re: Minimalist Superhet Radio.

Post by crackle » Thu Dec 19, 2013 4:58 pm

I experimented with a 100k pot in the cathode circuit of V2 (EF80) and found that 4.5k was the optimum value with regard to volume. So I have changed the 10k to a 4.7k resistor, cathode voltage has dropped to 1.5v, not sure yet if it has improved the distortion. It might be better if I switched a second 10k in parallel with the original one, I think this will make it easier to compare the quality.
Mike

PreviousNext

Return to Radio & Audio



Who is online

Users browsing this forum: No registered users and 3 guests