Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.tubes
Patrick Turner Patrick Turner is offline
external usenet poster
 
Posts: 3,964
Default Tuning indicator without magic eye or meter

Probably because I am slightly pefectionist, if something goes wrong with any bit of my own gear, I'm usually keen to fix it and improve on it if possible, and so the last 2 weeks has been spent on re-engineering a kitchen radio which began to just cut out after awhile although if I switched the audio amp to the FM tuner, it worked; but I need my AM stations.

Before dicussing the new tuning indicator, I did other things.......

OK, out to the shed with the radio. I soon find a mouse quite dead inside under the chassis - in the exact position where it must have first made contact with +400Vdc. It had been in there for 3 months from late autumn, and had found a way inside the house where I'd had an unpatched hole around a wire to outside through a wall for an experimental antenna. Quite a few mice around, and a kind of plague in good farming areas where maybe 200,000 mice can quickly breed up to eat a paddock clean. Anyway, after removal of corpse, and a Requium Mass at the local church, I got the radio working OK as I'd made it in 1999. Hardly anything wrong. All tubes fine despite 13 years of 4 hours use a day with maybe turning on/off 4 times a day. The second hand volume control had gone a bit noisy, and time to re-appraise the whole approach. I re-arranged for a much better layout and added a variable-mu 6N8 pentode as an RF input stage with a RF choke anode intalled a ferrite rod antenna coil. 6AN7 is used as the F-converter to give IF = 445.0kHz, because that seemed to work best with IF trannies from the junk box. Instead of switchable tertiary on IFT1 to broaden IF pass band, I just adjusted coils closer to each other to get a broader passband but not enough to make two big peaks.
There's a lot of talk about effectiveness of tertiaries, and about badness of twin peaked IF response, but I found a tertiary just wouldn't work right and coils being slightly too close gave 2 peaks, but not far enough apart, so band width remained stubbornly at 10kHz, allowing AF bandwidth of 5kHz.
So, I adjusted coil distance for as flat a band top as possible, tricky, because you have to adjust within about +/- 1mm in a distance of around 18mm. With that done on BOTH IFTs, I then tuned all to peak at 445kHz. Most ppl use 455kHz, but not everyone because for a given batch of manufactured IFTs, some designed for 455kHz just won't have their ferrite cores in a practical position, too far in, ot too far out, and so makers aim for 455kHz, but one might find +/- 20kHz either side of 455kHz.
With all peaked up and coils adjusted for best distance, I could get about 4.5kHz for -3dB AF response at detector. I then tuned IFT1 anode coil 3kHz high and grid coil 3 kHz low, and BW went out to 6kHz, without any peaking. I have 100k resistances across each IF coil to slightly lower Q. One might think that with 5 tuned circuits, "skirt selectivity" would be poor, but this radio is for locals as I don't ever want to listen DX, which really needs much higher Q and AF BW a typical 2kHz which is found is so many awful sounding AM tube radios. Its even worse in SS sets where there are just 3 tuned circuits with high Q and BW = 1.5kHz, and skirt selectivity bloody awful.
With the RF stage added which gives an average RF gain of 11 times, or about +20dB, the whole set came more alive and just worked better in every way.
There was 4 times the audio voltage available and I had to adjust the preamp gain down to about 1.5x between AF detector and power amp. Another tube was added. Anyway, now the FM level when used from a ****ant tuner with 200mV rated output sounds the same level as AM levels.
The preamp now has a 4 pole 4 pos double wafer switch from junk box ( good stuff from 1955 ) and this adjusts the shunt FB network and selects the source, AM or FM, ( or two more sources if I wanted them ). The FB used for AM allows a shelving network to boost treble above 5kHz. The following passive tone control also has some added shelving and AM BW is now out to 6kHz with 10kHz at -9dB when tone is set flat, and with full possible +5dB treble boost the AF response is -3dB at 9kHz. If I'd had ordinary typical IFTs, the rate of attenuation above 3kHz is far too steep to be able to use treble boost with RC networks, even if they had 12dB slopes.
I added shunt regulation to most critical B+ rails using a few zener diodes.. Most critical as to keep the screen voltage to RF, mixer, IF tubes constant, because that makes the applied AVC most effective. Most AVC is applied to RF, with about 1/3 applied to mixer, and none to IF amp which is a sharp cut off 6BX6 pentode with a constant Ia at 5mA for best linearity.
Average AVC voltage generated from IF anode circuit is around -10Vdc for locals.
I added a -320Vdc rail to allow cathode follower detector driving 1N914 diode and RC and its following cathode follower buffer to have 56k cathode loads, so the maximum possible AF can be up to about 30Vrms. No need though. But there are none of the bothers using all high Z circuitry with vacuum diodes.

BUT, how to make sure the set is tuned right on AM? The flat topped IF response
with slightly atagger tuned IFT1 manages to show that highest audio levels at lowest distortion and best sounding bandwidth occurs at a point away from where AVC measures its highest level. Most tuning meters just run the meter off a cathode resistor in RF or IF tubes, and that's OK where the IF BW is narrow for most sets, especially for SW, although tuning by ear is oftem all one can do because at SW there often isn't any change to AVC which is at the delayed AVC voltage from PSU of say -1.5Vdc. ( necessary to stop RF and IF tubes generating excessive gain, grid current, and instability. )
Tuning eyes and meters on many domestic sets are about useless for most SW.
But for my radio, I found the not uncommon problem of AVC peaking not being where best AF was to be had.
So, back to junk box to look at available old RF coils and IFTs to make an additional LC tuning circuit with as high a Q as possible. Many old IFTs I found I had didn't have a very high Q, and nothing like what RDH4 says should be available, showing yet again how so many makers, designers, entrepreneurs routinely laughed at RDH4, and did things to ensure a profit at lowest cost of production. After checking 5 coils, I did find a 200uH input coil with ferrite slug core, only 4ohms Rw, and Q about 60 when added C of 550pF gave Fo at 445.0kHz. I set up this LC fed from 47k from cathode of CF buffer driving diode detector. The live end of the LC has a 1N914 to 270pF to generate a positive voltage of about 1.3Vdc max and this drives a darlington pair of bjts which turns on an LED between its collector and a +17V supply derived off a rectified heater winding. A 1k8 current limiting R is between +17V and LED and bjts. It can't be destroyed easily.
This circuit works just fine, and after setting the LC to be tuned at 445kHz, it was found that by watching the LED turn on and off as stations were tuned, I could always get the set's generated IFT tuning F within +/- 1kHz of 445kHz, usually only 500Hz away, and good enough, and to do better, an F counter and far more complex circuit would be needed.
The beauty of this approach is that the IF FREQUENCY is the guide for whether or not the set is tuned. I don't need to know what the RF level is, or what the AVC voltage is; my ears tell me most I need to know, but exact tuning can always be wrong without something like I have now begun to use.

I have avoided the use of a tuning meter or a tuning eye such as EM84, even though I have a few. Both are much more difficult to fit in than a single 5mm LED. Besides, EM84 and most other tuning eyes need a lot of AVC voltage to work properly, and they don't necessarily indicate best tune, as I explain here. It seems to me the tuning eyes and fancy meters are gimicks on AM domestic radios which most ppl have learnt to tune by ear.

Deriving AVC is best if made from the anode circuit of the IF amp because more Vdc is easily available compared to the more feeble amount usually possible from the secondary of the last IF amp. And furthermore, I've found that using AVC from the last IFT sec tends to cause sibilance while tuning as the radio tries by fails to keep its level up after tuning off a station. Better to get AVC from an anode circuit using C = 33pF only, and diode to 0V, and 1M and 0.05uF following RC network. The loading effect on high impedance anode circuit is negligible.

I've also remade the front of the box and modified the whole tuning mechanism amd dial construction, using parts from a an old Kriesler. Now Kreisler must have had an army of bean counters, but some things they did were just fine, like the dual gang tuning cap with dissimilar gang sections to ENSURE good tracking was possible between tuned RF input and oscillator. Many sets just have two or three or even four identical cap gangs, and they'd use a series "padder" capacitor for the oscillator coil but then you will always find that good tracking is quite elusive, getting perfect tracking is impossible, even though the bottom and top of the band can be coaxed into good tracking. At the middle of the band, tracking can be way out. If you do all the calculations for LC resonances wanted across all the AM band, you will find out why tracking is always not so hot. In my set, I really would not have needed an RF input stage with original Kriesler tunung cap and the oscilator coil. Kriesler also used a Ferrite rod but had a terminal for log wire antenna to be used. To this they added a 5k6 series R with a choke wound onto the R to make a blocker for noise. But in 1960 they didn't foresee an age where fluroescent lamps would be compulsory, and each house would have myriad gadgets which generate appalling levels of noise and RFI which often appears as a 100Hz hum but only when tuned to a station. So the RF electrostatic portion of radiated wave at home is what gets picked up by a long wire, and it has statin signals with added hum modulation. If you have a ferrite rod which acts to pick up the magnetic portion of wave there is usually less than 1/10 of the added modulation - BUT - the wanted signal level to give tubes to work on is also much lower than a long wire antenna. Hence I'vs added the RF input stage. I've also mounted the ferrite rod on a flexible fix above the set but still inside the timber box height to allow 90 degrees of rod rotation. Reception is poor with rod mounted vertically. The rod remains horizontal in my set, but rotates around a vertical axis. I figure I can reach in behind the set and adjust the rod for best noiseless reception for most wanted stations, and no more needs to be done unless I move the set. Its been in the same place for 13 years, and It can stay there. The Pioneer FM tuner from 1976 sits on top. Hence I have good FM and AM in kitchen.

Recently DAB broadcasting has started where existing ABC FM and AM station content is available from 4 channels on DAB radios. But the sets I've seen are pretty awful with tiny speakers giving lousy bass. Well, in theory, the signal from the "radio" part of the digital radio set should be pretty good, so all that's needed is a good following amp or pair of amps if stereo is insisted upon. I figure I don't need stereo in the kitchen if I use my rather good sounding single DIY floor standing full range speaker which acts to support the radio set, which sounds well only 3 feet away.

A glass of goos shiraz seems to help the radio come alive. Especially good is to hear the Brandenburg Orchestra play anything they ever have over the years from their many recordings - fantastic music.

I will be posting the final schematic at my web site for other to follow if they wish.

Meanwhile, even though I have retired, there are more projects here in front of me than ever before, but not too many for paying customers, because not many have ever wanted to pay any better than the old age pension! And finally, I don't have to repair solid state crap. Give me tubes and I'm happy.

Patrick Turner.
Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Possible to hack tuning on Sony car stereo from Japanese to US tuning? Service manual? [email protected] Car Audio 0 April 25th 07 07:03 AM
How to set the tuning on Sony car stereo from Japanese to US tuning...service manual? [email protected] Tech 0 April 25th 07 07:02 AM
I need to set the tuning on my Sony car stereo from Japanese to US tuning; service manual? [email protected] Audio Opinions 0 April 25th 07 07:01 AM
Tuning indicator 6G-E12 Engineer Vacuum Tubes 2 February 28th 05 02:06 PM
FA:Telefunken EMM801 Tuning Indicator NOS/NIB for Dyna Brian McAllister Vacuum Tubes 0 November 17th 04 05:23 PM


All times are GMT +1. The time now is 06:46 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"