Wobulator
On Thursday, 31 October 2013 12:03:54 UTC+11, John L Stewart wrote:
Aw Yes, the wondrous Wobulator! Patrick mentioned a very useful piece of test equipment. I built one around 1956 to be used to align IF Transformers in AM receivers. To avoid the mess of mechanically wobulating, the circuit used a reactance modulator. Tuneable over the range of 450 to 475 KHz with a regular variable cap to go to the IF center frequency, the reactance circuit driven by a handy sine or triangular wave oscillator did the rest. The sine wave also drove the X-axis of a scope rather than the internal timebase. A plug-in slot, such as an octal socket provided a place where a crystal of say 465 KHz could be plugged in. That way a birdie showed up on the display to indicate mid-band. The schema is long gone. Think it was published in Audio Magazine. It used a 6BA7. But I expect a search on the web would find something. Cheers to all, John +-------------------------------------------------------------------+ |Filename: KLIBAN KAT A.jpg | |Download: http://www.audiobanter.com/attachment.php?attachmentid=359| +-------------------------------------------------------------------+ -- John L Stewart
About 17 years ago I built a two band RF oscillator with 300-500kHz and 530-1,650kHz. It was limited in usefulness so I added grid modulation for AM in 6BX6RF amp. Then I added a detector and applied NFB and reduced envelope THD and increased max mod to 95% without more THD than could be expected in most radios I serviced.
But for a wobbulator, I created a saw tooth gene which altered the biasing of 8 paralleled 68V x 5W zener diodes arranged to make a vari cap which could not conduct with the 20+ Vac generated in the Hartly LC oscillator tank.
This allowed me to FM modulate the low band up to +/- 10% at centre 455kHz, and then I could see the effects of alignment as obviously as dogs balls.
It was interesting to build, and sometimes useful, but the other way to align IF of AM radios is to shunt the AGC voltage, set a low level of RF below which overloads anything, and then use 455kHz applied to the antenna with set tuned low as possible, and then adjust for max AGC voltage generated starting with IFT2 grid, then plate, then IFT1 gid then plate, then repeating 3 times to make sure you have the best. Then apply square wave AM at 50% and the square wave audio detected should show little over shoot of detector output of set.
Having done that, then use an non AM RF input to antenna at 900kHz, or away from any station frequency ( to avoid any beating or chatter ) and then use F counter to ensure IT is 455kHz, then re-tune the 4 IF coils for max Vagc.
But then the tracking must be adjusted, so then you use 550kHz RF input, and tune set to that for max agc, and adjust the RF input coil ferrite slug to change coil inductance for max Vagc.
Sets with ferrite rod antennas need to have their coil located about 1/3 the way along the rod so that inductance adjustment is done by sliding the coil slightly one way or other on the rod. Rods are directinal, and so adjustments must be done with positions kept constant.
Then set F gene to 1,600kHz, and tune set to the 1,600kHz and adjust the trim cap on the tuning gang for RF input coil so max Vagc is attained.
Then move F back to 900kHz, and Vagc should be high, between the other two values. If you have a set that was designed well and which didn't suffer during the manufacturing process, then Vagc might be say -5Vdc +/- 1V for the whole AM band. But you will be very lucky to get this, and usually you must fiddle around a lot to get the tracking correct enough for say between 650kHz and 1,200kHz, outside of which the gain of the set seems to fade because tracking cannot be got right as RDH4 says it ought to be.
Once everything is tuned optimally for most gain, maybe the darn set oscillates at part of the band causing whistles while you tune, so stagger tuning the IFTs may help, and it ma be what was done originally. Maybe put 4k7 in series with IFT1 grid coil and IF amp tube, anyway, you can sure waste much time getting AM radios to work properly.
aodbout all Agc made
As I said, tracking is rarely perfect in many AM sets. You can try to calculate L&C values you should have to get good tracking. But accurate calculations don't correspond to perfect tracking, where the tuning gang has two equal C gangs, and a series C is used for the higher F of oscillator coil. When you replace the RF coils used in AM radios with ferrite rods ( to get rid of hum modulated AM waves caused by flourescent lamps et all ), then the turns of the coil on rod must be very carefully adjusted to suit the existing vari cap gang and get good tracking.
I've never bothered to make an audio F wobbulator which would be useful for speaker testing if the change of centre F is sufficient to prevent waves reflected off room walls etc to interfere with response measurement. But one could be made using a crystal oscillator at say 100kHz, and then have a VFO which has its output with FM, and the two signals are mixed to give a difference F which is detected with diode detector, and the FM is done using RF FM techniques. Some gear was sold with this idea.
But I just use pink noise for testing speakers and a switched F BPF with Q = 12 to plot speaker response with 33 F between 20Hz and 20kHz.
So, I avoided the need of an audio F wobbulator. Nothing I ever read on the subject led me to believe I was wrong.
Patrick Turner.
|