IF Rebuilding help.

5/9/2009 3:45:14 AMPeter G. Balazsy

I am rebuilding ALL the IF cans on this Zenith 10H20. I've already done the FM discriminator and the 1st FM IF, T1-A.
I am now doing the 1st AM IF, T1B and things seem a little wrong although the radio seems to play fine.
I selected the caps for T1B by measuring them before complete dis-assembly (just 1 coil leg lifted). The mica was rather oxidized but no static or shorting or anything so I think the values I measured are close to being correct.
I looked up and calculated the L/C resonance on-line calculator:

http://www3.telus.net/chemelec/Calculators/LC-Calculator.htm

for the 2ndary and it suggests ( 960uh-127pf) for 455khz...but all I had on hand was a 125pf as measured.... so I figured it good enough.

I couldn't do the same for the primary as it has a screwy-looking double cap arrangement.

Anyway.. after reassembly and alignment attempt... I was able to peak the primary slug reasonably well although it seemed a bit broad as opposed to a nice sharp peak.

The real problem though is that I get NO reaction at all while trying to peak the 2ndary. I screw the slug up and down a lot.. and there is [b]no change[/b] or peak that I can detect.
Any ideas?

5/9/2009 3:50:07 AMPeter G. Balazsy

UPDATE

Well ... duh... I found the problem.

I hate it when things[i] seem[/i] sort-of right but aren't and I know there's no such thing as "magic" here... so this just goes to show that "PPPO" works...lol
PPPO = "persistence and perseverance pays off"

After posting the above I dug back into the circuit and tore off all the leads to that IF can and cut away that 125pf cap on the 2ndary.

I remeasured it to make sure it was close and yes it was 125pf.

So then out of pure curiosity I REMEASURED the coil... and lo and behold if it didn't come up with a DIFFERENT inductance than I measured yesterday when the coil was out of the circuit.

NOW.. it measured 680pf ... DANG... there's the answer!...lol

Thanks to my handy-dandy L/C meter ($120)

Now... I went back to the L/C resonance calculator and it indicated that 680uh requires 180 pf to resonate at 455khz.

GREAT... so I dug out a nice silver mica 180 that actually measures 185pf and popped that on in. Only 5pf high...right?

Eureaka ! ... and voila... it peaks now.

However I had to back the 2ndary's slug way up to the top though to get a nice peak.

.... Seems that using 185pf vs 180pf needs 20uh less inductance than 180pf. So I guess backing out the slug gave that 20uh drop in inductance.
So unless I can find a cap a bit closer to 180pf... so that I can run the slug in closer to mid-range... this is where it's going to have to stay for now.

5/9/2009 12:04:44 PMEdd

My usual probem child is the LAST IF transformer at the detector stage, but on newer generation variants of that
specific Zenith IF transfoirmer design.
The only variant on that T1-B that I see, is that I would have ALSO lifted the tap lead wire along with EITHER one of the
coils end wires, which you mentioned doing, as otherwise, there is some interconnect between the two caps.

BUUUTTTT you do say that the primary SEEMS to tune up OK, with the problem being the secondary section.
Since I do remember you having a decent bandwidth-triggered Silly-Scope along with a signal generator, how
about trying this, in comparative evaluation of the resonant circuit if that T1-B secondary, along with an ALMOST akin, T2-B circuit.

What you will be doing. . .in a no power applied situation. . . . is initially having sig gen and scope grounds going to the
low side of T2-B secondary with the sig gen RF output going to the high side of the secondary winding, also clip onto that
same point with the scopes 10x probe.

Set the Hoz timebase such that you will not be getting a parade of individual signal nodes, such as you would with a setting of 1 us, but instead in the 10us or 100 us range, such they will then be jammed /clustered together and you can just concentrate on their overall “ green” amplitude.
Set the vert sweep sensitivity down into the ~ 1 or 5 Mv range, since you are 10x’ing it.
Then bring up the RF input at ~455 such that you can get a couple of centimeters of display amplitude.
Take the freq adj dial and rather rapidly manually sweep back and forth across the proximity of 455 to get an evaluation of the actual peak that the IF is tuned to as well as the bandwidth of its slope.

(Holy Mackerel ‘dere Andy, you done gots yerself a sweeeeep generator, with you being the sweepOR and the generator being ‘de sweepEE ).

With that evaluative data. . .locked in your mind. . .then move over to the T1-B secondary connect up the grounds to its low side and sig gen / scope to its high side and evaluate its like operation.
Mainly, as to its resonant point that you currently have left it tuned to, as per its being right at that same 455 range, and also how the
two side response slopes compare.

Should you want to be even more exacting, in comparison of the two units, you could roughly estimate the bandwidth by taking a P-P reading of waveform at the 455 peaked freq and then sweep to the lower side until there is a drop in amplitude down to ~ -75% level and then read out the frequency. . .I remember, you got one ‘ a dem too !. . .and then you swing the other side of 455 , acquire the same -75% level down from P-P and read it out. Then, you will at least have some empirical figures in comparing the two transformers bandwidths.

I’m expecting that problem T1-B secondary section to be either waaay down in gain, OR at a totally different resonant frequency peak than you were expecting, or else have a vewy broad response curve.

I am also wondering , in the specific manner of how you handled the problem of the now naked sets of metal ”petals” within the transformer which you will not need if you have eliminated the inner silver clad mica dielectric disc and are using
outboard silver micas ?

73's de Edd

5/10/2009 12:39:04 AMPeter G. Balazsy

hi Ed:
Well you can see from my 2nd post tht I found the problem.. but thanks for the info on how to "pseudo-sweep" it or to measure the fall off points for bandwidth.

As far as the 'naked' petals... well on the FM discriminator I slipped a disc of cardboard in there as an insulator.

But after that on the others I've done now I just cut off most of the petal leaving just enough to keep the lug from slipping through the holes and then the inner phenolic disc that the coil form is also mounted on will go back in place just holding what's left of the cut-petals down firmly.