Failure in building an Elenco AM/FM 108CK kit ...

[mossygreen]

A few weeks ago I started building an Elenco AM/FM-108CK kit. I've completed the AM-side assembly but am having problems at the alignment stage. I've asked Elenco for help but haven't heard back (I'll reach out again). I'm trying to do the instrument-based alignment w/ a Tek function generator and a Tek 2465B scope, along with a Brymen 869S DMM.

Basically, there appears to be a lot of AC content at various places and the voltages they give as test positions don't match for the AGC section. I've attached the schematic, the BOM, and the test conditions pages to this post. Here is a link to the whole instruction manual:

https://www.elenco.com/wp-content/uploads/2017/10/AMFM-108CK_low-res-2.pdf

When I measure the voltages as they describe (9V power supply, power on, volume at minimum, short the leg of C29 heading to L4 to ground), things look fine except right around TP5. Instead of being 1.4V, it's about 0.75-0.82V (it varies). That then messes up the Q8 transistor's base and emitter voltages.

If I ground out C29 on the "near" leg (closer to the tuning section), then TP5 is steady at 1.44V, but of course Q7's base is 0V. The collectors of Q7 and Q8 are at 8.8V, reflecting the drop across R38. Q9's collector is at 8.99V. The resistance between the antenna leads is spot on with their diagnostic data - I have the left 4-wire variant on that page.

My test voltages following their procedure are:

Q7 B - 1.53
     E - 1.01
     C - 8.85

TP5 - 0.75

Q8 B - 0.75
     E - 0.18
     C - 8.88

Q9 B - 1.74
     E - 1.10
     C - 8.99

The kit PCB is a little primitive - big pours w/ no through hole plating and no thermal relief for the pads. Soldering it with lead-free solder means that some heat is going into the components. I use an old-style Edsyn iron set to 725 and generally have good results soldering.

I just don't know what failure modes to look for. I'm a little suspicious of the gang tuner assembly. In soldering it in place they want you to short the guide pin with a lead that's next to it. That probably wasn't connected at first (again, no through hole plating) but I can't see how it wouldn't be now that I added a lot of solder at that spot - it's blind, though, so I can't be sure.

Any ideas on how to diagnose this would be greatly appreciated!

[floobydust]

Your voltages look fine. I wouldn't use the word 'failure', these kits are supposed to be fun and teach. Wow 64 page instructions that covers a lot. Common mistakes are mixing up the transformers (red, yellow, white, black) some IF transformers have a capacitor inside them, which may short to solder blobs.

I believe it uses negative AGC meaning the TP5 AGC voltage goes down with increasing signal in order to lower Q8's gain.
With the local oscillator disabled C29 to GND, there should be no AC signal downstream at TP5,4,3 and AGC voltage should go up to the highest possible ~1.4V at TP5 or TP2, and you get 1.44V which is ok.
If the AGC voltage is still low 0.7V and there's AC signal present, then the IF amplifier chain is oscillating and that can get annoying to stop. Maybe that is what you are seeing, especially when C29/L4 is grounded compared to TP7. I would put the scope on D4-K, knowing the probe capacitance detunes things when poking around TP4,3.

Try align the AM IF into the ballpark. I disable the local oscillator and inject a weak 455kHz sine wave into the IF amplifier (mV with cap+resistor from sig gen to TP6), then adjust the tuning IF transformers for max signal. with a scope, or AGC for lowest voltage as the manual suggests.
Then I enable the local oscillator and get it ballpark, using a freq counter or luck if I can pick up a station, between the five (!) tunables - L5, and the tuning capacitor and the trims. That can be difficult.

If this doesn't work, then the radio might have an instability issue to fix or sometimes they just misbehave until aligned. The IF transformer's can needs to be soldered to ground too, at the tabs.

[mossygreen]

Thanks very much - I'll check the components and follow the path you describe. You are right - failure is the wrong word. With no experience, I was stuck with no apparent path forward.

[mossygreen]

Okay, it took me a little time to get back to this to this. Someone else was suspicious of my solder joints so using some rosin a reflowed all of the joints. At that point, I still had the same voltage symptoms as before. Then I moved the board around a little on my work surface - and the AGC voltage went from the 0.7 to 0.8V I was measuring previously to 1.0-ish, slowly wandering up to 1.1V. I repositioned some things and took the measurement again - back to 0.7-ish.

Well, that got me to press on different components on the front and then the rear of the board. With finger pressure on the end of R34, backside of the board, the AGC voltage came up to 1.4V. That made me think that there was something wrong with that part (1M ohm resistor) or joint, so I replaced the part - same problem. I then put the part on the backside and ran a snipped lead from the end of the resistor to the destination of that trace. Same problem. If I measure the resistance, that trace (with or without my jumper) shows 0.0-ish ohms. I've attached a pic (IMG_0869.jpg) of the final installation, resistor underneath w/ a jumper wire duplicating the trace.

I then attempted the alignment procedure as you suggested. I set up my signal generator at 455kHz modulating a 400Hz tone at 80% - and they want you to start off with a very small signal. I'm at 2mV. With the signal generator on and power to the board I see 1.2V or so of an erratic signal (see IMG_0861.jpg). If I press down on R34 quite hard the noise virtually goes away - but to take the picture (IMG_0865.jpg) I couldn't get things quite right. When I press just right, things are quite tight and a nicely periodic signal is evident.
At this point I think I have a board that is bad (could be my fault) and/or a component that is bad (perhaps the transformer that R34 is headed into). Do these symptoms suggest a particular direction to investigate?

[KE5FX]

 Three bits of advice (well, four):

1) Usually, alignment is best tackled after overall functionality has been achieved, not before.  For troubleshooting, work backwards from the audio section towards the antenna, not the other way around.  If you have good audio, and you can demodulate a 455 kHz AM tone signal injected at the output of the last IF amp, you can start to focus on earlier stages.

2) A one-megohm load resistor (R34) across the primary of T6 is kind of unusual.  It makes me think that at some point they had problems with the IF stage breaking into oscillation... but a 1M resistance would make a rather small difference given the much lower RF impedance levels in question.  (Put it this way: 350 femtofarads of capacitance has a reactance of 1M ohms at 455 kHz.  That's a ridiculously small amount of capacitance.  Just moving your hands near the circuit will alter the stray capacitance at various points by that much.)  When the design came out, the transistors might have been slightly different (or, actually, a lot different), perhaps with less gain than the ones they're including with the kit now.  I would try lowering that resistor to 100K or so, to begin with, and see if the erratic behavior gets better.

3) The soldering I can see in your photo looks good.  If you aren't 100% confident in your soldering, life will be easier if you stick with lead-based solder, at least in the beginning.  It is not the least bit hazardous unless you make a habit out of holding it in your teeth. 

4) That's a nice scope, looks brand new. Turn down the intensity before you burn the CRT!   

[0culus]

FWIW, I had a chat with an engineer who works for elenco when I built one of these. Basically, the manual was written a long long time ago, and the components of the time were different. It is pretty much expected that you will find places where reality deviates from what the manual says it is, particularly when it is outlining tests to perform. There's also some typos IIRC.

It's a fun kit to build; some good advice above on how to approach aligning it after you've built it.

PS: I built mine using a 2456B as well; I still have it but I prefer my 7904A mainframe for general use now.

[floobydust]

Elenco uses same circuit for their AM-550K, with that 1MEG resistor which looks like too high in value. All of the local oscillator is coupled into the next (1st IF) stage, which might be way too strong especially with 9V power. I would also try lower R34 (it's ~150k in other radio kits) and add a ceramic bypass cap across C34, around 0.022uF to ensure no RF is on that DC rail as your C34 might not be great at 1MHz.

You can also run it at lower VCC say 6V, or lift a leg on C37 to reduce the last IF stage's gain for an experiment and see what that does. If it does nothing, then the LO and 1st IF are not getting along.
I think the radio's problem can get aggravated by unexpectedly high beta transistors, or defective transformer or just the large circuit layout and grounding, or an AGC voltage that is wrongly designed.

I note the chinese AM radio kit with the same transformers but 1.4V power instead use a series 100R resistor to keep it stable which makes more sense to me (and the 24k collector resistor).

At this point, hard to tell if the Elenco circuit is finnicky and there's something "special" about OP's parts.
Just be patient and try to enjoy what happens to all people experimenting with radio circuits. The scope is a wonderful tool, put it to use 

[mossygreen]

Thanks for all the responses. I tried the altered resistor value, worked through the debugging steps, etc. No dice. The issue appears to be correlated to a board level issue as I can make it go away by pressing on a small area of the board. I tried various things to try to compensate for broken/intermittent traces. I decided that I'd learned a bit already and I would be better off working on the "Learning the Art of Electronics" sequence which I was trying in parallel with the Elenco kit.

BTW - the scope was actually set pretty dim - the apparent brightness of the trace is due to the camera compensating for what it considers and underexposed image.

[ledtester]

Just wanted to mention that there is an ongoing series of youtube videos on building and analyzing this kit by "The Offset Volt" starting at video #80 and onwards:

https://www.youtube.com/channel/UCr5GSao6hWBvsVLGWha1hCw

[mossygreen]

Thanks, I'll have a look.

[CJay]

Appreciate the design is an old one but surely Elenco would be selling substitute parts in it if they didn't work.

Given a press on the board makes it perform differently I think perhaps it might be worth another really close inspection of the soldering as well as the tracks for breaks or shorts.

I think I'd be poring over the schematic and metering out connections, making sure inductors and transformers aren't short to ground or open

[0culus]

Just call them. When I built mine one of the IF cans was bad, so they sent me a new one free of charge. If the whole PCB is delaminating they might even replace the whole thing.

[mossygreen]

I actually reached out to them a couple of times and got zero response. These are strange times so I don't condemn them for it. I actually bought two kits and will try the other at some point, leveraging the videos referred to above.

[niko20]

Speaking of the Elenco fm/am kit what is the purpose of C36 in the second AM IF stage? It literally bypasses the base signal to ground on what is stated to be a common emitter amplifier. It makes no sense to bypass to gerund there, on the base (signal input) unless maybe they are trying to attenuate the signal on purpose? Bypassing the ground there is going to cut into your signal a lot! If you watch Offset Volts videos you can even see the i put signal on Q9 looks horrinle and it extremely tiny with a lot of noise.

They mist be cutting the signal down for some weird reason? What is the point of C36?

[mikerj]

C36 is an AC bypass for one side of the IF transformer secondary and the bias network R39/R40.  The transistor base is connected to the other end of the transformer secondary, so C36 is not shunting the base signal to ground.  If you removed C36 it would reduce the signal on the base and allow supply noise to be coupled in.

[niko20]

Hm


(C36 goes directly to ground here)

Ok  if I re-arrange the schematic I can see the transformer is a bit unconventional - you are saying C36 gives the transformer a ground, because the transformer is basically "between" the base bias resistors and the base of the transistor.

Like this:

That makes sense but why didn't they just ground one side of the secondary and input the other side of the winding to the bias resistor network like usual through a blocking capacitor? Seems like it would be almost similar function-wise and you don't even need more components and would be more conventional.. Perhaps this arrangement helps with input impedance? Maybe it blocks off the affect of the bias resistors better for impedance purposes?

[mikerj]

That makes sense but why didn't they just ground one side of the secondary and input the other side of the winding to the bias resistor network like usual through a blocking capacitor? Seems like it would be almost similar function-wise and you don't even need more components and would be more conventional.. Perhaps this arrangement helps with input impedance? Maybe it blocks off the affect of the bias resistors better for impedance purposes?

Because you'd then be loading the output of the transformer by the bias network impedance.