Elenco AM/FM-108ck Radio kit build. Failed Output Bias Test

[Glassman]

I've just started building this kit and already have an issue with the output bias test.

9v battery at time of testing measured 8.57v

Static Measurements -
- Power Test measured 3.60 milliamps. Manual says current should be less than 10.
- Output Bias Test only measures 0.67volts Between TP1 & TP15.
  Manual says it should read between 3 to 6 volts.
  Note; Turn power off and voltage reads 1.36v and continues to drop.

I have
- Double checked all resister values
- Checked capacitor values and pos / neg orientation
- Re flowed / soldered all joins
- Double checked orientation of IC is relevant to notch shown on PCB

Only thing different I have done is used slightly thicker gauge wire at the speaker / earphone jack.
Two pics attached.
More pics here https://mega.nz/folder/QpknjYyI#dAxlw08nsooE-6IJXjftMw

Happy to add further pics of schematics / manual etc

Appreciate your time and thanks for checking this post out.

[Circlotron]

Guessing that IC is an LM380.
Possibly C40 is bad, allowing leakage current into the IC input.

[Glassman]

Thanks for your reply Circlotron

The IC is marked LM386N, SUM 14L01S.
 
I checked C40 while still attached to board with a basic ESR meter and it read 0.608
I will de solder tomorrow and check again.
C40 is 10uf 25v

[Glassman]

Update:

I have replaced battery with a new 9v that measured 9.94 volts out of the box.
Repeated the power test as described in the manual, powered on and found there was a vibrating sound coming from the speaker and milli amps seemed to climb and thus turned it off.
Repeated the test with the earphone connected and the noise seemed more noticeable (amplified?) and again amps continued to rise.
I have gone back over all connections for shorts etc and checked continuity and all seems as it should.

[floobydust]

U1 LM386 pin 5/TP1 is ideally around 1/2 battery voltage (bias) at idle. But hard to measure if it is oscillating or motorboating, the multimter reading will be wrong with all that AC voltage there.
Did you solder the 0.1uF cap C45 correctly on the bottom side of the PC board, across pins 2 and 6, it's mirror-imaged the pin count. Same for C39 470uF. These caps are essential to keep the LM386 from going unstable.
What does the volume control do?

edit: I looked at your assembly pics and it looks fine, I think your soldering is a bit cool meaning leave the tip touching the PCB an extra 2 seconds to transfer heat to the PCB. The solder is not flowing and making a good fillet, which can lead to bad connections.

[Glassman]

Update 2:

After re flowing some of the solder joins I checked for continuity.

I found continuity between the headphone jack's positive and ground as well as between headphone jack positive & negative.

I separated the internal contact slightly within the jack as they appeared to be touching and re did the power test (figure 6). This time it seemed to pass and measured 3.92ma's.

Continued on to Output Bias test (figure 7) but only got a reading of 1.57v instead of 3 - 6v.

Plugged in the earphone and again got clicking (motorboating???)

[floobydust]

Because at this point in assembly, the volume control has nothing feeding it, the audio amp will pick up noise and maybe oscillate if the volume is turned up. That's why I asked if the volume control does anything, it should be stable when turned low. It used to be called "motorboating" when an audio amp oscillates and makes low frequency clicks and thumping.

[Glassman]

Volume control makes no difference to the sound or volume level and is constant.

The following link is a recording of the sound, from turning on and gradually increasing the volume to volume down and power off.

https://mega.nz/file/059TCbLZ#A6buQWHdgJT6feMy2qbJFwDCsMy6WL1xyxYcWUyeMnI

Is it possible that the cap bridging pins 2 & 6 is the wrong value? Manual says this cap, "0.1uf is to prevent the IC from oscillating. Markings on cap says 104.

[floobydust]

Check the volume control wiper is not open-circuit to C40, and it should be close to zero ohms at min. volume. At zero volume the IC should behave.
I would say you've got a bad component and C45 0.1uF under the board, the part might be too low in value. I usually go larger like 1uF-2.2uF ceramic to shut up the IC. Also C44 0.047uF and C39 470uF are important for LM386 stability so I would test those. The IC does like to oscillate at ~12MHz which can make troubleshooting difficult, and cause the bias to read way off.

[Glassman]

https://mega.nz/folder/QpknjYyI#dAxlw08nsooE-6IJXjftMw/file/R8UUHB4D

With volume control off and meter set @ 200k:

55.0k between control ground pin and + of C40. Power on and increase volume, reading increases to approx 65.0k and motorboating.

Power off, connected to wiper pin and + of C40 and get approx 55.5k. Power on and immediately increases to 65.0 – 66.0k and motorboating.  Sound and reistance increases as I turn up volume and peaks around 117.0k. 

Power off, connected control ground to + C41 and zero reading.
Powered on and zero reading but motorboating.

Checked R43 and measured 98.4k.

Re C45 0.1uF under the board, only other caps I’ve got are some 101, 102 & 103 ceramics or 104 polyesters. Lowest value old lytic I have spare is a 47uf.

Will de solder and test all caps tomorrow.

[Glassman]

Today I removed C44, C45 & C39 from the board and tested with 2 different ESR meters and they both seemed fine.

Also, I read the data sheet for the lm386 where it mentioned a bypass capacitor from pin 7 to ground.
I experimented with some old caps I have lying around. 47uf 10v, 100uf 10v, 220uf 10v, 100+220 in parallel, 1000uf 25v.

Each time the capacitor made a difference with a delay in the time it took for the motorboating to start but it inevitably happened.

The pad at pin 7 came off and pads for C39 are starting to lift.

The manual clearly states that "The amplifier in our kit with a gain of 150 is shown in Figure 5." I've noticed a couple of things:
1. Figures 4B with gain = 200 & Figure 4C with gain = 150 - Both show a bypass capacitor at Pin 7, value = ? yet both Figure 5 and the PCB board show nothing.
2. Coming off Pin 5 to ground, Figures 4A, 4B & FIGURE 5 show capacitor - resistor then ground yet Figure 4C with gain = 150 it shows resistor - capacitor then ground.
3. Resistor values are the same.
Also, re this data sheet https://www.ti.com/lit/ds/symlink/lm386.pdf diagrams showing a speaker connected have a 250uF capacitor before it where as C43 in this kit has a 470uF.

Thoughts anyone

[rdl]

Might try disconnecting pins 1 and 8 (gain), and/or try a lower value cap from pin 7 (bypass) to ground.

[Glassman]

I'm starting to think the 386 is dodgey???

[floobydust]

The volume control resistance measurements seem off. That looks like the volume control is defective (open wiper)? I could find no data on Elenco 622050 potentiometer, it might be 10kΩ or 50kΩ or 100kΩ etc. but 10kΩ is shown in Fig. 4's.
You only ever measure resistance with an ohmmeter when power is off. This is because the multimeter is injecting current which will clash with anything that already has power and current flowing. Always have power off, and wait a bit for caps to discharge when measuring ohms or diode-test.

On ohms function with power off (no battery connected), between the wiper (middle pin) and ground side, the resistance should go down to near zero at min. volume and climb up when you turn up the volume. Leakage current from C40 will interfere at the higher ohm readings but removing the IC (it's in a socket right) open circuits C40 at pin 3.
From end-end on the potentiometer you should see it's total resistance i.e. 10kΩ (it's not exact) but again there are capacitors like C41 (although only one leg is connected at this point) that can interfere with the measurement. Not so if one leg is connected to nothing which I think is the case right now.
Potentiometers can have a cracked wafer or open wiper, which would cause the LM386 to pick up anything including itself and oscillate.
Look at the IC markings, there are many LM386 copies coming out of china and they might not perform so great.

I would contact Elenco and see what they advise. Because you're just starting out on the kit assembly and your assembly is OK it's some defect on their end I believe.

[Glassman]

Might try disconnecting pins 1 and 8 (gain), and/or try a lower value cap from pin 7 (bypass) to ground.

I have lost a couple of pads now at the IC, pins 6 & 7 due, I guess, to an inferior board and soldering / de soldering + pads not the best at C39. Now soldering direct to pins 6 & 7.

Due to the above I was reluctant to remove resistor & cap between 1 & 8 so just held / connected a jumper wire between pin 1 side of resistor & pin 8 side of cap, don't know the proper tech term, bridged / jumped / bypassed ?? resistor & cap. Motor boating stopped immediately.
Tried bypassing resistor & cap individually and motor boating stopped only when I bypassed the cap!

Note: Prior to doing this I had experimented and connected a 100uF cap between pin 7 and ground which made little difference, just seemed to delay the motor boating starting up.
Also, due to the differences between Figure 4C & Figures 4B / Figures 5, I swapped positions of C44 & R45. This change made no difference and again, due to solder / de solder am reluctant to revert back unless I need to. https://mega.nz/file/E5l2wABZ#U4nChtCvYLrq773yUb-CkkkfmaYC4cVjVVTyjRvOyoU

As per manual power test, I now get 4.37ma, which is similar to what I've seen someone get in YTube vid of 4.45ma.
https://mega.nz/file/xocFnT7Q#ENeJeJXx4G-ZcUDVosEV3S1ZfJrbVdYatu7xLYKUg5c
Bias Test between TP 15 ground and TP 1 is .575volts. Should be 3 - 6 volts

[Glassman]

“The volume control resistance measurements seem off. That looks like the volume control is defective (open wiper)? I could find no data on Elenco 622050 potentiometer, it might be 10kΩ or 50kΩ or 100kΩ etc. but 10kΩ is shown in Fig. 4's.”

:  Page 9 of the manual shows,  Volume/S2 (50k Pot / SW). Without de soldering I’m sure this was marked on the pot before soldering in place.

“You only ever measure resistance with an ohmmeter when power is off. This is because the multimeter is injecting current which will clash with anything that already has power and current flowing. Always have power off, and wait a bit for caps to discharge when measuring ohms or diode-test."

:  Thankyou, shows how much of a newbie I am.

"On ohms function with power off (no battery connected), between the wiper (middle pin) and ground side, the resistance should go down to near zero at min. volume and climb up when you turn up the volume. Leakage current from C40 will interfere at the higher ohm readings but removing the IC (it's in a socket right) open circuits C40 at pin 3."

:  With meter set at 200k & no battery connected & IC removed, zero reading, powered on and eventually peaked at 52k @ full volume.

“Look at the IC markings, there are many LM386 copies coming out of china and they might not perform so great.”

:  Markings on top, “LM386N” “14L01S” bottom is K11 in a round circle. I have been reading some reviews about these chips and yes, understand that these can be hit and miss in quality / performance. I’m considering buying some more for potential future projects and spares. National Semiconducter seem to be a good brand / manufacturer?

[james_s]

You could just try building the rest of the thing and see if it works properly, it's possible that there's nothing wrong with it. You could also buy a LM386 from Digikey or some other reputable supplier and try that, it's possible it was defective to begin with or it could have been damaged by static or something.

[Glassman]

I will likely be buying some extra 386's and am now thinking that I may copy the circuit on a breadboard and test. Will have a look at digikey & mouser etc from what I've read so far I'll try and avoid Aliexpress / Ebay etc

[james_s]

I will likely be buying some extra 386's and am now thinking that I may copy the circuit on a breadboard and test. Will have a look at digikey & mouser etc from what I've read so far I'll try and avoid Aliexpress / Ebay etc

A LM368 on a piece of perfboard driving a little speaker can be handy as a bench amplifier, so you could make a dual purpose test amp/LM386 tester.

[Hank30721]

I know this is an old post - just curious how it turned out.  I'm finishing up the Elenco 108CK kit myself.