Toshiba IC70

[briskly]

Originally Posted by Herald1360

Hi,

a 2SB364 (nearly fooled me but a 2SB354 is a 100V 15A stud mounting device!) is TO1 and the spot is collector

Hopefully it's survived.....

Thanks! Sorry about the typo! Just checking - it is not a spot but an arrow on the Toshiba 2SB364 - is this still the collector?

[briskly]

Originally Posted by Hybrid tellies

if all fails building a small IC amp is still an option.

Thanks - do you have any idea what sort of gain I'd have to go for? Some of the IC amps need lots of extra components for more than 20dB.....

[briskly]

Originally Posted by Herald1360

I

Possibly cheaper option to just use silicon pnp and increase the 39R resistor to get bias about right.

Thanks - any idea what value for the resistor I should try for starters?

[julie_m]

If you've swapped over the collector and emitter, the transistor shouldn't have been damaged permanently; but it will have much lower hFE in this configuration. (A boiler ignition board I worked on once worked perfectly happily with a BC547 reversed. We had been shipping them with the transistor in wrong-way-round for about 3 months before anyone noticed it didn't match the silkscreen ).

Be suspicious of anything in a TO92 package! The middle lead is usually the collector, except when it's the base (as in BC547, 557, 337 and 327). If in doubt, test it -- even cheap multimeters have a transistor test function nowadays. Identify the base by the usual process of testing as diodes, then insert the transistor in the test socket both ways around. Whichever way gives the higher gain reading is correct.

I'd be tempted to fit a BC327 and 337 (both C-B-E) and a different bias resistor; try 100R for starters and adjust for minimum current with no signal, then apply a signal and adjust again for best compromise between distortion and current.

[Herald1360]

Second AJS's suggestion.

I would have suggested about 120ohm for a starting point, but go up or down as required to get the 2.5mA standing current in the output pair suggested by the circuit.

If you have one, put a 100ohm variable in series with the 39ohm and increase its resistance from zero until the current is correct.

Of course if the original 2SB264 is OK, you're probably almost there already.

[Herald1360]

Done a bit of poking about with a simulator.....

Looks like for all silicon devices, the 39ohm needs to increase to around 220ohm for output bias also the 100K reduce to 82K to get the midpoint voltage back near 3V.

Did the sim with a BC557 pnp but pretty much any small silicon pnp would do.

If the output standing current is too high, reduce the 220ohm, if it's too low increase it. You may need to use a pot or use two resistors in series or parallel to get things exactly right.

[briskly]

Thanks again for all your help and boost to morale! I'm going to try swapping the 2SB364 leads as soon as I get a chance, then try the other suggestions if necessary.

Andrew

[briskly]

Well, I put the 2SB364 back with the leads the right way around, but still no sound and only 0.2V at the the common emitter point.

In desperation, I made up another amp using an LM386N-1 IC. This seemed to work, after a fashion, for a while. I found that even when set up with minimum gain - no components bewteen pins 1 and 8 - the output was distorted. I then put a 500k resistor between the volume control and the input and the sound was pretty good, but the 10K volume control then wouldn't work, of course. There was also quite a bit of a sort of motorboating sound. I haven't connected anything to pin 7 - could this have an effect? Unfortunately, the waveband switch is faulty - AM hasn't worked for a while, just FM - and I've lost the tuner output. It is not clear from the board how I can by pass the waveband switch completely for FM only - does anyone have any ideas?

Andrew

[frsimen]

Andrew,

I think you still have one or more of your transistors connected the wrong way round. My guess is that either the base/ collector or base/emitter have been swapped on the 2SB364. You can check that the collectors of the other two transistors are wired correctly, as the collector is connected to the case of the transistor. The BC108 collector should be connected to the positive supply, the BC109 collector should go to the diode. You only have the other two leads to check on each then.


The problems you are having with the LM386 are fairly easy to explain. I assume you are working from the manufacturer's data sheet. The data sheet carries the bare bones of the circuit, but you need to add a few more components to allow it to work properly. For the motorboating, you need to decouple the power supply, so add a 220uF 10V capacitor across the supply, connecting it close to the LM386. Adding a 10uF capacitor from pin 7 to pin 4 won't do any harm either.
Returning to the volume control, feed the signal from the radio via a 10uF capacitor, with the negative connected to the volume control. That should overcome the distortion.

Perhaps some switch cleaner will restore the FM and AM operation on the radio.

Paula

[briskly]

Thanks Paula - I'll try all this tomorrow. Thanks for keeping my morale going!



Andrew

[geofy]

If you follow the switches on the diagram it is set in the FM position. One shows 13 which isn't on the separate switch connection pictorial. Maybe should show 10, 11, 12.

If you are using BC109 and BC108 (there is nothing wrong with a B gain type, a BC107 would work just as well) these are metal can types and the tag on the can is the emitter. EBC

There was nothing wrong with the original diode, these are quite tough high temperature diodes. Make sure the replacement is round the right way. The 0.2 could be the 2SB364 is either fully turned on, or has a EC short.

If you said a diode reading of the BC109 shows high all round then this is faulty, check this again.

[briskly]

Thanks Geof, I'll check this.

Meanwhile, I've managed to get the band selector switch working so at least I have a working set using the LM386! I wonder, Paula or anyone else if you could help again? I took your advice, fitting a 220 uF capacitor across the power supply and a 10 uF capacitor to the bypass (pin 7). Using a 10 uF capacitor at the volume control connected to the input of the LM386 removed the distortion and the set worked beautifully in a lash up. However, I had to lengthen all the leads so I could thread them through through the back. I then mounted the 10 uF capacitor on the LM386 board for the lead to the volume control, as there isn't room for this within the set. But I then found that a fair bit of distortion had appeared again. I have managed to remove some of this by fitting a 100k resistor between the 10 uF capacitor and the volume control but it is still present to some degree and gets worse if I increase the LM386 gain. I also get some funny squeaks if a hand goes anywhere near the LM386 board, which is a nuisance as I will have to mount it on the back of the set until I can make a proper repair.

Andrew

[frsimen]

Andrew,

Ah, the works as a lash up, but not when tidied up syndrome! When you said you lengthened the leads, I wonder if you have increased the distance between the LM386 and the 220uF capacitor? The capacitor must be connected close the the LM386, otherwise there is a risk it will oscillate. The other thing to watch is that you keep the output (loudspeaker) wiring away from the input. Sort those issues out and it should work well again.

Paula

[briskly]

Thanks Paula. The 220 uF capacitor is on the board, so unaffected by the increased lead length, I think! I've taken your advice and shortened the leads, keeping the output well away from the input. This has got rid of the squeals. I also seem to have fixed the distortion issue. I had moved the 10uF capacitor away from the volume control (in the lash-up) to the LM386 board. I measured the current drain and found it to be a wacking 70 mA, wheaeas it had been less than 20mA in the lash-up. Moving the capacitor back to the volume control got rid of the distortion and reduced the current to 20 mA, even when playing fairly loud. Maybe spurious inaudible high or low frequencies were driving the amp into distortion? I've just about managed to get the volume control back in the set with the cap and fixed the LM386 board to the back. It is now sounding absolutely magnificent - putting modern sets to shame.

I still hope to fix the original circuit when I get time, using everone's kind advice on the forum. I may have to ask some questions again! But thanks again everyone for your splendid help.

Andrew

[briskly]

I made a lash-up on a breadboard using a BC327 and BC337 for Q1 and Q2 and a 220R variable in line with the 39R and a new BC109C for Q1, as kindly suggested. It didn't work - 5.5V at common emitter point, zero current with no effect on changing the 220R variable. I was about to give up but then tried swapping the BC109 for a BC237. Bingo! 3.1V at the common emitter. 2.8V at the diode, 2.5mA quiescent.It seems to amplify too. I'll now try to wire it up on the original board. It seems using a BC109 was holding things up - this was the 3rd I'd used - anyone have any ideas why? I'm positive the connections were right.

Meanwhile, I'm using an LM386N-4 taped to the back. It seems pretty good at low levels, but distorts a bit a high volume (I would guess around 200mW). It seems to give less distortion than an N-1 I tried though. Are these chips well known for distortion at higher levels?

Thanks agian for all the help,


Andrew

[frsimen]

Andrew,

The LM386 data sheet shows distortion below 1% at outputs up to 200mW, rising rapidly after that to 10% by 300mW. I don't think you are going to achieve concert hall volumes with that chip.

On your other circuit, assuming you had the BC109C wired in correctly, I can't explain why it didn't work but the BC237 did.

Back in post 52 I mentioned my lash up version of the circuit. I added the capacitors and measured its performance. The distortion was much higher than the LM386 - at best I got 4% harmonic distortion, mostly 2nd harmonic and couldn't quite reach 200mW into 8 ohms. Hopefully, your one will work better.

Paula

[briskly]

Paula,

Thanks for this. Interesting that the distortion with this circuit is so high. I'll see what it sounds like with the tuner input as soon as I summon up the courage to open up the set again.....

Andrew

[briskly]

I'd like to thank everyone again for their invaluable help with this. I've just managed to buy a replacement set and wanted to mention that, despite having the mismatched transistors in the output stage (silicon npn with germanium pnp) it sounds pretty good - much better than the old set with the LM386 replacement. Maybe that old design wasn't so bad after all?

Andrew