Nordmende Globetrotter 808 output stage problem.

[MotorBikeLes]

I have had this radio from (almost) new, certainly 40 years. I bought it directly from the NordMende UK importers. It had been sold, and was returned as faulty, and that was how it came to me.
One or both of the output transistors was/were faulty. I replaced them with a beefier pair: BD241/BD242 to replace the OEM BD239/BD330.
I have always used the optional dry lead accumulator (8v, comprised of a four cell battery). However, over the last 12 months or so, if I have plugged it into the mains to recharge the battery, the radio has frequently "locked up". There is a click and the audio disappears, though the tuning indicator tells me it is still tuned in OK. Switching off the mains, then the radio OFF and ON again normally restores things to normal, or even pressing various band change press buttons. It then behaves until next recharge, which I tended to do overnight to get around it as a problem. I assumed it was the battery needing to be replaced.
As reported elsewhere, I have just converted it to take lithium ion cells. Nominal 3.7v xs 2 = 7.4 v. However, they will reach around 9.2v under control by a dedicated charge controller. Whilst fitting the lithium ion cells, I decided to replace the various electrolytic capacitors in the PSU and output section.
The fault is still present, so NOT due to failing Accu. or electrolytic capacitors!
Before I started the battery conversion exercise, I connected the Accu connectors (battery removed) and ran it up on my Farnell power supply. I increased the voltage from around 5v ( = drift of tuning + distortion) steadily up to 15v. No misbehaving.
I had on my list to check quiescent current (10mA at 9v battery output (4 off "U2" types) which showed about 3.5mA, so I tweaked the control. It shot up to 37mA, and no adjustment of the pot would change things. I had ensured it was on FM, no station, min volume, but it occurred to me it may have once more "locked out". Turning up volume and altering FM tuning confirmed this. Would anybody suggest a first line of attack?
Les.

[Leon Crampin]

Your problem is almost certainly due to poor circuit design around the bias adjustment potentiometer. When the wiper of an old pot makes intermittent contact with the track, the circuit shown will turn off the Vbe multiplier transistor and drive the output stage hard on. This results in a high current draw and no audio.

Replace the pot and preferably reconfigure a replacement to be between the base and emitter of the Vbe multiplier. If wired as a rheostat, connect the wiper to one end of the track. Cermet pots are best. A reconfigured circuit will need the resistance values changing.

Leon.

[MotorBikeLes]

Leo, thanks for reply. I think there may be another fault, but I will certainly replace the pot with suitable resistors.
However, I don't fully understand how you advise to reconfigure the circuit.
Les.

[Cruisin Marine]

However, I don't fully understand how you advise to reconfigure the circuit.
Les.

Hi Les,
I guess he means limiting the variable range of the bias pre set by placing it in the middle/ either side of some resistors that limit the range total of the bias adjustment, just in case it goes "crazy"
It means a smaller value pre-set, with much less adjustment range of course.

[turretslug]

My impression was that Leon was saying that connecting the base of the Vbe multiplier to the wiper alone was distinctly unwise- a similar vulnerability arises quite frequently in PSU designs where loss of contact causes the comparator to drive the output full high. Instead of which, the variable component should be rheostat-connected (with one end connected to the wiper for safety) in the lower (base-emitter) arm of a re-configured divider with a fixed resistor in the upper arm.

[Leon Crampin]

Correct.

Leon.

[Cruisin Marine]

The end of the track of a pre-set can often go open cct just as much as the wiper.

[MotorBikeLes]

Right, I understand. That is my intention, but I thought Leon meant ALTER the effective circuit. I will start with a physically large preset, find correct setting, then replace with discreet "R"s.
It was working OK earlier, I will set it up and fit the separate resistors as soon as I get a few moments.
Thanks all.
Les.

[Leon Crampin]

I would alter the circuit, very slightly, to ensure that the Vbe multiplier turns hard on when the wiper lifts off the track, or its slipring tarnishes. In some cases, a highish value fixed resistor connected base to collector of the multiplier will achieve this without adversely affecting the adjustment range. Re-calculating the circuit and wiring the pot as a rheostat in the b-e leg of the network is not difficult and may give a better range of adjustment.

Leon.

[MotorBikeLes]

I temporarily wired in a good 470 pot and quickly set the quiescent current to 10mA with 9v input, and in fact it was just under 5mA when run from mains only. I checked the two resistances of the pot, getting 202 and 249 ohms respectively. I decided to try a 220 and a 270 to replace the pot with as a first try, but the QC was 7.3mA, so not suitable. Thinking further, I dug out a 7.5K pot of correct size to fit, set it mid point, and put the 220 and 270 resistors back in position Set exactly at 10mA with a minimum of 5mA and max of 20mA. I will leave it at that, as there is no chance of O/C pot causing havoc now.
I had fastened it up before seeing Leon's last reply, so will not proceed further with his suggestion. That way, at least, I won't make things worse.
Thanks for replies, Les.

[paulsherwin]

Thread reopened by request.

[MotorBikeLes]

About three weeks after my last post, I connected the mains to recharge the Lithium battery pack I previously created. Next time I switched on, no sound, just like it did before. However, this time it was permanent. A definite failure.
Too busy to touch it, I dragged a Yacht boy into position and put the Globetrotter on the bench. A quick look a week or so later got me nowhere, so today I spent an hour measuring voltages. As previously, the set tunes in according to the indicator, and the 'scope confirms audio from volume control. These voltages make no sense to me, unsurprisingly. I could make a start changing every semicon in sight, but maybe somebody here can make a reasoned suggestion of which (component if not semicon) to look at first.
Les.

[buggies]

What looks wrong to me is T507 has 7.2 volts on the base and nothing on the emitter so would be checking the transistor with an avo (or diode test of DVM).

[MotorBikeLes]

Buggies, thanks, I hope to check it tomorrow. I tried previously to check the various voltages, but struggled to visualise solder side to component side. As the years pass, some skills lessen. This morning I first scanned the solder side from the manual, selected a smallish section, expanded it to A4 and printed it to help the task. I did not study the results as well as I should have, but now agree something looks wrong there. Furthermore, I recall a tiny bit of print damage, suggesting I have either had it out, or changed it at some time in the last 40 years.
Les.

[MotorBikeLes]

Buggies, you win the prize. T507 it was, BC338, I only had a 337. Switched on, perfect.
It was NOT T507 which had been previously removed, but T508.
When tested with an analogue meter, T507 checked out OK.
Cheers, Les.

[buggies]

Good news! I have great respect for the members that can understand these direct coupled amps. If it was not T507 I would have been out of my depth.