Fender deluxe 112 Plus. Really hot but sounding fine?

[Alex R]

Useful info here, which I used to solve the problem on this thread https://music-electronics-forum.com/...112-oscillates

[loudthud]

... First, these aren't ordinary 0.7 drop diodes.

These diodes drop from 1.3V to 2.5V (at 175 degrees). This has two implications: first, common 0.7 drop diodes are only ~half so please do not exchange the BYV26 for 1n4xxx one to one.

Secondly, these diodes are extremely sensible to heat.

I often see people on this forum misreading semiconductor data sheets. In this case the Vishay data sheet says in abbreviated form: The Max Voltage Drop at 1 Amp is 2.5V Max @ 25 ^oC and 1.3V Max @ 175 ^oC. Link: https://www.vishay.com/docs/86040/byv26.pdf

If we do a little math, that is -8mV / ^oC at 1 Amp. That is pretty high for a forward Voltage tempco, which is usually in the range of -2 to -2.2mV / ^oC.. Looking at figure 4 (which is mislabeled on the Vishay data sheet), at 10mA, the Voltage drop is just over 1V and the tempco is more like -4.4mV / ^oC. Still high for a silicon diode, but about normal for two regular silicon diodes in series.

Biasing darlingtons is always tricky because the tempco isn't linear.

1V is just about right for biasing a darlington transistor in the output of a class AB amp. The input side transistor is On and the output side transistor is nearing conduction. Four 1N4xxx diodes wouldn't be that far off, you may need a resistor across one or two to bring the bias down.

Edit: added minus signs to tempco's.

[Dave H]

I don't understand. I used to make thermometers using 1N4148 as the sensor. Their tempco was -2mV/deg C. If you look at the BYV26 data sheet you'll see that the voltage drop is lower at 175 deg C than at 25deg C (a negative tempco) that's why they are in thermal contact with the heatsink. As the output stage gets hotter so do the diodes reducing their Vf and therefore the bias voltage.

[loudthud]

Yes, tempco of a silicon diode junction in negative. Voltage drop goes down as temperature increases. I have edited my post above.

[jmaf]

I often see people on this forum misreading semiconductor data sheets. In this case the Vishay data sheet says in abbreviated form: The Max Voltage Drop at 1 Amp is 2.5V Max @ 25 ^oC and 1.3V Max @ 175 ^oC. Link: https://www.vishay.com/docs/86040/byv26.pdf

If we do a little math, that is -8mV / ^oC at 1 Amp. That is pretty high for a forward Voltage tempco, which is usually in the range of -2 to -2.2mV / ^oC.. Looking at figure 4 (which is mislabeled on the Vishay data sheet), at 10mA, the Voltage drop is just over 1V and the tempco is more like -4.4mV / ^oC. Still high for a silicon diode, but about normal for two regular silicon diodes in series.

Biasing darlingtons is always tricky because the tempco isn't linear.

1V is just about right for biasing a darlington transistor in the output of a class AB amp. The input side transistor is On and the output side transistor is nearing conduction. Four 1N4xxx diodes wouldn't be that far off, you may need a resistor across one or two to bring the bias down.

Edit: added minus signs to tempco's.

You're absolutely right and I stand corrected. I drew the wrong conclusions and my post was misleading about the tempco and thermal runaway, so I requested it be deleted.