Supply Voltage for LTP

[xXBrunoXx]

Hi , what is the max rails voltage for a long tailed pair using 2sc1815, 2sa1015. ?
Does the LTP see only one rail ? I know the Vas should be rated for both rails since it swings both rails.

Here rails are +-35v , are 2sc1815 enough for that ? They have Vce 50v and Vcb 60v.

 

[rayma]

That should be fine. The emitters stay within a few volts of ground.

 

[xXBrunoXx]

In LTSpice I have 35v on the LTP transistors, Base-Collector or. Collector-Emitter.

 

[xXBrunoXx]

Full schematic.

 

[rayma]

The 50V rating is ok, but more is better.
Are the rails regulated? That would be safer if using the 50V devices.

 

[xXBrunoXx]

@rayma No. Can you please explain I can't find anywhere. only one rail is seen on the LTP transistors? ( Vce , Vbc ) How do they swing the input signal,
This is with 200mv ac input. it swings 200mV on the 35V rail ?.

 

[xXBrunoXx]

So , for 1V input it will swing 0.5V to + rail and - rail ? but 35v being the middle ?

 

[steveu]

Normally, yes, but don't ignore the possibility of a latch-up condition. This is why you see clamping diodes on the feedback decoupling cap etc. I have actually seen a simulation latch up until I clamped the LTP. The model of the LTP transistors likely does not include the Vbe avalanche voltage (~7V), but you can and should add it to the model. I posted about this a while back in the LTC spice thread. Also, what happen when the preamp creates a startup ~40V transient that adds to the rail voltage. If you don't cover these possibilities, you get an amp that mysteriously fails one day. Clamping also protects against ESD on the input when you connect it hot.

 

[xXBrunoXx]

@steveu I'm not sure where you mention " clamping diodes on the feedback decoupling cap " C6 ? , Also, Please take a look here I have a few questions. https://www.diyaudio.com/community/threads/simple-sub-amp.406002/

 

[rayma]

The Q11 gives most of the voltage gain. Think of the junction of the two emitters (at the CCS) like a virtual ground.

 

[rayma]

I have actually seen a simulation latch up until I clamped the LTP.

That's a pretty good simulation.

 

[steveu]

Yes, C6. Normally very little DC (nor AC) voltage happens on C6, but if the output is held at one of the rails, C6 will follow in seconds, unless it has shunt diodes. These diodes never conduct under normal conditions. You will also see clamping diodes between LTP bases as used by Bob C. A clamp on the input is not so easy because the signal voltage could be more than a diode forward drop.

 

[xXBrunoXx]

What if I put diodes like this ( but dioes are 0.7v ) maybe two diodes in series or LED's, I can " stop " the input being higher then 1v , 2v , 3v dependin on diodes, it wil clip the signal make it square almost, but won't allow more voltage in right?. anyone used this approach?.

 

[xXBrunoXx]

@steveu wait, that is what you said " clamp on the input is not so easy because the signal voltage could be more than a diode forward drop. "

 

[xXBrunoXx]

@steveu Like this >? " You will also see clamping diodes between LTP bases as used by Bob C. " what does this ? compared to the diodes on the input post #13 .

 

[rayma]

Input is typically up to 2V rms for full output, or around 3V peak.

 

[wg_ski]

Common mode voltage is 2-3 volts. The actual differential between the two bases never gets more than a vbe. It only takes a few mV of voltage change to steer the current all the way from one side to the other.