I am now designing a discrete balanced preamp, but I cannot make a decision if I must add protection circuits.
within the circuit I used an opamp as dc servo to monitor the output DC,
resulting lower than 1mv DC output.
But then I dont know if I must add a DC output protection that having a relay to switch off the output and circuit to monitor.
And also a overload protection to protect the output stages.
(I dont think it will overheat so I am not asking overheat protection here 😛 )
This question may look stupid but in the design PCB space is critical so,
this become a question to me.
So I would like to ask if you guys will play safe and add all those protections or,
you think it is fine that to have no these protections for preamp. DC servo is enough. DC output protection in power amp is enough to protect the system.
I just a bit afraid if there is any failure in the preamp will result destruction of the power amp and speaker. Or am I overthinking?
What will you do for your preamp?
Thanks for any suggestions and advises.
within the circuit I used an opamp as dc servo to monitor the output DC,
resulting lower than 1mv DC output.
But then I dont know if I must add a DC output protection that having a relay to switch off the output and circuit to monitor.
And also a overload protection to protect the output stages.
(I dont think it will overheat so I am not asking overheat protection here 😛 )
This question may look stupid but in the design PCB space is critical so,
this become a question to me.
So I would like to ask if you guys will play safe and add all those protections or,
you think it is fine that to have no these protections for preamp. DC servo is enough. DC output protection in power amp is enough to protect the system.
I just a bit afraid if there is any failure in the preamp will result destruction of the power amp and speaker. Or am I overthinking?
What will you do for your preamp?
Thanks for any suggestions and advises.
Always use a coupling capacitor in series with the output. Nothing more is needed to stop any DC affecting your amplifier.
It is possible, if your power amplifier does not have an input coupling capacitor.
Guitar amp manufacturers often add clamping diodes from output to +/-V rails, for the very good reason that such equipment is portable, so being plugged/unplugged all the time (which means errors are possible) ,often in dark crowded noisy Club stages, and sometimes by hurried/drunk/stoned operators, with a high possibility that a plug to plug cable goes from a, say, 100W speaker out into a 10/100mV sensible input, a preamp or effects loop, etc, NUKING it.
Not sure about your environment but clamping diodes are easy to add and do not affect output sincen they are reverse biased all the time unless a catastrophe happens.
Here they are applied to input:
but you can do same with output.
Not sure about your environment but clamping diodes are easy to add and do not affect output sincen they are reverse biased all the time unless a catastrophe happens.
Here they are applied to input:
but you can do same with output.
Thanks but I am designing a dc coupled preamp so an output capacitor is not my solution.
Guitar amp manufacturers often add clamping diodes from output to +/-V rails, for the very good reason that such equipment is portable, so being plugged/unplugged all the time (which means errors are possible) ,often in dark crowded noisy Club stages, and sometimes by hurried/drunk/stoned operators, with a high possibility that a plug to plug cable goes from a, say, 100W speaker out into a 10/100mV sensible input, a preamp or effects loop, etc, NUKING it.
Not sure about your environment but clamping diodes are easy to add and do not affect output sincen they are reverse biased all the time unless a catastrophe happens.
Here they are applied to input:
but you can do same with output.
If I do this to output, this will only clamp the output within the power rails?
But the amplifier should not be possible to output higher than the rails.
this is a good idea.
even if the output is shorted, the transistors may not be killed immediately.
only burns the resistor.
But how about DC at output?
only DC servo?
A DC coupled preamp with a DC servo is just a complicated and potentially unreliable way to achieve the same result as an AC coupled preamp: an LF rolloff with zero DC gain from input to output. In both cases any LF signals around the LF rolloff point will be subject to whatever effects, if any, are due to capacitors. Don't make the mistake of thinking that caps in the servo are 'not in the signal path'.
Agree.
With due respect: no down-to-DC Musical sources, no down-to-DC speakers so although it´s an interesting theoretical challenge worth pursuing as far as it´s practical, do not get deeply committed to it.
Admit that we do have down-to-dc ears, as anybody diving, flying in an unpressurized plane, rising/falling fast in an elevator or simply going through a runny nose cold can attest (ouch!!) ... but that´s not "music".
As a side note: yes, clamping diodes do nothing under normal conditions but clamp external voltage source to +/- 15V if it tries to go higher.
A series resistor limits current.
With due respect: no down-to-DC Musical sources, no down-to-DC speakers so although it´s an interesting theoretical challenge worth pursuing as far as it´s practical, do not get deeply committed to it.
Admit that we do have down-to-dc ears, as anybody diving, flying in an unpressurized plane, rising/falling fast in an elevator or simply going through a runny nose cold can attest (ouch!!) ... but that´s not "music".
As a side note: yes, clamping diodes do nothing under normal conditions but clamp external voltage source to +/- 15V if it tries to go higher.
A series resistor limits current.
When designing for live sound, where frantic techs do a lot of plugging too quickly, I try to protect all jacks for 100V 10A-- if not for "no problem", at least so only cheap parts smoke. (I have had 48V on 30V mikes, and 100W amp out to an input is far too easy.)
For home hi-fi, I would not worry.
I do not make this mistake of thinking.
I use film caps at the integrator of the servo loop.
Of course, it achieves the same result as a huge coupling cap.
What potentially unreliable ?
Sure it clamps to +/- 15V. It can be good to have them, but it is not bullet proof....Do not think, the rails are rock solid constant voltages.
Regulators do regulate loads that sink but not for abnormal loads that source.
BTW I do not design for drunk or stoned users messing with cables.
Many op-amp have output protection. I do not take a chance, the 47 ohm limits the current to a safe value. That is not the only reason, I want nothing to destabilize the op-amp, for instance a too capacitive cable.
Generaly speaking, it is bad practice to expose inner outputs directly at outside connectors.
DC at the ouput is taken care by the servo loop that forces 0V +/-1mV.
Thanks for reply.
I will take your advice adding an output resistor.
So in your opinion, having both dc servo and output resistors is already making the preamp safe enough?
I can only assume the preamp is followed by a DC coupled power amplifier.
This sort of thing scares me - as it is fully capable in NORMAL OPERATION of destroying any speaker connected to it.
Unless it has a DC protection circuit. Which is kind of oxymoronic, isn't it!
To be honest if I were you, and building this "on principle" I would add a dirty big cap on the output as this is ultimately the best protection.
As an aside, how are you managing the input? If there is DC here your servo will be working against it and may saturate. How is this handled (apart from the fact that a competently designed source will generally be AC coupled)?
Additionally... where is your volume control in this chain? Making sure there is zero DC here is essential to avoiding crackly volume controls....
This sort of thing scares me - as it is fully capable in NORMAL OPERATION of destroying any speaker connected to it.
Unless it has a DC protection circuit. Which is kind of oxymoronic, isn't it!
To be honest if I were you, and building this "on principle" I would add a dirty big cap on the output as this is ultimately the best protection.
As an aside, how are you managing the input? If there is DC here your servo will be working against it and may saturate. How is this handled (apart from the fact that a competently designed source will generally be AC coupled)?
Additionally... where is your volume control in this chain? Making sure there is zero DC here is essential to avoiding crackly volume controls....
Indeed the only fully safe way is a big coupling cap at the speaker.
This cap is a must for tweeters.
When using full DC to avoid caps in the signal path, it is clear that a failure can destroy your speaker. Sure, a component or a solder joint failure can blow up your speakers.
Wether that failure comes from a servo or from any other part of a power amplifier or a preamplifier is the same.
I understand, fiddling with a servo can be unsafe but this is the same when fiddling with any DC coupled power amplifier.
Glitches from volume pots is another issue that asks for serious design when going DC coupled.
Actually, the main point is: Removing electrolytic capacitors from the signal path.
This cap is a must for tweeters.
When using full DC to avoid caps in the signal path, it is clear that a failure can destroy your speaker. Sure, a component or a solder joint failure can blow up your speakers.
Wether that failure comes from a servo or from any other part of a power amplifier or a preamplifier is the same.
I understand, fiddling with a servo can be unsafe but this is the same when fiddling with any DC coupled power amplifier.
Glitches from volume pots is another issue that asks for serious design when going DC coupled.
Actually, the main point is: Removing electrolytic capacitors from the signal path.
I used to design DC coupled amp as capacitor directly placed in the signal path for me give some coloration, which I don't like.
In most cases I will try to avoid such.
For the input, in the design at this stage,
I use an opamp with DC servo to combine the balanced signal to SE with no gain first,
Then come to a volume control pot.
Then Amplification and convert back to balanced (both with DC servo) after volume pot.
Every extra component increases unreliability, especially active components. As a minimum, a capacitor is replaced by an opamp, several resistors and at least one capacitor.mchambin said:
What about the capacitors in the servo? They are 'in the signal path' too.snowcatmaiden said:
- Status
- Not open for further replies.