In the application note AN-1192 there are diodes at the input of the DC servo. What is the purpose of these?
I can think of two possible reasons.
1 Limiting recovery time in case of saturated servo (will almost never happen)
2 Protecting the opamp against too high differential voltage (almost any opamp will take +- 15 V at least). When will this happen? Never?
Conclusion: These diodes are hardly necessary. Am I wrong?
I can think of two possible reasons.
1 Limiting recovery time in case of saturated servo (will almost never happen)
2 Protecting the opamp against too high differential voltage (almost any opamp will take +- 15 V at least). When will this happen? Never?
Conclusion: These diodes are hardly necessary. Am I wrong?
And the answer is ......................2
If you look at the way that an OpAmp with negitive feed back works; the amp attemps the adjust it's output until the two inputs are equal. The only time there will be a difference between the two inputs is when the amp has "run out of gain". This can also happen when the input change is faster the the output and feed back network together can't respond fast enough. Since the open loop gain of the amp is very large, you notice that the open loop gain times 0.6v (the drop of the diodes) is well beyond the supply voltage, and the output is therefore saturated a long time before you turn on the diodes.
Therefore the diodes "clamp" excessive inputs to the circuit.
If you look at the way that an OpAmp with negitive feed back works; the amp attemps the adjust it's output until the two inputs are equal. The only time there will be a difference between the two inputs is when the amp has "run out of gain". This can also happen when the input change is faster the the output and feed back network together can't respond fast enough. Since the open loop gain of the amp is very large, you notice that the open loop gain times 0.6v (the drop of the diodes) is well beyond the supply voltage, and the output is therefore saturated a long time before you turn on the diodes.
Therefore the diodes "clamp" excessive inputs to the circuit.
Usually we put input limiting on a servo to keep it from 'latching up' This can happen because the drive voltage to the servo 'can' exceed the power supply voltage. Diodes are a crude fix, but series reversed polarity zeners, maybe +/- 6.2 V or more might be better. Diodes will always conduct a little and with a high Z source, can conduct enough to create a non-linearity. Zeners will not come on as easily until a certain threshold is reached.
Back to the diodes, I don't recommend leaving them out. Although theoretically they would not be needed, I'll bet you'll regret not having them. If you go back to the original BPA200 app note they recommend using 1N456A which are very low leakage diodes. Also if you read the app notes on the opamp they warn about over voltaging the inputs. The design rule of thumb (for real circuits, not theoretical circuits) say's you will see ~35V at those input some time. I don't think 6V zeners will work any better and will allow the amplifier to saturate, slowing recovery from the abnormal condition that they protect against.
I used the recommended LM833 which work fine. There is a heck of a lot more distortion and noise from all those caps, no matter how good a quality you use. Besides, that servo is a LOW PASS FILTER, also called an integrator. By definition it doesn't pass any audio unless you like listening to things below 5Hz! It's function is to remove any DC offset to keep the power dissapation down.
I built the BPA200 a few years ago and couldn't measure the distortion from it ( I don't have a distortion meter, but I do have use of a 500MHz Sampling scope).
I used the recommended LM833 which work fine. There is a heck of a lot more distortion and noise from all those caps, no matter how good a quality you use. Besides, that servo is a LOW PASS FILTER, also called an integrator. By definition it doesn't pass any audio unless you like listening to things below 5Hz! It's function is to remove any DC offset to keep the power dissapation down.
I built the BPA200 a few years ago and couldn't measure the distortion from it ( I don't have a distortion meter, but I do have use of a 500MHz Sampling scope).
Sorry; spent the last few weeks writing test procedures. I seem to have forgotten how to get to the point!
The BPA200 works great! I used a chunk of heatsink I scavanged from a 6kW power inverter about 12"x6"x6" and a PC fan which solved the cooling problem.
The biggest problem WAS the power transformer... until I started working in the test equipmet lab at a transformer factory! We just walked about 100' from the lab and wound a 2KVA torroid. Took about 30 minutes. (WAAAY OVERKILL by the way)
I meant that I had no good way to measure distortion other than putting in a signal and dividing the output down and subtracting the input from it on a 500MHz scope.
The BPA200 works great! I used a chunk of heatsink I scavanged from a 6kW power inverter about 12"x6"x6" and a PC fan which solved the cooling problem.
The biggest problem WAS the power transformer... until I started working in the test equipmet lab at a transformer factory! We just walked about 100' from the lab and wound a 2KVA torroid. Took about 30 minutes. (WAAAY OVERKILL by the way)
I meant that I had no good way to measure distortion other than putting in a signal and dividing the output down and subtracting the input from it on a 500MHz scope.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.