Now that I look at it I see that I don't need a protection diode for the pre reg. D2 and D3 will do the same thing🙂 . Stand by for yet another schematic.
D1
Change D1 a 1 or 2 watt 20 Volt to 25 Volt zener connected with same polarity. This will protect from over voltage during start up before the output voltage comes up. The zener will still provide the origional function of D1 as well.
Again from date sheet:
ABSOLUTE MAXIMUM RATINGS
Input-to-Output Voltage Differential
“C” Grades .......................................................... 30V
“I” Grades ............................................................ 30V
“M” Grades .......................................................... 35V
Art
Change D1 a 1 or 2 watt 20 Volt to 25 Volt zener connected with same polarity. This will protect from over voltage during start up before the output voltage comes up. The zener will still provide the origional function of D1 as well.
Again from date sheet:
ABSOLUTE MAXIMUM RATINGS
Input-to-Output Voltage Differential
“C” Grades .......................................................... 30V
“I” Grades ............................................................ 30V
“M” Grades .......................................................... 35V
Art
Old and new
When the LT1085 was released there were several supply projects using it, even advising it for negative supplies provided you had separate secondaries.
They also suggested rather big output caps, 2200uF or more, which I wonder how could get away with. Maybe because it was also advised putting a load resistor at the output, like 1K or so.
This output resistor might be a good thing to add on your present supply.
Though I never tried it, some designers suggest large capacitors for bypassing the adjust pin. This seems to improve ripple rejection and noise. But I would like to hear some comments from someone who actually did it and saw the improvement.
But if the principle is the same as the multiplier, it should work.
BTW: isn't replacing a zener as D1 the way to use these regulators with high voltages?
Carlos
When the LT1085 was released there were several supply projects using it, even advising it for negative supplies provided you had separate secondaries.
They also suggested rather big output caps, 2200uF or more, which I wonder how could get away with. Maybe because it was also advised putting a load resistor at the output, like 1K or so.
This output resistor might be a good thing to add on your present supply.
Though I never tried it, some designers suggest large capacitors for bypassing the adjust pin. This seems to improve ripple rejection and noise. But I would like to hear some comments from someone who actually did it and saw the improvement.
But if the principle is the same as the multiplier, it should work.
BTW: isn't replacing a zener as D1 the way to use these regulators with high voltages?
Carlos
Hi Carlos,
From what I understand (little). You can use the LT1085 and several other regulators for high voltage as long as the supply is floating and the input to output difference is within the recommended 30v. It seems to me though that such a supply would be a potential hum monster.
From what I understand (little). You can use the LT1085 and several other regulators for high voltage as long as the supply is floating and the input to output difference is within the recommended 30v. It seems to me though that such a supply would be a potential hum monster.
LT1085
And if someone shorts the output or when the regulator is powering up the voltage on the output comes up slowly.... you could wind up with this sequence of events!
Procede at your own risk.
Typical audio regulator circuit for adjustable three terminal regulators and one man's story:
http://www.welbornelabs.com/ps1.htm
http://www.stereophile.com/showarchives.cgi?54:6
Art
And if someone shorts the output or when the regulator is powering up the voltage on the output comes up slowly.... you could wind up with this sequence of events!
Procede at your own risk.
Typical audio regulator circuit for adjustable three terminal regulators and one man's story:
http://www.welbornelabs.com/ps1.htm
http://www.stereophile.com/showarchives.cgi?54:6
Art
DIY supply kits
Art,
Those below are two examples where the output cap are large:
http://www.welbornelabs.com/ps1.htm
http://www.stereophile.com/showarchives.cgi?54:6
It's 3300uF on Welborne's and 4700uF on Stereophile's. It was also large on several DIY supplies The Audio Amateur published for the LT1085.
In both supplies it would nice if someone tested them with a several MHz scope and see if they oscillate or not. And also see what happens if you swap the output cap and the adjust cap.
--------------------------------------------------------------------------------
You can use the LT1085 and several other regulators for high voltage as long as the supply is floating and the input to output difference is within the recommended 30v.
--------------------------------------------------------------------------------
Well, that's what I thought too. But some suggested it may blow the output cap at startup when using high voltages. And I wonder if something could not happen to the regulator then.
Carlos
Art,
Those below are two examples where the output cap are large:
http://www.welbornelabs.com/ps1.htm
http://www.stereophile.com/showarchives.cgi?54:6
It's 3300uF on Welborne's and 4700uF on Stereophile's. It was also large on several DIY supplies The Audio Amateur published for the LT1085.
In both supplies it would nice if someone tested them with a several MHz scope and see if they oscillate or not. And also see what happens if you swap the output cap and the adjust cap.
--------------------------------------------------------------------------------
You can use the LT1085 and several other regulators for high voltage as long as the supply is floating and the input to output difference is within the recommended 30v.
--------------------------------------------------------------------------------
Well, that's what I thought too. But some suggested it may blow the output cap at startup when using high voltages. And I wonder if something could not happen to the regulator then.
Carlos
Love the symbols ART. LOL. I don't plan on using the LT1085 or any other regulator IC for High voltage. I think that I would have to use a discrete regulator using high voltage Mosfets in that case.
Also notice Carlos that the welborne labs supply is using not an output resistor but a 1K resistor in parallel with the load installed before the 2200uF cap. I wonder what the purpose of this resistor is?
1K resistor preloads regulator for a minimum output current. This also forces the pass transistor in the regulator to operate baised further into the Class A region. Notice the specs are rated for a minimum 10mA DC load in the data sheet. Also there is a mimimun current required for the regulator to operate.
Attachments
I see. So with loads that utilize less than the minimum current load you would preload the regulator?
preloading
Actually I have seen preloading used to improve regulator performance even at loads above minimum current
Art
Actually I have seen preloading used to improve regulator performance even at loads above minimum current
Art
Well unless there is something else that is glaringly wrong with my diagram(values) I will start laying out the PCB for it. Thanks for the lessons guys.
Art. Do you think that the circuit will be stable with the 15000uF cap before the pre reg? I can certainly reduce the value.
So what are you using this regulator for?
Sorry, read the post, but did not see what this thing will be used for. I assume a pre-amp?
Some thoughts on many of the points:
- Large output capacitors. Most regulators I have found are stable with large output capacitors. The reality, is that the load produced by that large output cap will damp any oscillations that may occur. Fortunately, the ESR of that big cap is probably not great at high frequencies (but no guarentee).
- Floating regulator - Yes it can work, but you need to be very careful. When I do this, I put a Zener across the regulator at a voltage less than the max for the regulator, but of course more than the expected operating drop. Shorts will always be an issue, but that is life. This will prevent issues of slow startup which is exactly what will happen with the large output capacitor.
- Caps -- Well the world is not moving away from Tantalums. When you need low ESR and good high frequency decoupling, they are still considerably cheaper than organic electrolyte, cheaper than MLCC, and more effective than standard electrolytics including most switching power supply electrolytics. That said, Tantalums have high dielectric absorption. That should only be a problem if the voltage level on your power supply is changing which is exactly what you are trying not to do. My real experience is that they work just fine in many power supplies (but never in a signal path), however, for the cost of DIY, I usually err on the side of caution and use organic. That said and done, you probably don't need 15,000 uf of organic electrolyte, but some caps to handle the high frequency would be good as standard ones are not. The rest of the higher frequency filtering should happend right at your application circuit.
- So why not use a regulator designed for higher input voltages? The LT parts are good, but after floating them, the benefits could and probably will be gone. There are a few that I remember do 60V (load dump for automotive), and lots that have 40 volt ranges.
.. just some thoughts from an arm chair quarterback...
Sorry, read the post, but did not see what this thing will be used for. I assume a pre-amp?
Some thoughts on many of the points:
- Large output capacitors. Most regulators I have found are stable with large output capacitors. The reality, is that the load produced by that large output cap will damp any oscillations that may occur. Fortunately, the ESR of that big cap is probably not great at high frequencies (but no guarentee).
- Floating regulator - Yes it can work, but you need to be very careful. When I do this, I put a Zener across the regulator at a voltage less than the max for the regulator, but of course more than the expected operating drop. Shorts will always be an issue, but that is life. This will prevent issues of slow startup which is exactly what will happen with the large output capacitor.
- Caps -- Well the world is not moving away from Tantalums. When you need low ESR and good high frequency decoupling, they are still considerably cheaper than organic electrolyte, cheaper than MLCC, and more effective than standard electrolytics including most switching power supply electrolytics. That said, Tantalums have high dielectric absorption. That should only be a problem if the voltage level on your power supply is changing which is exactly what you are trying not to do. My real experience is that they work just fine in many power supplies (but never in a signal path), however, for the cost of DIY, I usually err on the side of caution and use organic. That said and done, you probably don't need 15,000 uf of organic electrolyte, but some caps to handle the high frequency would be good as standard ones are not. The rest of the higher frequency filtering should happend right at your application circuit.
- So why not use a regulator designed for higher input voltages? The LT parts are good, but after floating them, the benefits could and probably will be gone. There are a few that I remember do 60V (load dump for automotive), and lots that have 40 volt ranges.
.. just some thoughts from an arm chair quarterback...
Brain Fart,
I was following a few schematics back.
The 15K is on the input... hmmm that may not provide any where near the benefit you may think. Has anyone put this into spice and determined what the regulation bandwidth is of the regulator? That cap on the input has limited effective bandwidth. Within that limited bandwidth, the regulator will probably provide enough regulation to not warrant such a large cap. It is not going to make it worse, but it may not make it better.
However, the output cap looks pretty small, 150uF? If you get heavy instantaneous draws on the output, it is going to pull that cap down no matter how many uF are on the input. You are limited by the bandwidth of the regulator and you can't change that except by improving your regulator bandwidth or increasing the size of the output capacitor. You seem set on the regulator, so the output cap needs to increase. Also remember that the output of the regulator is not a true push pull. If the circuit you are driving has any inductive effects, then you may actually get feedback from the circuit that tries to drive the rails up. Big fast caps are the only thing that will fix this.
.... More thoughts of an air chair quarterback who put his glasses on this time and apologizes for the not completely accurate last post.
I was following a few schematics back.
The 15K is on the input... hmmm that may not provide any where near the benefit you may think. Has anyone put this into spice and determined what the regulation bandwidth is of the regulator? That cap on the input has limited effective bandwidth. Within that limited bandwidth, the regulator will probably provide enough regulation to not warrant such a large cap. It is not going to make it worse, but it may not make it better.
However, the output cap looks pretty small, 150uF? If you get heavy instantaneous draws on the output, it is going to pull that cap down no matter how many uF are on the input. You are limited by the bandwidth of the regulator and you can't change that except by improving your regulator bandwidth or increasing the size of the output capacitor. You seem set on the regulator, so the output cap needs to increase. Also remember that the output of the regulator is not a true push pull. If the circuit you are driving has any inductive effects, then you may actually get feedback from the circuit that tries to drive the rails up. Big fast caps are the only thing that will fix this.
.... More thoughts of an air chair quarterback who put his glasses on this time and apologizes for the not completely accurate last post.
I would also consider a zener between VI and GND on the regulators...... remember that those caps between the GND on the regulator and the actual GND are shorts when it first turns on. The resistor and inductor in front of the regulator will probably prevent a problem and the parts will no doubt take over-voltage for a while, but you may want to err on the side of caution to start.
Sorry, I keep liking this less and less the more I look at it....
By doubling up the regulators you are drastically increasing the gain of the regulator and hence the regulation at low frequencies. However, at higher frequencies, you could be making things worse. I would expect that the gain at low frequencies is so substantial that the effect of the huge input cap is minimal. The gain at high frequencies and the input to output resistance at high frequencies is so low that the large input cap again is useless. You may also introduce some interactions between the regulators that may increase the settling time.
I am not saying it won't work, but I am not convinced it will be better at what you want it to do. I would model it in spice if you have it and play around with the values.
I should also clarify the zeners across the regulators. I realize you have one across both, but the output of the first regulator will be at gnd when this thing first turns on. I don't know how robust these things are and if they can handle 40V across them while the cap charges. Again, the input resistor/inductor may prevent this.
By doubling up the regulators you are drastically increasing the gain of the regulator and hence the regulation at low frequencies. However, at higher frequencies, you could be making things worse. I would expect that the gain at low frequencies is so substantial that the effect of the huge input cap is minimal. The gain at high frequencies and the input to output resistance at high frequencies is so low that the large input cap again is useless. You may also introduce some interactions between the regulators that may increase the settling time.
I am not saying it won't work, but I am not convinced it will be better at what you want it to do. I would model it in spice if you have it and play around with the values.
I should also clarify the zeners across the regulators. I realize you have one across both, but the output of the first regulator will be at gnd when this thing first turns on. I don't know how robust these things are and if they can handle 40V across them while the cap charges. Again, the input resistor/inductor may prevent this.
Hi Alvaius,
I have to be honest here. I don't have a clue(yet) what you are talking about. I'm sure that your right but I don't understand. This all started out because I wanted a good regulator for the circuit that's posted in this thread. I'm not very far along in electronics. I know just enough to be dangerous. I'm learning but I'm not there yet. The basic schematic for this circuit came from the LT317 datasheet. I'm always open for any suggestions. That was not meant to be sarcastic. Really. Any help would be appreciated. My brain is a sponge.
I have to be honest here. I don't have a clue(yet) what you are talking about. I'm sure that your right but I don't understand. This all started out because I wanted a good regulator for the circuit that's posted in this thread. I'm not very far along in electronics. I know just enough to be dangerous. I'm learning but I'm not there yet. The basic schematic for this circuit came from the LT317 datasheet. I'm always open for any suggestions. That was not meant to be sarcastic. Really. Any help would be appreciated. My brain is a sponge.
- Status
- Not open for further replies.