Hi warm folks!
Can we use a positive regulator to produce a negative voltage?
I'm looking to build a +-15v (adjustable) using 1 or 2 rectifier bridge, and use 2 identical positive regulator (LM1086, LT1086, LT1963, etc.).
I've found a schematics but they use 2 bridges. Can we use this with only one bridge, and remove the first stage of the schematics (darlington and opamp).
Possible ?
.
Can we use a positive regulator to produce a negative voltage?
I'm looking to build a +-15v (adjustable) using 1 or 2 rectifier bridge, and use 2 identical positive regulator (LM1086, LT1086, LT1963, etc.).
I've found a schematics but they use 2 bridges. Can we use this with only one bridge, and remove the first stage of the schematics (darlington and opamp).
Possible ?
.
Attachments
One bridge and center-tapped xformer: NO
Two bridges, isolated secondaries (and basic topology as shown): YES
- Klaus
Two bridges, isolated secondaries (and basic topology as shown): YES
- Klaus
KSTR said:
Thanks.
I tried with only one bridge. It was funny. Boom.
If I follow you, I can have for each:
15v secondary + bridge + caps + LM + caps
And I connect the ground at output (or one ground plane for both ?)
.
Hi,
a dual positive supply can be series connected AT THE OUTPUT to give a dual polarity supply.
But you cannot use a centre tapped nor a single bridge to create the two completely separate supplies needed for the commoned output dual polarity.
a dual positive supply can be series connected AT THE OUTPUT to give a dual polarity supply.
But you cannot use a centre tapped nor a single bridge to create the two completely separate supplies needed for the commoned output dual polarity.
If fact, if you already have the right transformer, the cost is only one bridge more and more work to design the pcb?
For the ground plane, on the PCB. Can we build it using all the component in red or we MUST build 2 separated ground planes et connect them at the exit (green).
.
For the ground plane, on the PCB. Can we build it using all the component in red or we MUST build 2 separated ground planes et connect them at the exit (green).
.
Attachments
Hi,
the sensing resistors feeding the regulator control circuits MUST SENSE the OUTPUT voltage, not some arbitrary point on a ground plane.
the sensing resistors feeding the regulator control circuits MUST SENSE the OUTPUT voltage, not some arbitrary point on a ground plane.
AndrewT said:
Like this?
The objective is to produce someting better than LM317/LM337 couple with the minimum component and lower input voltage.
.
Forgot the first PDF. I did a mistake. I not used the right pinout for the LT1965.
I've still some problems with grounding...
.
I've still some problems with grounding...
.
Hi Stef,
you need to do more reading on basic circuits before you try to make them more complex.
Fuses F1 & F2 might blow on first switch on. You can either make them bigger (less protection) or place them after the main smoothing caps.
The upper bridge rectifier is back to front.
The 10VA will give 410mAac maximum continuous current.
After rectifying and smoothing the maximum continuous DC current is ~200mA. A 150mA transformer (3.6VA) will not do.
If you want 15Vdc then you need a 15Vac transformer.
10VA will now give 165mA @15Vdc.
The inductor before the first cap has not been designed to make this a choke regulated PSU.
It will still act as a capacitor input PSU.
the inductor in the lower 0v lead should be on the lower -ve lead.
The sense resistors should be connected to the output.
the ground in the lower reg should be connected to the negative supply line for this regulator (just like the upper reg is connected to it's negative supply line.
Are these regulator stable with a low ESR cap on the output?
where should C7 & C17 be connected?
Sorry to appear so hard, but you would be better building a simple PSU first and understand what you have achieved, before you go regulated.
you need to do more reading on basic circuits before you try to make them more complex.
Fuses F1 & F2 might blow on first switch on. You can either make them bigger (less protection) or place them after the main smoothing caps.
The upper bridge rectifier is back to front.
The 10VA will give 410mAac maximum continuous current.
After rectifying and smoothing the maximum continuous DC current is ~200mA. A 150mA transformer (3.6VA) will not do.
If you want 15Vdc then you need a 15Vac transformer.
10VA will now give 165mA @15Vdc.
The inductor before the first cap has not been designed to make this a choke regulated PSU.
It will still act as a capacitor input PSU.
the inductor in the lower 0v lead should be on the lower -ve lead.
The sense resistors should be connected to the output.
the ground in the lower reg should be connected to the negative supply line for this regulator (just like the upper reg is connected to it's negative supply line.
Are these regulator stable with a low ESR cap on the output?
where should C7 & C17 be connected?
Sorry to appear so hard, but you would be better building a simple PSU first and understand what you have achieved, before you go regulated.
AndrewT said:
Sorry but yesterday night I worked as a pig on this drawing. Too fast and with not edited copy/paste. This not my first psu (go to see my last one, the dedicated DEQ/DCX psu on the forum). I uploaded a corrected version of the drawing. Most of your comments was already corrected.
The fuses are not fuse but polyswitch. The LT1965 is a new low dropout regulator with low noise from Linear. It works with low esr caps. This is an example. We can replace it with others as the LT1086, LT1763 or others from National.
The idea is to build something different of the regular 317/337 couple or the more detailed Sulzer type psu (too many component in my case, no space).
AndrewT said:
Interesting. I use VK200 for their low price and small size since years. May be a bad habit. What is this story of "input psu capacitor"? Do you prefer other type of ferrite bead?
AndrewT said:
The LT1965 is a 290mV dropout. We can may be try 13 or 14v at input to get 15v.
AndrewT said:
No problem at all. If I ask for advice, it's because I have no problem to be corrected when wrong. I'm looking for that.
.
Hi,
that looks much better.
Check the datasheet again, LDO regulators are more susceptable to oscillation than the conventional regulators. Look for any advice and permitted capacitance on the output for alternative loads.
that looks much better.
Check the datasheet again, LDO regulators are more susceptable to oscillation than the conventional regulators. Look for any advice and permitted capacitance on the output for alternative loads.
AndrewT said:
I was in fact again reading the LT1965 datasheet... It is stable with Ceramic, Tantalum or Aluminum capacitors. The minimum is 10uF.
But some questions:
- Is the LT1965 really adapted to work with 15v output (look at noise and ripple chart)?
- The application note talk of the value of the resistor divider depending of the "9v clamp". I don't understand this part. My English is not always good.
In situations where the ADJ pin connects to a resistor
divider that would pull the ADJ pin above its 9V clamp voltage
if the output is pulled high, the ADJ pin input current
must be limited to less than 5mA. For example, a resistor
divider is used to provide a regulated 1.5V output from the
1.20V reference when the output is forced to 20V. The top
resistor of the resistor divider must be chosen to limit the
current into the ADJ pin to less than 5mA when the ADJ
pin is at 9V. The 11V difference between the OUT and ADJ
pins divided by the 5mA maximum current into the ADJ
pin yields a minimum top resistor value of 2.2k.
- Currently, I use 100nF Wima MKS2 to double the large capacitors. At output, we have 47uF (or 10uF is enough) tantalum and 100nF MKS2. Following the datasheet, may be an 100nF XR7 will be better. Any advice?
- What about the WK200?
Finally, I discovered a new regulator, the LT3080 (not yet available). It seems to be better for "high output voltage" than the LT1965 and I can reduce component count as it use only one external resistor. But, it's a "floating device" and I'm a little afraid by that. Any comment?
Hi,
no-one else seems to want to contribute.
Is the diode d1 going to the correct pin. I have seen it going to the output pin to discharge the cap if the output get shorted.
Is the LED the right way round?
C8 looks a bit small.
Have you worked out the minimum and maximum voltage on the smoothing caps for the worst range of mains voltage and output current?
You need this to check for dissipation and heatsink sizing.
no-one else seems to want to contribute.
Is the diode d1 going to the correct pin. I have seen it going to the output pin to discharge the cap if the output get shorted.
Is the LED the right way round?
C8 looks a bit small.
Have you worked out the minimum and maximum voltage on the smoothing caps for the worst range of mains voltage and output current?
You need this to check for dissipation and heatsink sizing.
Look at the new schematics. I changed the filters and component numbering changed too.
D1 and C9 is a soft start.
For the C6 filter cap, it should be ok for the needed current (150mA max). The resistor devider is just used to use a small value cermet. It can be replaced by only one resistor. I'm not sure of the correct values as it is not explained in the datasheet.
Heatsink = (18,5v - 15v) x (0,150A/58) + (18,5v - 15v) x 0.150A = 528mW = 40°C with a T0220 package = no need of heatsink or a small one.
Correct me if wrong. I'm bad with numbers...
D1 and C9 is a soft start.
For the C6 filter cap, it should be ok for the needed current (150mA max). The resistor devider is just used to use a small value cermet. It can be replaced by only one resistor. I'm not sure of the correct values as it is not explained in the datasheet.
Heatsink = (18,5v - 15v) x (0,150A/58) + (18,5v - 15v) x 0.150A = 528mW = 40°C with a T0220 package = no need of heatsink or a small one.
Correct me if wrong. I'm bad with numbers...
Hi,
what are the lowest and highest voltages that feed the input to the regulator? The answer is not 15V.
what are the lowest and highest voltages that feed the input to the regulator? The answer is not 15V.
AndrewT said:
As well if we have a 15V transformer.
Between 18.5V and 20V at input. But depends of the transformer and the bridge, etc.
A tried to spice the LT3080 with the Linear tool. I don't understand yet how the Rset system work.
.
In fact, the regulator is easy to use. You only need to use a capacitor with a minimum ESR of 0.5 Ohms for the output capacitor with a minimum value of 2.2uF.
Inclosed new spice.
Inclosed new spice.
- Status
- Not open for further replies.