Author Topic: uSupply - LM350T instead of LT3080  (Read 10723 times)

0 Members and 1 Guest are viewing this topic.

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #25 on: December 22, 2016, 04:05:01 pm »
It will still be unstable, I'm afraid. As long as you have another gain stage inside the op-amp's feedback loop (T1 in this case) it will be prone to oscillation.

So I need to just completely remove T1 and it will be fine?
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: fi
  • Country: Broken Britain
Re: uSupply - LM350T instead of LT3080
« Reply #26 on: December 22, 2016, 04:25:43 pm »
So I need to just completely remove T1 and it will be fine?

I'd try swapping it for a PNP, and maybe taking it's collector to the -3V. You'd have to swap the two inputs to the op amp around. I don't think the large reverse b-e V would be a problem, keep the 10K.
CML+  That took much longer than I thought it would.
 
The following users thanked this post: IrejectYourReality

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #27 on: December 22, 2016, 04:34:50 pm »
Removing T1 makes things a little better, but you still have the problem of not having a defined minium resistance at the output of the LM350. So an extra resistor at the output (before voltage feedback of cause) is more or less needed.  It is possible too, to move the resistor to the GND side of the load and than use it as the shunt.   

Removing T1 also means the OP for current regulation should be relatively fast, as it sets the speed in preventing a big current spike. So the LM358 would be rather slow in this case.

An other problem might be the shunt at the input side of the LM350. The current sensing amplifier should be relatively fast - the may408x is not that fast.
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #28 on: December 22, 2016, 04:56:05 pm »
What value resistor should I use to define the minimum output resistance? I would guess something around 1-10 ohm? (The maximum output of the supply should be 25V 2A)

If the LM358 is too slow in this case, what opamp would be better?
« Last Edit: December 22, 2016, 04:58:29 pm by IrejectYourReality »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #29 on: December 22, 2016, 05:20:32 pm »
For a 2 A supply I would suggest something like 0.1 to 0.3 Ohms. This is not very much, but could make a big difference compared to a few mOhms parasitic resistance and the output impedance of the LM350. If you already have a negative supply, one could use something like an TL072. Many OPs to choose from in this range.

The problem with the LM358 is it's very limited slew rate (e.g. 0.3V /µs) and the additional cross over distortion (add's another about 2-50 µs delay if current changes sign).
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #30 on: December 22, 2016, 06:05:42 pm »
If I add said resistor at the output of LM350, could I use that resistor as a current shunt, or is it better to measure at the input of the regulator? Also, if I wanted to put the resistor to the lowside, what amp should I use for the measurement, would INA214 work well? (if I understand correctly the 4080 is high side only.)
« Last Edit: December 22, 2016, 06:30:30 pm by IrejectYourReality »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #31 on: December 22, 2016, 07:30:22 pm »
One could use the resistor behind the regulator as a shunt, but this is not that easy, as there can be lot of common mode voltage with changing output voltage. It should be possible with a floating current regulator - though the circuit is not that simple with central control. A floating current regulator has some benefits.

If the shunt on the low side is used for current regulation one could use a "simple" OP to amplify the current signal. This also needs an extra amplifier to transfer the set-point signal for the voltage regulation.
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #32 on: December 22, 2016, 09:27:21 pm »
So, what will be better/more accurate? Measuring the current at the input of LM350 or measuring it at lowside? I would like to get around 1mA resolution if possible. And if the max4080 is too slow for the cc to work properly, what should I use instead of it?
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9902
  • Country: us
  • DavidH
Re: uSupply - LM350T instead of LT3080
« Reply #33 on: December 23, 2016, 01:35:00 am »
I am still concerned about sinking the current to the negative supply. As I said, I want to use a PWM voltage inverter, which is a very weak supply and can generate only couple of mA. If I sink the 12.5 mA to this supply, isn't it basicaly the same process as if the supply itself sourced equivalent opposite current? I fear, that by sinking too much current into this supply, the negative voltage will drop and the supply will no longer be able to go close to 0V at the output. Is this a valid concern?

It is a valid concern if you need the output to get all the way to 0.0 volts and the negative bias supply cannot support the sink current required.  The LM350 requires a minimum load current of roughly 10 milliamps (worst case) whether that comes from the resistor between the output and adjustment pin or from a separate load on the output.

If your negative bias supply is not going to support the needed current, then your only choice if you want to continue to use the LM350 will be a low compliance current sink from the output to ground and a larger resistor between the output and adjustment pin to supply the current through the diodes into the operational amplifier outputs.  1.25 volts divided by 1.2k will yield about 1 milliamp through the diode which is currently forward biased.  I usually run the entire LM317/LM350 10 milliamp minimum output current through the diodes into PNP emitter followers.  This is also a good place to put a pair of LEDs to show if the output is voltage or current limited.

The minimum output voltage will then be limited by the load and the current sink to ground.  With no load, is a minimum output voltage of about 0.3 volts enough?  How close to 0.0 volts do you need?

You could use a different regulator (1) which does not have such a high minimum current or replace the regulator with a discrete transistor stage.  Personally I like using integrated regulators because of their built in protection circuits.

While not the power supply I would emulate exactly, the well documented Tektronix PS501 and PS503 use discrete pass transistor output stages and *still* pull a current to below ground so that they can achieve a 0.0 volt output with good stability.  The PS503 actually deliberately regulates to a few millivolts below ground to account for the worst case offset voltage of the error amplifier.

(1) I did a quick search and did not find any direct replacements for the LM317/LM350 with a lower minimum output current.
 
The following users thanked this post: IrejectYourReality

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: fi
  • Country: Broken Britain
Re: uSupply - LM350T instead of LT3080
« Reply #34 on: December 23, 2016, 04:01:02 am »
If you're brave and rich you could try the www.linear.com/product/LT3083
Minimum load 1mA and goes down to 0V. Max 23V though.

You might be able to fiddle the 1mA min load something like this, I'll try it when I get a round tuit.
« Last Edit: December 23, 2016, 04:21:23 am by StillTrying »
CML+  That took much longer than I thought it would.
 
The following users thanked this post: IrejectYourReality

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9902
  • Country: us
  • DavidH
Re: uSupply - LM350T instead of LT3080
« Reply #35 on: December 23, 2016, 05:01:31 am »
I designed it myself, when I noticed many people were using separate voltage references and trimmer resistors to get 0V from the LM317. I realized there's a more effective way to zero the output.

I do not remember seeing this idea before.  It is a great solution if you do not want to add a buffered -1.25 volt reference and have a spare operational amplifier.  Old application notes often use an LM329 or similar shunt regulator.
 
The following users thanked this post: IrejectYourReality

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9902
  • Country: us
  • DavidH
Re: uSupply - LM350T instead of LT3080
« Reply #36 on: December 23, 2016, 06:30:18 am »
The LM317/LM350 already has a series output resistance for its own current sense and emitter ballasting.

An added output resistance is an ideal place for the current sense if you make provisions to handle the wide common mode voltage swing.  If the regulator can be floating, then placing the current sense resistor on the ground side makes this trivial but either high or low side can work fine.  The first bench supply I designed for myself did this using high side current sensing and a 4 resistor difference amplifier but the frequency compensation needed to make it stable (see below) made for poor performance.

The 324/358 is slow but no slower than the 741 which is plenty fast for this application.  JFET amplifiers have an advantage in slew rate which will make recovery faster.  Beware of using an operational amplifier which is too fast; the extra frequency compensation needed will slow it down reducing its slew rate and and increasing its recovery time so there is little advantage.  There are ways to fix this but they add complexity.
 
The following users thanked this post: IrejectYourReality

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #37 on: December 23, 2016, 09:28:07 am »
The LM358 and 741 have a rather low slew rate and limited bandwidth. For the shown type of regulator with a low output impedance power stage voltage regulation is easy (no problem with 358). But the cross over from CV to CC is somewhat difficult and a high slew rate and also a not so low bandwidth is a good idea. Worst case the OP for current regulation has to go all the way from around 20 V to near zero before current limiting even starts to work. At 0.3 V/µs , this is some 60 µs without active current limit and thus a larger overshoot in the current. Delays from compensation and recovery of the LM350 come on top.  I know there are circuits using the LM358/324 in such a circuit - but this really compromises performance. To really take advantage of a faster OP, the compensation needs to be done in a way to keep the extra delay short. This is possible at least to a large part for the critical CV/CC transition.

The easy way is having the shunt at the low side - it also gets rid of the slow max4080.
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #38 on: December 23, 2016, 11:04:03 am »
If your negative bias supply is not going to support the needed current....
What if I just added a small inverting switchmode powersupply? It is cheap and I have plenty of room on the pcb.


This is my first try at a lowside current sensing. Is this enough, or do I need something more complicated? I googled and used the MCP6V81UT-E/LTY beacause I do not need it to operate at high voltage and it is much more precise than TL072 (correct?). I would get the -1.25V by using a voltage divider or some sort of reference/regulator. I would like to have 1mA accuracy for current sensing. I do not really care that much for the cc accuracy, for example if I set max current at 200 mA and the supply limits it at 195 mA, it is fine.
 
« Last Edit: December 23, 2016, 02:10:45 pm by IrejectYourReality »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #39 on: December 23, 2016, 05:06:25 pm »
The circuit diagram still has a few points missing / wrong:

There should be a compensation for the voltage drop at the shunt. This could be adding a faction of the voltage at the shunt to the + input together with the set signal.

The current sensing resistors have to go to GND and the amplifier sensing from the other side.

There is still the output capacitance and the compensation parts missing.
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #40 on: December 23, 2016, 05:51:47 pm »
There should be a compensation for the voltage drop at the shunt. This could be adding a faction of the voltage at the shunt to the + input together with the set signal.

Added it in, I supose this means that 1V from DAC no longer gets me 10V at the output, but 9.09V, correct? But this can be compensated for easily in software.

The current sensing resistors have to go to GND and the amplifier sensing from the other side.
Of course, that is just o total brainfart on my side   :palm:

 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #41 on: December 23, 2016, 06:20:50 pm »
There is no absolute need for R18 if you have the shunt on the low side. The shunt at the low side also works to limit current gain. The 0.1 Ohms value could be close the the lower limit though.

The output capacitor should be directly across the load, not to GND. Likely a value of only 1 µF is also too small. I would more expect something like a 10-100 µF electrolytic in parallel with 100-300 nF ceramic.

I am not sure about the right value for R17. My crude look at the circuit gives me more like 91 K as an appropriate value. With 100 K I would expect a little residual DC output resistance (e.g. 10 mOhm range).

The part still missing is frequency compensation for the control loops. Usually this are RC series circuits from the neg input of OP1A/B to the output (or the adjust pin of the LM350). OP1B would need an extra resistor from the current signal. To improve stability with large capacitance at the output one might consider something like a 5 n in series with 10 K in parallel to R4.
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #42 on: December 23, 2016, 06:54:23 pm »
The datasheet of LM350 suggests either 1uF tantalum or 25 uF electrolytic cap on the output.
About R17, shouldnt it be 90K? Then Vout=10*(V2*1/10)+(V1*9/10), where V1 is the DAC1 IN.
I tried to add the frequency compensation, but I am not 100% sure that I understood everything you said. Also, I just guessed the values for R21, C6, R20 and C5, because i have no idea what kind of values should be used for this.



So, how did I do?  :)
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6119
  • Country: de
Re: uSupply - LM350T instead of LT3080
« Reply #43 on: December 23, 2016, 08:14:39 pm »
The circuit looks good now.

R17=90K would be the theoretical value - to make sure the output resistance is not negative, a slightly larger value might be a good idea.

The frequency compensation as shown is still rather slow, but I have not done a full simulation / calculation. I would expect C5 and C6 more in the 100pF to 1 nF range. R21 should smaller (e.g. <= 5 K) - currently the voltage loop gives a gain of about 1 at high frequencies - this is a little too much. To get a better guess for the compensation, I would use a simulation. I am not sure how different the LM350 behaves from a more normal darlington transistor / MOSFET.

For a faster transition from CV to CC mode, it might be a good idea to move the right side of C5 to behind the diodes (adjust pin).
 
The following users thanked this post: IrejectYourReality

Offline IrejectYourReality

  • Contributor
  • Posts: 25
  • Country: cz
Re: uSupply - LM350T instead of LT3080
« Reply #44 on: December 23, 2016, 09:30:14 pm »
Thank you very much for all the advice, I will design the rest of the circuit and the pcb and will report back with results.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf