Author Topic: Non-linear CV/CC control in schematic  (Read 5663 times)

0 Members and 1 Guest are viewing this topic.

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #25 on: March 03, 2019, 11:49:36 pm »
I drew this up quickly. Ill check for possible mistakes later.
The values aren't supper critical.

Thank you. :)

I was hooking the pot up wrong.. everything else was correct.
I'm sorry you guys have to dumb things down sometimes. I appreciate it though :)

Circuit has been modified and seems to function well. I adjusted the voltage divider and adjusted R1 slightly to get a bit smoother voltage control.

Vdroop under a 12A load is about 0.1V. There's not a lot I can do about that eh? Something to do with VRef or current sense maybe?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #26 on: March 04, 2019, 12:33:06 am »
Good to see things work.
What is the minimum regulation voltage?
The lowest loaded output voltage while still being controllable with the voltage adj Pot.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #27 on: March 04, 2019, 01:03:50 am »
The cause of the voltage droop should be traceable by looking for voltage changes while loading and unloading the output.
Such as, Vref, Ninv, Inv and the top of R20, all WRT pin7 -V.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 
The following users thanked this post: Silver_Pharaoh

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #28 on: March 05, 2019, 01:37:35 am »
The cause of the voltage droop should be traceable by looking for voltage changes while loading and unloading the output.
Such as, Vref, Ninv, Inv and the top of R20, all WRT pin7 -V.

Seems Pin 2 and Pin 10 have large Voltage fluxuations:

Pin 2:
12A load: 5.575V
Low load: 5.109V

Pin 10:
12A load: 7.05V
Low load: 5.80V

I have not had nay luck much in modifying the circuit.

The Vdroop gets worse after I pull 11A from the circuit. Below 11A the Vdroop is very minimal.
At 11A, Vdroop is 0.02V and gets worse the more I pull after that. At 11.5A Vdrop is 0.05V...

Could is be the balancing resistors?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #29 on: March 05, 2019, 01:52:19 am »
The cause of the voltage droop should be traceable by looking for voltage changes while loading and unloading the output.
Such as, Vref, Ninv, Inv and the top of R20, all WRT pin7 -V.

Seems Pin 2 and Pin 10 have large Voltage fluxuations:

Pin 2:
12A load: 5.575V
Low load: 5.109V

Pin 10:
12A load: 7.05V
Low load: 5.80V

I have not had nay luck much in modifying the circuit.

The Vdroop gets worse after I pull 11A from the circuit. Below 11A the Vdroop is very minimal.
At 11A, Vdroop is 0.02V and gets worse the more I pull after that. At 11.5A Vdrop is 0.05V...

Could is be the balancing resistors?
The change at pin 10 is expected as the output transistors need more drive with increased load .
The change at pin 2 is very close to causing the CL transistor to conduct especially if the LM723 is warm.
I would be curious to know how much drive current pin 10 is having to source. This can be done by inserting a 10Ω series resistor and measuring its voltage drop.
« Last Edit: March 05, 2019, 04:31:48 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #30 on: March 05, 2019, 01:59:54 am »
You would be able to tell how close operation is to the current limit threshold by moving the Pot.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #31 on: March 06, 2019, 12:52:03 am »
You would be able to tell how close operation is to the current limit threshold by moving the Pot.

Not sure what you mean by this?


Voltage drop is very low across the 10 \$\Omega\$ resistor.

I had to change the current limit pot value back to 10K. I forget why I changed it but it didn't let me reduce the current below 10A. I must have been tinkering with it and forgot to change it back.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #32 on: March 06, 2019, 01:01:54 am »
The idea is to figure out wether the CL transistor inside of the LM723 is starting to conduct or not.
The simplest way is to apply a load that just starts to make the output voltage drop then make a small adjustment to the current limit Pot to see if it affects the output voltage.
Measuring the voltage between pins 2 and 3 would be useful.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 
The following users thanked this post: Silver_Pharaoh

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #33 on: March 06, 2019, 01:27:45 am »
The idea is to figure out wether the CL transistor inside of the LM723 is starting to conduct or not.
The simplest way is to apply a load that just starts to make the output voltage drop then make a small adjustment to the current limit Pot to see if it affects the output voltage.
Measuring the voltage between pins 2 and 3 would be useful.

Ahh I see.

Can't tinker much tonight, but I think you are correct in saying the CL transistor is starting to conduct.
It's hard to tell because I don't have fine control with the current adjust and the voltage drops near full load, so the current adjust pot is already maxed out..


I reduced the balancing resistors to 0.8 \$\Omega\$ each and the voltage shot back up to 5.1V like I have it set. This pulling 12A at the output.
Are the balancing resistors limiting max output current too much?
 

Offline diane

  • Newbie
  • Posts: 3
  • Country: us
Re: Non-linear CV/CC control in schematic
« Reply #34 on: March 06, 2019, 01:51:58 am »
Wow, such a difference in if of possibility. I only did a patent on the idea,  but if there is a possibility of creating without too much grevience, I will recommend that person's idea. Of course it may take another patent for idea. Read articals. About the building of assisted living quarters. 4 New ones were just finished within 5 to 10 miles of my area. They are filled and with waiting lists. Thank you for your help.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #35 on: March 06, 2019, 04:22:43 am »
The idea is to figure out wether the CL transistor inside of the LM723 is starting to conduct or not.
The simplest way is to apply a load that just starts to make the output voltage drop then make a small adjustment to the current limit Pot to see if it affects the output voltage.
Measuring the voltage between pins 2 and 3 would be useful.

Ahh I see.

Can't tinker much tonight, but I think you are correct in saying the CL transistor is starting to conduct.
It's hard to tell because I don't have fine control with the current adjust and the voltage drops near full load, so the current adjust pot is already maxed out..


I reduced the balancing resistors to 0.8 \$\Omega\$ each and the voltage shot back up to 5.1V like I have it set. This pulling 12A at the output.
Are the balancing resistors limiting max output current too much?
To increase the max current setting, just put a resistor from the junction of the 100Ω summing resistors to the junction of the sharing resistors(power supply's output). Start with 100Ω.
« Last Edit: March 06, 2019, 05:24:59 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #36 on: March 06, 2019, 05:05:54 am »
Wow, such a difference in if of possibility. I only did a patent on the idea,  but if there is a possibility of creating without too much grevience, I will recommend that person's idea. Of course it may take another patent for idea. Read articals. About the building of assisted living quarters. 4 New ones were just finished within 5 to 10 miles of my area. They are filled and with waiting lists. Thank you for your help.
Is this the thread that you meant to post in?
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #37 on: March 06, 2019, 08:51:38 am »
I would rather see R14 placed between B and E of Q1.  Then R15 should be something lower in resistance.
The low end of the current range will have a lot of thermal drift.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #38 on: March 07, 2019, 10:58:44 pm »
I would rather see R14 placed between B and E of Q1.  Then R15 should be something lower in resistance.
The low end of the current range will have a lot of thermal drift.

Isn't that where R14 currently is?

I've added in the 100 \$\Omega\$ resistor. Having it around 0.5 \$\Omega\$ gets be a Vdrop of a mere 0.01V which is sweet! Short circuit is around 14A
That resistor has the same effect as lowing all the balancing resistors to 0.08 \$\Omega\$

It is better to lower this new resistor instead of the balancing resistors isn't it? I imagine finding 0.08 \$\Omega\$ resistors might be pricey/hard. I haven't checked yet.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #39 on: March 07, 2019, 11:17:41 pm »
That looks right except for R11 which is label as "1" but must really be 100Ω because 1Ω would result in a very high maximum setting like 100A!
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #40 on: March 08, 2019, 12:33:11 am »
That looks right except for R11 which is label as "1" but must really be 100Ω because 1Ω would result in a very high maximum setting like 100A!

At 100 \$\Omega\$ max current seems to be around 12.8A?

Unless this is a quirk of LTSpice?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #41 on: March 08, 2019, 12:43:13 am »
That looks right except for R11 which is label as "1" but must really be 100Ω because 1Ω would result in a very high maximum setting like 100A!

At 100 \$\Omega\$ max current seems to be around 12.8A?

Unless this is a quirk of LTSpice?
That could be right. In situations that use a transistor'a Base to Emitter threshold voltage as a reference, some try and see needs to be done.
Also, the short circuited current setting will have to be a fair bit higher than the expected normal current draw to avoid voltage droop.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #42 on: March 08, 2019, 01:48:06 am »
That looks right except for R11 which is label as "1" but must really be 100Ω because 1Ω would result in a very high maximum setting like 100A!

At 100 \$\Omega\$ max current seems to be around 12.8A?

Unless this is a quirk of LTSpice?
That could be right. In situations that use a transistor'a Base to Emitter threshold voltage as a reference, some try and see needs to be done.
Also, the short circuited current setting will have to be a fair bit higher that the expected normal current draw to avoid voltage droop.

Make sense, otherwise the CL transistor would start to reduce the voltage as it conducts more and more right?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Non-linear CV/CC control in schematic
« Reply #43 on: March 08, 2019, 02:47:42 am »
As the Base to Emitter voltage increases with increasing output current, the CL transistor starts turning on gradually at first. This is what would be causing the voltage droop.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 
The following users thanked this post: Silver_Pharaoh

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #44 on: March 09, 2019, 05:09:31 pm »
Well I think this is the final version!
All seems to work well.

I included 2 other versions with a more linear & tighter voltage adjustment.
 

Offline Silver_PharaohTopic starter

  • Contributor
  • Posts: 32
  • Country: ca
Re: Non-linear CV/CC control in schematic
« Reply #45 on: March 09, 2019, 11:38:39 pm »
Wait, I just thought of the inrush current.
63kA? That can't be right...

I know the input cap is for smoothing but 47,000uF seems like a lot. Following this]https://www.electro-tech-online.com/threads/how-to-calculate-the-value-of-a-smoothing-capacitor.106374/]this] formula I get about 56,000uF assuming 14A with a 2v drop. That's a lot of capacitance.
Would I be able to reduce the capacitance to say 5,600uF instead?

Using an NTC thermistor with a resistance of 2.5 \$\Omega\$ @ 25C would limit the inrush to 10A roughly which is fine, but having say a 30mOhm resistance at full load reduces the max current I can draw at 24V to 10.5A. I can get around this buy suppling the LM723 with 28V instead of 27V.

Am I over thinking this too much?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf