Author Topic: how to design fast bench supply with CC and CV?  (Read 10262 times)

0 Members and 1 Guest are viewing this topic.

Online exe

  • Supporter
  • ****
  • Posts: 1805
  • Country: nl
  • self-educated hobbyist
how to design fast bench supply with CC and CV?
« on: April 07, 2020, 08:35:46 pm »
Hello!

This is a continuation of https://www.eevblog.com/forum/testgear/output-capacitance-of-higher-end-bench-power-supplies/ .

I would like to build a high-precision, high-performance power supply. How would I do that?

Some general specs to give direction:
1. Digital control and readback
2. Output voltage range 0..15V (min 12V)
3. Current ranges 0..100mA and 0..3A (min 2A) with 1mA step, and 0.25mA accuracy. Or, alternatively, I can simply build two channels with different current capabilities.
4. Precision current set and readback. Many power supplies have cannot do that because of different parasitic currents, although error can be pretty small
5. Remote sensing. Well, remote up to output terminals, I don't long external connectors.
6. Stability with minimum output capacitance
7. Stability with large output capacitance
8. Performance... Well, how about recovery time from 0 to 100% load and back? Say, in less than 20us. Also fast CV/CC switch without glitches. Not sure how to define that.

Bonus points if PSU can:
1. Sink current
1. Bipolar
1. Multiple voltage ranges, but I don't really need steps more than 1mV. But nice to have :).
1. True remote sensing.

I decided to split task into two. The first one is choosing power supply architecture. The second one is choosing components.

I'm more or less set on parts to be used. And, I think, it's less important than architecture. I need you help there. May be gurus of analog engineering want to share their designs and ideas? :)

Currently the best I can do is the following. I didn't figure out how to make a beefy, precise and fast power supply. So, I decided that I only need precision and performance for low-power and low-voltage ranges. So, I'd make a low-power channel that would consist of simple emmitter follower with a fast bjt (like 2n2222 or something, point is it has much lower collector-base capacitance than, say, 2STA1943).

To reduce recovery time, I'd add a current sink at the output. This will create an offset current as it goes through the shunt. So, this needs to be taken into account. But even if we resolve this problem, I still haven't figured out how to switch fast between CC and CV mode. I'm currently looking at opamps that have ability to limit output swing, but there are not many around. So, clearly, not the best design.

Now to start answering from that thread:

The TIP2955/TIP3055 are more like normal slow BJTs. For fast response I would consider 2SC5200 or similar transistors made for high end audio. For a smaller version D44H / D45H could be an option.

Thanks for suggestion, I'll try 2SC5200. As of D44H and TIP2955, I think the performance of these part varies a lot from manufacturer to manufacturer. My measurements show that original parts from Toshiba (okay, I have only one sample of 2ta1943) was much slower than 2STA1943 (ST). As of D44H, I think showed similar performance to TIP2955/TIP3055. I can't find curves for D44H8 the moment, but I'll do measurement again and share them here. Here is methodology I'm using for measurements: https://www.eevblog.com/forum/beginners/evaluating-bjts-for-a-linear-psu-am-i-doing-it-right/msg2950816/#msg2950816 .

As for other questions, I'll answer them next time. It's a lot of topics yet to discuss, and I want to start with something. So, here is the thread for that.

I'm not investigating push-pull stages, but I have zero previous experience with them, so it will take time how to figure out how they work, how to optimize, say, biasing, and how to sense current.
 

Online exe

  • Supporter
  • ****
  • Posts: 1805
  • Country: nl
  • self-educated hobbyist
Re: how to design fast bench supply with CC and CV?
« Reply #1 on: April 07, 2020, 08:53:33 pm »
Ah, I just got to know that 2SC5200 is a brother of 2SA1943. In this case I'm fairly confident that modern TIP3055/2955 from onsemi and ST are no worse. In fact, it can be just the same die. I'll share curves later.
 

Offline Smokey

  • Super Contributor
  • ***
  • Posts: 1629
  • Country: us
Re: how to design fast bench supply with CC and CV?
« Reply #2 on: April 07, 2020, 10:52:30 pm »
Since I'm stuck at home in this stupid coronavirus quarantine, I've been working on my Mars rover for Mars again.  It sounds like your power supply would be a perfect addition to my project and it sounds like you are SUPER excited about electronics just like me!!!!!!!!!!111111

https://www.eevblog.com/forum/projects/i-would-like-to-build-a-mars-rover-for-mars-please-help/

Lets work together!!! 

Since the coronavirus will be over by Easter, that's my new deadline for my Mars Rover for Mars.  It sounds like your project should be done by then too!!!!!! I'm so Excited!!!
 
The following users thanked this post: exe

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11901
  • Country: us
  • DavidH
Re: how to design fast bench supply with CC and CV?
« Reply #3 on: April 08, 2020, 12:07:27 am »
For fast response and low output capacitance, take a look at how class-AB audio power amplifier output stages are designed.  And there is a very small step from a class-AB output stage to 2 or 4 quadrant operation.

I do not necessarily recommend using them but Linear Technology, now Analog Devices, makes some power output stage controllers which might be useful.  Check out their datasheets and application notes for some ideas:

https://www.analog.com/en/products/lt1166.html
https://www.analog.com/en/products/lt1970.html
https://www.analog.com/en/products/lt1970a.html

Keep the voltage and current control loops as simple as possible, which almost always means one operational amplifier for each.  Do *not* cascade amplifier stages within the control loops unless it is absolutely necessary, and it almost never is.  Level shift control and read-back signals to your digital common outside of any control loops.

As a corollary, this also means that you will likely get maximum overall performance with relatively low bandwidth operational amplifiers because faster ones will just require external compensation for stability anyway.  And external compensation creates problems with integrator windup unless special steps are taken.
 
The following users thanked this post: exe, Neomys Sapiens

Offline H713

  • Regular Contributor
  • *
  • Posts: 103
  • Country: us
Re: how to design fast bench supply with CC and CV?
« Reply #4 on: April 08, 2020, 04:53:22 am »
Highly doubtful that the TIP3055 and 2SC5200 have the same die- the voltage ratings, power ratings and speeds are all different.

The 2SA1943 and 2SC5200 are extremely high performance transistors and are probably completely overkill. Also commonly counterfeit- so buy from a reputable supplier. Other overkill (and extremely high performance) transistors include the MJL3281 and MJL1302. These are ON-Semi reproductions off the 2SA1302 and 2SC3281. The original Toshiba parts have been out of production since the late 1990s. If you find "Toshiba" parts for sale, be careful. The TIP3055s and TIP2955s that I have tested were dog slow.

The MJ21194 is also a pretty nice choice. It is also available in plastic packages.


 

Online Radiohead

  • Contributor
  • Posts: 31
  • Country: nl
Re: how to design fast bench supply with CC and CV?
« Reply #5 on: April 08, 2020, 06:42:57 am »
Hello exe,

Your requirements look a lot like an Source Measurement Unit (SMU), which often have very little output capacitance and very high programmable output precision. I haven't found any articles of someone designing a true SMU as a hobby project, but I have a few interesting resources you might want to read:
https://poormanssmu.wordpress.com/research/

The lt1970A is also an interesting chip. The current limiting isn't precise, but the application example in the datasheet to build a four quadrant power supply with the LT1970A is still worth a look.

Of course Blackdog has an (unfinished) design of a high-speed, low capacitance and low current power supply project: (Warning: it's in Dutch!)
https://www.circuitsonline.net/forum/view/135399
 
The following users thanked this post: Neomys Sapiens

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 7890
  • Country: de
Re: how to design fast bench supply with CC and CV?
« Reply #6 on: April 08, 2020, 09:57:20 am »
A first point to decide for the design is the type of output stage: One has to decide between a current controlled output stage
and a voltage controlling output stage with a low output impedance. The later may split the problem to the amplifier / control part and the output stage. Remember that even just a emitter follower can oscillate with capacitive load. 
With the low voltage aimed for both variants may be feasible.

For the transistors the TIP3055 and similar may be fast from some sources, but they usually should not be fast.
The main limiting factors for the control loop should be parasitic inductance (e.g. of the shunt resistors, connections to the output) and the
speed of the power transistors. At low current even the faster power transistors can become relatively slow. So one may have to operate them with some minimum current. It is tricky enough to cope with a variable load impedance.
 
The following users thanked this post: WattsThat

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11901
  • Country: us
  • DavidH
Re: how to design fast bench supply with CC and CV?
« Reply #7 on: April 08, 2020, 05:40:49 pm »
The 2SA1943 and 2SC5200 are extremely high performance transistors and are probably completely overkill. Also commonly counterfeit- so buy from a reputable supplier. Other overkill (and extremely high performance) transistors include the MJL3281 and MJL1302. These are ON-Semi reproductions off the 2SA1302 and 2SC3281. The original Toshiba parts have been out of production since the late 1990s. If you find "Toshiba" parts for sale, be careful. The TIP3055s and TIP2955s that I have tested were dog slow.

Ring and perforated emitter transistors are common in high performance regulators.  They used to show up in linear point of load regulators for Pentium processors.  Old timers may recognize the D44H11/D45H11 series of transistors which would be a good choice except for needing several in parallel for this application to handle the worst case power dissappation.  The MJL4302AG/MJL4281AG would be better choices both for economy and a simpler design and there are lots of similar parts in the same family which will work. 

Besides being fast, ring emitter transistors also have less hfe and ft droop at high current.  All three make them popular for high performance audio power amplifiers which provide enough demand to make them inexpensive.

If I used power MOSFETs for the output transistors, then I might use ring emitter transistors to provide gate drive.
 

Offline Vovk_Z

  • Frequent Contributor
  • **
  • Posts: 443
  • Country: ua
Re: how to design fast bench supply with CC and CV?
« Reply #8 on: April 08, 2020, 08:52:50 pm »
Isn't it easy to achieve bipolar <=3 A output with something like TDA7293 at output? (It needs precision op amp to control it, of cause).
« Last Edit: April 09, 2020, 08:29:51 pm by Vovk_Z »
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1009
  • Country: ca
Re: how to design fast bench supply with CC and CV?
« Reply #9 on: April 09, 2020, 03:16:46 am »
I've been out of bipolar power transistors for a while; the MJL4302AG/MJL4281AG Ft are quite impressive, especially considering the breakdown voltages.  I remember that in olden times 2N3055 were single diffused and slow.  Sometime along the way they became the epitaxial 2N3055E, about 3X better Ft, and more prone to oscillation in unmodified circuits.  The above devices appear to be an order of magnitude faster.

The D44H11/D45H11 also rings a bell.  I have a Tek FG503 that I got for almost nothing because there was no AC output but a very large DC offset.  I expected it to have a 2N3866/2N5109 combo but it had something from the D40 series, now apparently unobtainable.  I kludged in a 2N2219 & 2N2905 soldered on a copper heat spreader and let it go at that until I found something with more muscle.  I was perusing Digikey and found the 2SC6144SG/2SA2222SG.  They were cheap, had Ft and hfe to burn so I bought some but haven't tried them out yet.  They use something called the "MBIT" process.

I don't know much about ring or perforated emitter devices but apparently it's a transistor geometry used in the VHF 2N3866?  Any more info on what's being done in these devices to pull this off?
 

Offline profdc9

  • Regular Contributor
  • *
  • Posts: 195
  • Country: us
Re: how to design fast bench supply with CC and CV?
« Reply #10 on: April 09, 2020, 05:59:54 am »
I designed a supply based on a popular op-amp design.  It also supports external programmability.

http://www.github.com/profdc9/LinearPS

The output capacitance is on the order of 10-47 uF.
 
The following users thanked this post: exe

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 14584
  • Country: gb
  • 0999
Re: how to design fast bench supply with CC and CV?
« Reply #11 on: April 09, 2020, 12:12:22 pm »
Switching between CC and CV modes is tricky/impossible to do without glitches, resulting in voltage spikes or current surges. Most power supplies will do one or the other, with the latter being more common. You might want to add a switch to allow the power supply to be changed between optimal voltage and current regulation modes, so it will work just as well for CC loads such as LEDs, as the usual CV loads like microcontrollers.
 

Offline mzzj

  • Frequent Contributor
  • **
  • Posts: 985
  • Country: fi
Re: how to design fast bench supply with CC and CV?
« Reply #12 on: April 09, 2020, 12:32:06 pm »
For fast response and low output capacitance, take a look at how class-AB audio power amplifier output stages are designed.  And there is a very small step from a class-AB output stage to 2 or 4 quadrant operation.


HP 6632A or 6632B are pretty much like that.
Actually quite heavily biased, IIRC around 2A, possibly to lower output impedance.
Separate shunts for low and high range, 20mA range really accurate for a PSU.
Output capacitance selector switch, normal mode and fast more. IIRC also controls the feedback loop.

6632B’s are common on second hand market and often real bargain for what you gets. Going rate for these has been around 150usd and this was over 2 grand per piece from Keysight just a few years ago.
 
The following users thanked this post: exe

Online exe

  • Supporter
  • ****
  • Posts: 1805
  • Country: nl
  • self-educated hobbyist
Re: how to design fast bench supply with CC and CV?
« Reply #13 on: April 09, 2020, 01:00:43 pm »
Hello exe,

Your requirements look a lot like an Source Measurement Unit (SMU), which often have very little output capacitance and very high programmable output precision. I haven't found any articles of someone designing a true SMU as a hobby project, but I have a few interesting resources you might want to read:
https://poormanssmu.wordpress.com/research/

Oh, man, is that your blog? I read it all, too bad it somewhat abruptly stopped, so I'm trying to carry on :). Thanks for all the info! I'm now experimenting with push-pull too, have promising results already.

Switching between CC and CV modes is tricky/impossible to do without glitches

Yep, figured that the hard way :(. Blackdog's power supply offers an elegant solution with clamping diodes. See schematic here, for example: https://www.eevblog.com/forum/beginners/how-does-blackdog_s-psu-work/ . Very nice unit, really hard to beat.

Another approach: "anti-soar clamp" in AoE x-chapters. Afaik, essentially the same concept, but using an opamp, bjt, and a few passives to do the same. I'll post a picture of it.

6632B’s are common on second hand market and often real bargain for what you gets. Going rate for these has been around 150usd and this was over 2 grand per piece from Keysight just a few years ago.

Yeah, but I want to build mine :).


Sorry folks, can't reply to all of you, you gave me readings for weeks ahead. Nice to get to know new parts. I thought I knew all the bjts on the market, but I was wrong :).
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 14584
  • Country: gb
  • 0999
Re: how to design fast bench supply with CC and CV?
« Reply #14 on: April 09, 2020, 02:43:08 pm »
Switching between CC and CV modes is tricky/impossible to do without glitches

Yep, figured that the hard way :(. Blackdog's power supply offers an elegant solution with clamping diodes. See schematic here, for example: https://www.eevblog.com/forum/beginners/how-does-blackdog_s-psu-work/ . Very nice unit, really hard to beat.

Another approach: "anti-soar clamp" in AoE x-chapters. Afaik, essentially the same concept, but using an opamp, bjt, and a few passives to do the same. I'll post a picture of it.
The problem with the diodes is the voltage is fixed. The anti-soar clamp sounds like a good idea. I haven't seen the circuit but I can imagine what it looks like.

The output capacitors are responsible for the huge current circuit but it'll still happen without them. Here's a quick design I've done in LTSpice to simulate the transient response. It's not supposed to be a final design and isn't very good.
[attachimg=1]
[attachimg=2]
« Last Edit: April 09, 2020, 06:32:26 pm by Zero999 »
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 7890
  • Country: de
Re: how to design fast bench supply with CC and CV?
« Reply #15 on: April 09, 2020, 04:01:51 pm »
For the current spikes with an emitter follower configuration there are 2 problems:
One is that the amplifier for the voltage control has a limited slew rate. The other problem is wind of the current regulator. So in the case of a sudden short there has to be some additional fast limit. Often this limit may be fixed.

Bockdogs circuit has a limited anti-windup with the diodes in FB at the OPs. This is more like windup limiting, not fully preventing windup. Still this is a good idea if it works (depends on the configuration and needed control voltages).

One step closer anti-windup can be done with extra transistor(s) to couple the 2 regulators. There is a slight disadvantage the leakage can effect the precision. Attached is a simulation file for such a anti-windup for the voltage regulator. The simulation shows the CC to CV transition were the anti-windup part helps.  Beside the anti windup the circuit also has a nice simple constant current sink.
 

Online exe

  • Supporter
  • ****
  • Posts: 1805
  • Country: nl
  • self-educated hobbyist
Re: how to design fast bench supply with CC and CV?
« Reply #16 on: April 09, 2020, 05:57:41 pm »
Bockdogs circuit has a limited anti-windup with the diodes in FB at the OPs. This is more like windup limiting, not fully preventing windup.

Could you please elaborate this? As far as my understanding goes, it works almost perfectly: it clamps both opamps just above the current output voltage (+1-2V above the output voltage). For CV this is done by raising voltage at output feedback pin. For CC this is done by clamping voltage to the highest side of high-side shunt. CV/CC switching works really well. Both opamps still track the output voltage.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 7890
  • Country: de
Re: how to design fast bench supply with CC and CV?
« Reply #17 on: April 09, 2020, 07:00:15 pm »
In the circuit from Backdog the diodes limit the OPs output voltage to about +0.6 V, maybe +1 V for the current limit.
The active channel is at some -1 to -2 V. So on the transition the OPs have to slew from some +0.6 V to some -1 ot -2 V.  One has little room to reduce the range, as there has to be room for the regulator to operate.

The additional transistor Q6 in Blockdogs circuit adds an additional fast current limit, though at a fixed level. Ideally one would not need this, but it does not hurt to have this extra safety net.

With the transistor the step for the regulator to overcome is at some 1.1 V. Slightly smaller, though not much.
The other point that can help is to have some of the feedback from behind the diodes. However this can also sometimes lead to the current limit to engage a little earlier on voltage transients, when just at the edge. So this is a little 2 sided.

Both versions are quite a bit faster to react than the simple circuit where on transition between modes the regulator has to slew quite a bit, like some 5 v set by the supply.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 635
  • Country: it
Re: how to design fast bench supply with CC and CV?
« Reply #18 on: April 09, 2020, 07:28:29 pm »
Since I'm stuck at home in this stupid coronavirus quarantine, I've been working on my Mars rover for Mars again.  It sounds like your power supply would be a perfect addition to my project and it sounds like you are SUPER excited about electronics just like me!!!!!!!!!!111111

https://www.eevblog.com/forum/projects/i-would-like-to-build-a-mars-rover-for-mars-please-help/

Lets work together!!! 

Since the coronavirus will be over by Easter, that's my new deadline for my Mars Rover for Mars.  It sounds like your project should be done by then too!!!!!! I'm so Excited!!!

Easter? Yes... may be 2021 or 2022...
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11901
  • Country: us
  • DavidH
Re: how to design fast bench supply with CC and CV?
« Reply #19 on: April 10, 2020, 12:05:06 pm »
Isn't it easy to achieve bipolar <=3 A output with something like TDA7293 at output? (It needs precision op amp to control it, of cause).

An integrated audio amplifier could be used but they are commonly decompensated for a minimum gain limiting their flexibility or performance.  I am sure it has been done though.

I've been out of bipolar power transistors for a while; the MJL4302AG/MJL4281AG Ft are quite impressive, especially considering the breakdown voltages.

What really matters are the secondary breakdown and power ratings.

Quote
I remember that in olden times 2N3055 were single diffused and slow.  Sometime along the way they became the epitaxial 2N3055E, about 3X better Ft, and more prone to oscillation in unmodified circuits.  The above devices appear to be an order of magnitude faster.

The original 2N3055s also had outstanding secondary breakdown characteristics making them very tough.  Does anybody have a link to the original 2N3055 datasheet with some characteristic curves?

Quote
The D44H11/D45H11 also rings a bell.  I have a Tek FG503 that I got for almost nothing because there was no AC output but a very large DC offset.  I expected it to have a 2N3866/2N5109 combo but it had something from the D40 series, now apparently unobtainable.  I kludged in a 2N2219 & 2N2905 soldered on a copper heat spreader and let it go at that until I found something with more muscle.  I was perusing Digikey and found the 2SC6144SG/2SA2222SG.  They were cheap, had Ft and hfe to burn so I bought some but haven't tried them out yet.  They use something called the "MBIT" process.

I don't know much about ring or perforated emitter devices but apparently it's a transistor geometry used in the VHF 2N3866?  Any more info on what's being done in these devices to pull this off?

The way I understand it, the D44H11/D45H11 and similar ring or perforated emitter transistors are distinct from RF transistors which commonly have multiple ballasted emitters.  From its description, MBIT seems like another variation of ring or perforated emitter.

Switching between CC and CV modes is tricky/impossible to do without glitches, resulting in voltage spikes or current surges. Most power supplies will do one or the other, with the latter being more common. You might want to add a switch to allow the power supply to be changed between optimal voltage and current regulation modes, so it will work just as well for CC loads such as LEDs, as the usual CV loads like microcontrollers.

Good frequency compensation and provisions to remove excessive charge from the output devices go a long ways toward minimizing glitches during mode changes.  That plus clamping the error amplifiers to prevent integrator windup can solve this problem.

The external compensation connections on some operational amplifiers, like the LM301A, can be used as clamp inputs.
« Last Edit: April 10, 2020, 12:13:27 pm by David Hess »
 
The following users thanked this post: exe

Offline duak

  • Super Contributor
  • ***
  • Posts: 1009
  • Country: ca
Re: how to design fast bench supply with CC and CV?
« Reply #20 on: April 10, 2020, 04:35:38 pm »
Here's a link to the 1966 RCA Transistor Manual:  http://www.rsp-italy.it/Electronics/Databooks/RCA/_contents/RCA%20Transistor%20Manual%20SC12%201966.pdf

The 2N3055 specs start on p 228.  I don't see an SOA curve however, some other power devices have a DC SOA only graph, but no pulse ratings.  I suppose it took a few more years before more complete SOA graphs were produced.  Motorola certainly had more complete graphs in the early 70s.  Does anyone remember the big, brown Motorola semiconductor data book?

Here's a link to a later RCA datasheet:  https://www.silicon-ark.co.uk/datasheets/2n3055-datasheet-rca.pdf but no SOA graph.  It appears to be the older, slower part.  I see it was sent from GE Solid State after RCA was folded back into GE.

https://en.wikipedia.org/wiki/2N3055 has some history and info but no data.

BTW, I think RCA were the first to come out with low threshold voltage power FETs in '86 or so.  I thought they might work well as pass transistors in a bench supply as they were touted as having Positive Temperature Coefficients and freedom from secondary breakdown.  In the end, I gave up and went back to bipolar darlingtons, as the LVFETs clearly suffered from Electro-thermal Instability.  If memory serves, they couldn't always handle a sudden short circuit load.
« Last Edit: April 10, 2020, 07:04:22 pm by duak »
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 7890
  • Country: de
Re: how to design fast bench supply with CC and CV?
« Reply #21 on: April 10, 2020, 04:53:59 pm »

BTW, I think RCA were the first to come out with low threshold voltage power FETs in '86 or so.  I thought they might work well as pass transistors in a bench supply as they were touted as having Positive Temperature Coefficients and freedom from secondary breakdown.  In the end, I gave up and went back to bipolar darlingtons, as the LVFETs clearly suffered from Electro-thermal Instability.  If memory serves, they couldn't always handle a sudden short circuit load.

The problem with using MOSFETs and a sudden short is especially there with a source follower configuration. In this case the current goes up very fast (with MOSFETs the current goes up even faster than linear with the gate voltage). In addition there is some danger to even exceed the maximum gate voltage.  With BJTs on contrast, the current gain goes down once on gets very high in current. So even if the base drive is not going down fast enough the current does not peak so bad.

The early MOSFETs had a rather high R_on and relatively large die. There SOA was reasonable good. Later generations got smaller dies and later a very limited SOA, especially for the lower voltage ones. In additions there are lots of unreliable data-sheets around ignoring the thermal instability SOA limit. So even if there is a DC SOA curve one has to check if one can depend on it.

The old, slow 2N3055 version is still available as 2N3055H.  Newer chips tend to get smaller and smaller dies for cost reasons.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 14584
  • Country: gb
  • 0999
Re: how to design fast bench supply with CC and CV?
« Reply #22 on: April 10, 2020, 05:33:27 pm »
Adding a diode to limit the slew rate is a good idea. Ideally the stage needs to be unity gain, otherwise it ends up amplifying the diode forward voltage. If needed a gain stage added to the reference, but it would require another op-amp. I thought I'd try it with my circuit, leaving it with a gain of 10, just to see what would happen. In this case it needs to go on to the voltage, rather than current amplifier, because my circuit is a current source first and a voltage source second. Obviously it makes no difference to the current surge, but the voltage spike is reduced, as expected.

Yes, something to clamp Q2's base very quickly, could be used to reduce the current surge. Ideally it should be variable, but the sense resistor's value would need to be switchable anyway, to get the range desired by the original poster. I have an idea, but will post it later.
[attachimg=1]
[attachimg=2]
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1009
  • Country: ca
Re: how to design fast bench supply with CC and CV?
« Reply #23 on: April 10, 2020, 08:39:29 pm »
Kleinstein points out the reasons why MOSFET pass transistors may not be the best without some extra considerations.  For my supply, I had considered a fast current limiter that reduced gate drive like that in the 723 regulator, ie., an NPN transistor driven from a current sense resistor in the pass transistor circuit.  (Q3 in the attached file)  But, the board was done and it was easier to go with a Darlington.

I'd suggest this type of fast current limiter in addition to the current control loop to protect the output devices in case of an accidental short circuit.
 
The following users thanked this post: exe

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 14584
  • Country: gb
  • 0999
Re: how to design fast bench supply with CC and CV?
« Reply #24 on: April 10, 2020, 10:36:21 pm »
Here's my shot at the fast current limit. The idea is to clamp the voltage on Q2's base to a low level, when the current limit is drastically exceeded. The main current limiting amplifier U2 starts to cut-off Q2's base drive, when the voltage across R1 exceeds one tenth of the current limit reference voltage: i.e V2/10. The fast current limiter is set at around 50% higher, than the main current limit, so V2/6.667. U3 is a non-inverting summing amplifier. It has two inputs, the output voltage of the power supply and the current limit reference voltage V2. R10 and R11 divide V2 by 62/3, which is added onto the output voltage. The output is connected to C1, via R7, which is needed for stability.

VOUT+VC1 = VOUT+ V2/6.667

So when V2 = 1
VOUT+VC1 = VOUT + V2/6.667 = 0.15V

When the output is shorted, the current through R5 will drastically increase, pushing the emitter and thus base voltage up. When the base voltage exceeds that on C2 by a diode drop, D3 will clamp the base voltage, pulling it down into C1, cutting off the drive. C1 will charge a bit, causing the current to slowly rise, but U2 starts to react before the current gets too high.

The advantage of this circuit is the higher current limit scales with the current setting. I will work better, if a greater voltage drop is allowed across R5, which will still need to be switched to different values, to meet the OP's specification.

I've changed the voltage amplifier to unity gain. The reference will have to have a separate amplifier, if some gain is required.

[attachimg=1]
[attachimg=2]
« Last Edit: April 10, 2020, 10:39:10 pm by Zero999 »
 
The following users thanked this post: exe


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf