Author Topic: lab power supply  (Read 13119 times)

0 Members and 2 Guests are viewing this topic.

Offline tooki

  • Super Contributor
  • ***
  • Posts: 13157
  • Country: ch
Re: lab power supply
« Reply #25 on: December 15, 2019, 12:47:22 pm »
It never came to fruition, but instead morphed into the still unfinished μSupply portable PSU project.

It has always been the uSupply portable USB PSU project.
Well then what’s this, then? It makes no mention of portable or USB:

https://youtu.be/CIGjActDeoM

It wasn’t until part 11 that you expressly mention wanting to make it portable:
https://youtu.be/xa9Lyb45oJM


And not until part 14 did you name it uSupply.

So while it might have always been envisioned in your mind as the uSupply, anyone watching the PSU series from the beginning (as I did) would see it as morphing from a more traditional PSU design into the uSupply. And for sure, for someone looking for how the “lab PSU” series continues, it’s not self-evident from the video titles that the uSupply is the continuation. Looking for “PSU part 14” comes up dry.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: lab power supply
« Reply #26 on: December 15, 2019, 01:18:18 pm »
Below is the lab power supply schematic that I would emulate from National Semiconductor linear brief 28.  Instead of using expensive LM395 integrated power transistors, the power pass element could be replaced with suitable integrated regulators or bare transistors or some combination of the two.

that's a nice circuit (unlike the other ones >:D) but it looks like LM308 is no longer available and LM101 is hard to find and very expensive, although there are cheap LM301 in sot8 package from onsemi.

replacing those opamps with modern parts would require a different compensation network and careful selection of parts complying with the same input range (and other features) of the original ones or mirroring the design (regarding polarity) to rather use the so-called single power supply opamps that usually include the negative rail as input range.

The LM101 is the military temperature range version of the LM301 and the LM301 can be replaced with the LM301A which is readily available.  (1) The LM308 as used here may be replaced with practically any device although the LT1008 from Linear Technology is an improved direct replacement.  The LT1012 and LT1097 are also improved replacements which include internal compensation.

The LM301A would be very difficult to replace in this circuit because:

1. Its input common mode range includes the positive supply making it especially suitable for a high side current sense amplifier or in this case, a high side current error amplifier.  There are many modern rail-to-rail input parts and several older JFET input parts which meet this requirement however ...
2. Its external compensation pin can is used here to clamp its output to improve overload recovery time.  The list of modern operational amplifiers which can support this mode of operation is very short.  The aforementioned LT1008, LT1012, and LT1097 are the only common modern possibilities I am aware of.  Analog Devices makes some parts which might be able to do this but since they do not publish schematics, there is no way to know without reverse engineering them.

I mostly listed this example circuit as something to study because it includes many lessons in practical design like low output capacitance, active pull-down, and direct high side current sensing.  There are ways to externally clamp an internally compensated operational amplifier however they get complicated very quickly so it is very rare.

(1) The LM301A is an LM301 with improved input bias current over temperature.  The LM301 was only produced for a very short time before being replaced with the A version.
« Last Edit: December 15, 2019, 07:04:20 pm by David Hess »
 
The following users thanked this post: not1xor1

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #27 on: December 15, 2019, 02:20:06 pm »
progress so far well used the 2 tanilum caps i think they are called and works great so far very stabe of course no load no heating up but anyway its cool so far
« Last Edit: December 15, 2019, 02:23:42 pm by glinjik »
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline Jwillis

  • Super Contributor
  • ***
  • Posts: 1725
  • Country: ca
Re: lab power supply
« Reply #28 on: December 17, 2019, 05:53:08 am »
The yellow caps you showed don't look like tantalums . They look more like monolithic ceramics . I could be wrong . If they are indeed tantalums there should be a marking on them that shows a polarity.Monolithic Ceramics will just have a number code. If they are monolithic ceramics it won't be a problem using them in your circuit.
 
The following users thanked this post: glinjik

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #29 on: December 17, 2019, 06:19:25 am »
thank you for clearing that up you are correct not tantalums that bit of the circuit worked fine yesterday untill i put a 5w resister load to see what amps im getting and my 4 1000mf caps went up in a kind of smoke cloud lol i just dont have a big enough cap for this at the moment oh i think the two 2n3055 transistors blew as well cant complain at least i only fryed a few components and not myself this time at least the fuse didnt blow  :palm:
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline Jwillis

  • Super Contributor
  • ***
  • Posts: 1725
  • Country: ca
Re: lab power supply
« Reply #30 on: December 17, 2019, 08:03:36 am »
What is the voltage rating of your electrolytic capacitors and The voltage of your transformer secondary.
If you measure the AC voltage of the secondary you take that measurement and multiply by 1.4 to get the rectified voltage With filter caps . So for example if your transformer is putting out 12VAC . 12 X 1.4 = 16.8 VDC after rectification and filter caps . 16V capacitors will be right on the edge of destruction at that voltage. Try to double the capacitor voltage rating over the rectified voltage.  Don't just take the voltage of the rectifier and try to match the capacitors to that voltage because as soon as you put a filter capacitor in circuit that voltage will rise potentially exceeding the rating of the filter caps your installing . If you can salvage old televisions there are usually some big high voltage ones in those . For your simple power supply a couple 220uF or 470uF 200 - 400 V capacitors would be more than adequate.
If your going to parallel transistors you need to bias them  to prevent load hogging and thermal runaway. Usually this can be done simply by putting a resistor at the base and a very low resistor ( 0.1 ohms is plenty) with a high wattage (5W) at the emitter of each transistor.  Even in a single transistor circuit, biasing will keep that transistor from exceeding its rated current capability if the power supply is over loaded by shorting . 
 
The following users thanked this post: glinjik

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #31 on: December 17, 2019, 09:50:08 am »
hi thank you for your reply the caps were 4 x 1000mf 50v transformer is 20.02 so with you calculation 28.02v should have been ok i was trying to get as close to the capacitance as possible had in series i thought the capacitance was really important ie got loads of big voltage caps but with low capacitance 22omf 400v, 150mf 450v, 100mf 450v etc i the emmiters have 0.5 \$\Omega\$5w ceramic resistors, should i put a big chunky diode at the output to prevent shorting? thanks for your reply hope you dont mind my noob questions it was kinda fun to watch those caps pop though lol
« Last Edit: December 17, 2019, 09:53:11 am by glinjik »
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #32 on: December 17, 2019, 10:05:54 am »
this is what ive done minus the dead caps :palm:
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4004
  • Country: nl
Re: lab power supply
« Reply #33 on: December 17, 2019, 11:01:27 am »
You can also have a look at:
https://hackaday.io/search?term=powersupply
for inspiration.
Some of the projects there looks suspiciously like Dave's old project, such as:
https://hackaday.io/project/23249-digital-battery-operated-powersupply

Also, think about adding a switched pre-regulator.
You can build one by modifying a cheap SMPS module from Ali / Ebay / China.
The basic way to do it is to add a PNP transistor to your circuit with the emitter to the power input, and the base connected to the output voltage.
(With added resistor for current limiting and 1 or 2 diodes in series to increase the voltage drop.
The collector of the PNP transistor then gets connected to the voltage sens pin of your SMPS board.

Such SMPS modules are cheaper then the big aluminimum heat sink it replaces.

Stuff like this is also so easy to find on the 'net.
How about a picture search for schematics?
  https://duckduckgo.com/?q=eevblog+power+supply+schematic&t=h_&iar=images&iax=images&ia=images


« Last Edit: December 17, 2019, 11:07:24 am by Doctorandus_P »
 
The following users thanked this post: glinjik

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #34 on: December 17, 2019, 11:09:29 am »
You can also have a look at:
https://hackaday.io/search?term=powersupply
for inspiration.
Some of the projects there looks suspiciously like Dave's old project, such as:
https://hackaday.io/project/23249-digital-battery-operated-powersupply

Also, think about adding a switched pre-regulator.
You can build one by modifying a cheap SMPS module from Ali / Ebay / China.
The basic way to do it is to add a PNP transistor to your circuit with the emitter to the power input, and the base connected to the output voltage.
(With added resistor for current limiting and 1 or 2 diodes in series to increase the voltage drop.
The collector of the PNP transistor then gets connected to the voltage sens pin of your SMPS board.

Such SMPS modules are cheaper then the big aluminimum heat sink it replaces.



hi thank you for your reply but i dont want to put modules in, im learning an awful lot from doing this so to buy modules would be pointless if you know what i mean but i did see somewhere about adding pnp transistors ill have a look at hackaday in a bit but thank you
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39026
  • Country: au
    • EEVblog
Re: lab power supply
« Reply #35 on: December 17, 2019, 11:10:17 am »
So while it might have always been envisioned in your mind as the uSupply, anyone watching the PSU series from the beginning (as I did) would see it as morphing from a more traditional PSU design into the uSupply. And for sure, for someone looking for how the “lab PSU” series continues, it’s not self-evident from the video titles that the uSupply is the continuation. Looking for “PSU part 14” comes up dry.

These prototypes must be a figment of my imagination.
From left to right, dated 7/8/09, 3/5/10, 31/8/10
I showed that first prototype to Leo Simpson at Silicon Chip when he visited Altium in 2009 and dropped by my cubicle to say hi. Almost got it published in SC, but I just kept fiddling with it.

 
The following users thanked this post: glinjik

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39026
  • Country: au
    • EEVblog
Re: lab power supply
« Reply #36 on: December 17, 2019, 11:12:47 am »
Would You mind to elaborate on the issues of the LT3081.

Sorry my memory of the exact issue is rusty. Something to do with back diode protection killing the chip. I wasn't the one that found it.
I think there was talk on the forum at the time whether LT would recall the chip.
 
The following users thanked this post: glinjik

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2764
  • Country: us
Re: lab power supply
« Reply #37 on: December 17, 2019, 01:16:41 pm »
Would You mind to elaborate on the issues of the LT3081.

Sorry my memory of the exact issue is rusty. Something to do with back diode protection killing the chip. I wasn't the one that found it.
I think there was talk on the forum at the time whether LT would recall the chip.

LT3080 wierdness - dave's power supply (eev224) gone mad
 
The following users thanked this post: glinjik

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #38 on: December 17, 2019, 04:57:03 pm »
update and question so after advice from
What is the voltage rating of your electrolytic capacitors and The voltage of your transformer secondary.
If you measure the AC voltage of the secondary you take that measurement and multiply by 1.4 to get the rectified voltage With filter caps . So for example if your transformer is putting out 12VAC . 12 X 1.4 = 16.8 VDC after rectification and filter caps . 16V capacitors will be right on the edge of destruction at that voltage. Try to double the capacitor voltage rating over the rectified voltage.  Don't just take the voltage of the rectifier and try to match the capacitors to that voltage because as soon as you put a filter capacitor in circuit that voltage will rise potentially exceeding the rating of the filter caps your installing . If you can salvage old televisions there are usually some big high voltage ones in those . For your simple power supply a couple 220uF or 470uF 200 - 400 V capacitors would be more than adequate.
If your going to parallel transistors you need to bias them  to prevent load hogging and thermal runaway. Usually this can be done simply by putting a resistor at the base and a very low resistor ( 0.1 ohms is plenty) with a high wattage (5W) at the emitter of each transistor.  Even in a single transistor circuit, biasing will keep that transistor from exceeding its rated current capability if the power supply is over loaded by shorting . 
i installed a 450v 150mf cap and that seems ok my little melt down did destroy both 2n3055's what im thinking now is no load i get a steady22.02v under load it drops to 19.59v its only a 12v led light as a load what im wondering is this can it be made more stabe? the variable resistor pot at P3 does nothing for caibration of volt meter im using a digi one im guesing that pot was designed to calibrate a analogue version also if i connect volt meter as per diagram it only shows 9 volts? maybe volt meter should be connected direct to output? can i get rid of c8,c9.d10.r8p3 amnd c10 and put a big output filter cap there instead i know a lot of questions and you all prob think im a thick noob but im loving this project :)
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #39 on: December 17, 2019, 05:01:04 pm »
also want to incorporate this fan control
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2764
  • Country: us
Re: lab power supply
« Reply #40 on: December 17, 2019, 07:11:27 pm »
Since you only want an adjustable supply with less than 1A, I suggest you go back to basics and get that working first.

If the digital volt meter you want to use covers the maximum output voltage the LM317 will be set to, you can just wire it directly to the outputs.

   891346-0


i installed a 450v 150mf cap and that seems ok my little melt down did destroy both 2n3055's what im thinking now is no load i get a steady22.02v under load it drops to 19.59v its only a 12v led light as a load what im wondering is this can it be made more stabe?

If the 150uF 450V cap is all you have on the input, it's way too small. 
You will have a large ripple on the input to the LM317 with it.
You need something in the range of 6800uF @ 50V to not have that voltage drop on the output.
What is the rating of the transformer secondary you are using?

Put the fan control you shown across the 'input' of the LM317.
« Last Edit: December 17, 2019, 07:29:51 pm by MarkF »
 
The following users thanked this post: glinjik

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #41 on: December 17, 2019, 07:41:04 pm »
Since you only want an adjustable supply with less than 1A, I suggest you go back to basics and get that working first.

i have it working  :) ijust want to inprove it the fan control does work its on a seperate board at the moment if i have space i will put it on main board also want to add short circuit protection with transistors not relay its coming together nicely th cap will have to wait untill my next trip to town maybe a few weeks unless i scrounge one someplace the transformer is i cant get any data on although it came from a drill battery charger i think i got about 3 amps out of it befor the caps blew up
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline rdl

  • Super Contributor
  • ***
  • Posts: 3667
  • Country: us
Re: lab power supply
« Reply #42 on: December 17, 2019, 09:01:57 pm »
What is the output voltage of the transformer secondary?
 

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #43 on: December 18, 2019, 05:44:49 am »
output voltage is 19.81vac and 26vdc after rectification and output from circuit 22.38vdc interesting i can get as low 3.4mv unloaded mv under load is mainly impossible to control need to put in a secondery volt control
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #44 on: December 18, 2019, 04:25:24 pm »
update to my beginner’s power supply all working as it should no problems yet although lm317 does not seem to be getting warm i added a small heatsink anyway as with the bridge rec just in case also added connectors for volt control. the 2n3055's and volt meter next want to change the current control to a potentiometer and also put that on a connector the output will also change to a clip connector I figured if I blow it up it will be east to remove the board from the case just using clips, next is fan control once that’s working I will add it to the schematic for any other beginners I’ve part done the schematic for any one’s interest
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline glinjikTopic starter

  • Regular Contributor
  • *
  • Posts: 200
  • Country: gb
Re: lab power supply
« Reply #45 on: December 19, 2019, 09:30:49 am »
after looking at various ways to control a fan rather than just having it running found a little 555 timer fan control uses a npn transistor and thermistor so I’ve incorporated it into the scheme but before I build it in was wondering if any of you enlightened chaps might cast your eyes over the scheme so far oh i re drew the 555 to reflect it in reality rather than schematically its easyer to follow for me
All the truth in the world adds up to one big lie (Bob Dylan)
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
Re: lab power supply
« Reply #46 on: December 19, 2019, 11:17:15 am »
Here are a few DIY circuits you can play with:

(Attachment Link)


Was going to start a new thread, but where's the fun in that when I can hijack this one.

The above PSU specifies garden variety parts. What if I wanted to feed in 12vdc for a output range of up to 9 volts. What is a good solution for that -5v reference using off the shelf parts?

Are those buck/boost IC's considered off the shelf?

« Last Edit: December 19, 2019, 11:21:35 am by Ed.Kloonk »
iratus parum formica
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2764
  • Country: us
Re: lab power supply
« Reply #47 on: December 19, 2019, 12:46:06 pm »
Here are a few DIY circuits you can play with:

(Attachment Link)


Was going to start a new thread, but where's the fun in that when I can hijack this one.

The above PSU specifies garden variety parts. What if I wanted to feed in 12vdc for a output range of up to 9 volts. What is a good solution for that -5v reference using off the shelf parts?

Are those buck/boost IC's considered off the shelf?

If you don't need it to go below 1.2V, I believe you could use a slightly different version with a "single supply" or "rail-to-rail" op-amp. 
I think I would try a LM358 with 0V and 12V for its supply.    Edit:  Or the TL071 shown on the other version.
I'm not so sure you need a negative supply rail.

892386-0
« Last Edit: December 19, 2019, 12:56:03 pm by MarkF »
 
The following users thanked this post: Ed.Kloonk

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
Re: lab power supply
« Reply #48 on: December 19, 2019, 02:21:12 pm »
Here are a few DIY circuits you can play with:

(Attachment Link)


Was going to start a new thread, but where's the fun in that when I can hijack this one.

The above PSU specifies garden variety parts. What if I wanted to feed in 12vdc for a output range of up to 9 volts. What is a good solution for that -5v reference using off the shelf parts?

Are those buck/boost IC's considered off the shelf?

If you don't need it to go below 1.2V, I believe you could use a slightly different version with a "single supply" or "rail-to-rail" op-amp. 
I think I would try a LM358 with 0V and 12V for its supply.    Edit:  Or the TL071 shown on the other version.
I'm not so sure you need a negative supply rail.

(Attachment Link)

Nice. Thanks.
iratus parum formica
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: lab power supply
« Reply #49 on: December 19, 2019, 05:53:02 pm »
Another way to make an adjustable current and voltage regulator without an operational amplifier is shown below.
 
The following users thanked this post: glinjik


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf