List of changes and additions: 1) Transformer changed from 24V to 20V to reduce stress on the Op-Amps and pass transistor. 2) The expected maximum output of the power supply was then reduced from 30VDC to 20V DC. This helped deal with the ripple problem. 3) Replaced the fake Op-Amps with LM741s. Later the voltage error amp was changed to a real TL081 as the increased voltage swing gives better ripple rejection. 4) Changed D7 (negative supply Zener) from 5.1V to 3.9V to further reduce stress on the Op-Amps. 5) Changed R12 from 56K to 27K to limit maximum output to 20V 6) Changed R18 from 56K to 80K to maximize range of control over current limit. 7) Changed R2 from 1/4W to 1/2W (you could even got to 1W). 8) increased filter capacitor from 3300uF to 9900uf and changed from troublesome Won Haung Lo brand to Nichicon. 9) Dramatically Increased size of heatsink from original design. 10) Added fine controls for both Voltage and Current adjustment. 11) installed polarized headers for the fan and to facilitate remote mounting of current limiting LED. 12) Installed terminal block to facilitate remote mounting of the pass transistor. And of course we added a nice LCD panel meter with V/A/P and put it all in a cabinet.
Wow! You actually made a really wonderful and serviceable power supply, from that rather poor kit. Kudos! Bravo! I think the look of it is quite handsome, as well. I am sure that it will be a joy, for years to come. Cheers!
It needs some kind of calibration scale for the current limit knob, so it can be set reasonably accurately without having to short the output to set the required current. Not that shorting them is a problem, it just makes it slightly easier and you can just look at the knob to see what current it is set to. Set assuming the fine knob is centered. A very interesting video series by the way, AliExpress should link to it!
Nice project and thought out construction. I would have added an off/on switch for the output. That would provide an extra chance to make sure the the voltage has been set correctly.
Success! Looks just like a bought one😆 A larger Fan, running at lower rpm, could be quieter. Nine or eleven sickle-shaped blades. Smaller the gap between blades - the quieter the noise. Less Static Pressure fluctuations, and also quiets bearing chatter. Monitor the Temperature of those Capacitors. Maybe 8 or so 1/8" holes in the base below the Caps, alow the Fan to draw some air over them. BUT hot Capacitors are an indication of a fault in-circuit. btw. You can thank Radio Shack for the holes in the sides of the terminals. Pretty smart, imo. I've start building a new bench PSU.
Very nice. I figure all serious techies should have a variable PS and a multi-meter wherever they sit down in their house. It's mandatory. I'd also check the noise voltage on the output when there was no load. PS, you should be fined for having a fine fine.
Great improvements. Some questions. What is the minimum required negative voltage? Could it be further reduced from -3.9V e.g. to -1V? Next, what is the minimum dropout voltage to achieve proper regulation? I think you have also foreseen fine control for voltage and current. What are the values for coarse and fine control potentiometers?
Thanks. I did not try to go lower than 3.9V. That was lowest voltage zener I had but I think you could get it down further but you would need it to be low enough for the op-amps output to reach 0V. The drop-out voltage for this arrangement would be about 3V given the darlington output stage and the voltage swing of the op-amps. I used single-gang 10K for the coarse and dual-gang 1K for the fine.
Pretty good build, the only thing I don't like is that the box is too small really. I've fallen into that trap too many times, nice box on shelf turns out to be an inch or so too small... 🙂 One thing I will point out about those blue boxes, many have plastic surrounds that the front and back panels screw into, and there is NO continuity between them, so if your design requires an earthed case then you need to ensure that each panel have earth straps connecting them.
Yea it looks very nice! I was wondering the same question. Was this plate ordered from other company or did you make it yourself (same with the painting and the drilling)?
Nice results. Do you know what your total costs are to build that? It's hard to find a relatively low power linear variable power supply these days. Seems most of the single output bench linear supplies are around 150W and $150, or dual+ output at 200-340W and $240-$1200 CAD. I've been looking for a dual rail one and don't need 300W but I can't find anything small that would be a good bench supply for playing around with OpAmps, etc. Might be useful to people who want to build their own if you put together a parts list with both the improved and additional parts.
Look at what spec you want carefully The kit based one here has a couple, of expensive parts, (transformer and display) that you can offset against the cost of a ready built unit. Search for linear power supplies, I've found 2A 0-15V units at about 32 GBP , higher voltage units go for around 50GBP Of course these are only single output, but if the lower spec units will suffice, then two units can be used to provide a +- type supply. There are some very cheap 2 channel liner supplies on ali - I don't know how good they are but the price is attractive for the bits inside alone if a rebuild is needed. Andy
This was about $75 CAD total. The transformer was the most expensive ($34) part and the cabinet was next ($16). The display was actually not too bad (about $5) but I got that on an AliExpress sale. It's normally twice that.
They are just PC boards without any traces. You can get precision holes made and use the silkscreen for lettering. It's a very economical way to do that. I left links to the Gerber files in teh description.
1000W! I usually power up a few op-amps, some transistors and maybe the odd micro controller or 8-bit computer. My BIG power supply is a Siglent SPD3303X rated at 220W and I have never needed all that. Just as a matter of interest, what would be powering that you need a 1000W power supply?
I routinely test electronic devices that require less than one watt. Often times these tests go on for weeks or months at a time. It’s much better to have more than one small power supply rather than one big honking power supply. You wouldn’t go out and buy a big tractor and trailer rig to haul a lawnmower.
@@mikebond6328 Sure there are small requirements for some things, that isnt the point, a 2.5AMP power supply is a waste of time to build. You are just going to be building more later for somethiing stronger, I get the point of having low power requirements but you act like you are only going to use this supply for one and only one thing
as a hobbyist i rarely set my supply current limit beyond 500mA. ( i usually set at 100mA when turning circuits up first time - just so that parts don't blow up in front of my face) why you need 1000W for.
List of changes and additions:
1) Transformer changed from 24V to 20V to reduce stress on the Op-Amps and pass transistor.
2) The expected maximum output of the power supply was then reduced from 30VDC to 20V DC. This helped deal with the ripple problem.
3) Replaced the fake Op-Amps with LM741s. Later the voltage error amp was changed to a real TL081 as the increased voltage swing gives better ripple rejection.
4) Changed D7 (negative supply Zener) from 5.1V to 3.9V to further reduce stress on the Op-Amps.
5) Changed R12 from 56K to 27K to limit maximum output to 20V
6) Changed R18 from 56K to 80K to maximize range of control over current limit.
7) Changed R2 from 1/4W to 1/2W (you could even got to 1W).
8) increased filter capacitor from 3300uF to 9900uf and changed from troublesome Won Haung Lo brand to Nichicon.
9) Dramatically Increased size of heatsink from original design.
10) Added fine controls for both Voltage and Current adjustment.
11) installed polarized headers for the fan and to facilitate remote mounting of current limiting LED.
12) Installed terminal block to facilitate remote mounting of the pass transistor.
And of course we added a nice LCD panel meter with V/A/P and put it all in a cabinet.
Well done for this ps ckt and your design of the equipment !😊
Thanks. The circuit is not mine, but I had to make a few changes to it to get it working. I'll leave those in a pinned comment.
Wow! You actually made a really wonderful and serviceable power supply, from that rather poor kit. Kudos! Bravo! I think the look of it is quite handsome, as well. I am sure that it will be a joy, for years to come. Cheers!
Thank you very much!
It needs some kind of calibration scale for the current limit knob, so it can be set reasonably accurately without having to short the output to set the required current. Not that shorting them is a problem, it just makes it slightly easier and you can just look at the knob to see what current it is set to. Set assuming the fine knob is centered. A very interesting video series by the way, AliExpress should link to it!
Yeah, a scale on the current adjust would have been a good idea. Thanks!
Nice project and thought out construction. I would have added an off/on switch for the output. That would provide an extra chance to make sure the the voltage has been set correctly.
Great idea, I'd actually put it in between the voltage error amp the the pass transistors. Might still do that.
looks nice.
Thanks!
Nice job. Now that you have a TL081 as IC2, you can accurately adjust the 0V with the offset trimmer on the board.
You are right. I had it set for the 741 that was in there before. I'll have to give it a tweak to get rid of the last 7mV or so.
Success! Looks just like a bought one😆
A larger Fan, running at lower rpm, could be quieter. Nine or eleven sickle-shaped blades. Smaller the gap between blades - the quieter the noise. Less Static Pressure fluctuations, and also quiets bearing chatter.
Monitor the Temperature of those Capacitors. Maybe 8 or so 1/8" holes in the base below the Caps, alow the Fan to draw some air over them. BUT hot Capacitors are an indication of a fault in-circuit.
btw. You can thank Radio Shack for the holes in the sides of the terminals. Pretty smart, imo.
I've start building a new bench PSU.
LOL! Thanks and good information!
Very nice. I figure all serious techies should have a variable PS and a multi-meter wherever they sit down in their house. It's mandatory. I'd also check the noise voltage on the output when there was no load. PS, you should be fined for having a fine fine.
LOL! I did check it a low voltage and at zero. Still about 50mV - just the ambient noise.
Great improvements. Some questions. What is the minimum required negative voltage? Could it be further reduced from -3.9V e.g. to -1V? Next, what is the minimum dropout voltage to achieve proper regulation? I think you have also foreseen fine control for voltage and current. What are the values for coarse and fine control potentiometers?
Thanks. I did not try to go lower than 3.9V. That was lowest voltage zener I had but I think you could get it down further but you would need it to be low enough for the op-amps output to reach 0V. The drop-out voltage for this arrangement would be about 3V given the darlington output stage and the voltage swing of the op-amps. I used single-gang 10K for the coarse and dual-gang 1K for the fine.
Thanks 👍
Thank you too!
Pretty good build, the only thing I don't like is that the box is too small really. I've fallen into that trap too many times, nice box on shelf turns out to be an inch or so too small... 🙂
One thing I will point out about those blue boxes, many have plastic surrounds that the front and back panels screw into, and there is NO continuity between them, so if your design requires an earthed case then you need to ensure that each panel have earth straps connecting them.
Yeah, It's a bit tight. I do have (should) to ground the top, but the front and back panels are fiberglass (PC boards).
how do you made the labeling in front panel? its look neat
Yea it looks very nice!
I was wondering the same question. Was this plate ordered from other company or did you make it yourself (same with the painting and the drilling)?
They are actually just PCB boards with no traces on them. Easiest and cheapest way to get custom panels made.
Great videos, please keep it going! Have plans to make your own power supply design? Would like to build one of those :)
I have been thinking about designing my own. My first look at this design I thought it was fairly good, but there are things that could be improved.
Nice results. Do you know what your total costs are to build that? It's hard to find a relatively low power linear variable power supply these days. Seems most of the single output bench linear supplies are around 150W and $150, or dual+ output at 200-340W and $240-$1200 CAD. I've been looking for a dual rail one and don't need 300W but I can't find anything small that would be a good bench supply for playing around with OpAmps, etc.
Might be useful to people who want to build their own if you put together a parts list with both the improved and additional parts.
Look at what spec you want carefully The kit based one here has a couple, of expensive parts, (transformer and display) that you can offset against the cost of a ready built unit. Search for linear power supplies, I've found 2A 0-15V units at about 32 GBP , higher voltage units go for around 50GBP Of course these are only single output, but if the lower spec units will suffice, then two units can be used to provide a +- type supply.
There are some very cheap 2 channel liner supplies on ali - I don't know how good they are but the price is attractive for the bits inside alone if a rebuild is needed.
Andy
This was about $75 CAD total. The transformer was the most expensive ($34) part and the cabinet was next ($16). The display was actually not too bad (about $5) but I got that on an AliExpress sale. It's normally twice that.
@@andye2005 Yeah, have to agree. This was more a project to see if I could get this thing working properly more than anything else.
Will you be adding the front and rear panels to the pcbway shared projects?
No, but I'll put a link to the Gerber files in the description in a day or two.
Links are there now!
Great job and beautiful face how did you make that, with such nice lettering?
They are just PC boards without any traces. You can get precision holes made and use the silkscreen for lettering. It's a very economical way to do that. I left links to the Gerber files in teh description.
@@uni-byte Good idea I should try it sometime.
Thanks
@@ultimateworkshop2000 You're welcome!
Of all the junky things that come out of China the junky power supplies bother me the most. They can cause so much damage and wasted time.
Yeah, they can. Most of them can be fixed .. like this one though. It started out not so good from a cheap Chinese kit.
Why would you make such a weak power supply anything less than 1000W is a waste of time imo
For most hobbyists who want to make their own lab power supply, this is not too expensive, not too complex, and it will cover their needs.
1000W! I usually power up a few op-amps, some transistors and maybe the odd micro controller or 8-bit computer. My BIG power supply is a Siglent SPD3303X rated at 220W and I have never needed all that. Just as a matter of interest, what would be powering that you need a 1000W power supply?
I routinely test electronic devices that require less than one watt. Often times these tests go on for weeks or months at a time. It’s much better to have more than one small power supply rather than one big honking power supply. You wouldn’t go out and buy a big tractor and trailer rig to haul a lawnmower.
@@mikebond6328 Sure there are small requirements for some things, that isnt the point, a 2.5AMP power supply is a waste of time to build. You are just going to be building more later for somethiing stronger, I get the point of having low power requirements but you act like you are only going to use this supply for one and only one thing
as a hobbyist i rarely set my supply current limit beyond 500mA. ( i usually set at 100mA when turning circuits up first time - just so that parts don't blow up in front of my face)
why you need 1000W for.