# electro[epsco] power supply

#### KMoffett

Joined Dec 19, 2007
2,784
OK, I took a look through the louvers on the top of my "Epsco NFB 32V/15A" this morning. I was wrong on the rating. It's the same as yours. It's a Variac style variable auto transformer.

You can buy any number of 30VDC/20A CV CC bench supplies from $250 to$400... if that's what you need. If I was going to have only one bench supply in that rating range, I would go for one of those. Which I did. But when the Epsco showed up, I found many uses for it off the bench...electrolysis derusting, foam cutter, sheet acrylic bender, and wind tunnel fan. None require any special regulation, just adjustable voltage with a lot of current.

Ken

#### ebp

Joined Feb 8, 2018
2,332
Using a microcontroller for PID is certainly possible but is even more demanding in terms of knowledge. It could produce better regulation if done extremely carefully and through use of operating mode variation such as transitioning from fixed frequency to fixed on time or fixed off time. Standard analog control is PID and it is making that work properly that is a significant part of the challenge. Do you even know what poles and zeros are? Do you have any concept of criteria for stability and the objectives for gain, phase and slope?

Digital ICs normally play no role in switchmode power supplies other than in supervisory positions. There are logic elements internal to controller ICs to steer outputs for the variants of bridge type converters and to manage the array of protections and interlocks that exist in most modern controllers.

No knowledgeable designer is even going to consider IGBTs for less than 600 volts and probably not for less than 1 kV. They have been much improved over the years, but FETs even more-so. IGBT switching losses, which are a big player in SMPS efficiency, will be substantially higher. Conduction losses are higher until you get well up in voltage handling requirement. At under a hundred volts they would typically be grossly inferior.

"And you can use IGBTs for the high power switching that could then go to a filter and use BJTs with a varying voltage applied to get reliable constant current."
Whut? You propose to use a linear regulator downstream of a switcher? That can certainly be done and can have benefits but it has be be managed extra carefully for efficiency, dynamic performance and device safe operating are. I've used such a technique to accurately control current at microwatt power level from a cathodic protection 500 W switcher. You now have introduced the requirement for another control loop with more of the pesky poles and zeros. You can't just "vary voltage applied." The switcher output voltage must be made to track the final output voltage unless you are wiling to accept efficiency approaching zero.
I will also point out that a current sensor, typically a resistor, that is suitable for say 10 amperes is putting out 0.1% of full-scale at 10 mA - 100 parts per million at 1 mA. That's 50 µV at 10 mA for a typical shunt; 5 µV at 1 mA. You better be very well versed in design of precision analog amplifiers to cope with that. The typical input offset voltage for cheap op amps will run 20 times the signal voltage at 10 mA and 200 times at 1 mA. The drift in offset voltage with an ill-selected amp could easily cause the set current to vary by a factor of 10 or 20 or 50. You can certainly ease requirements if you don't mind a current sense shunt that eats 1 or 2% of output power. Know what Manganin is?

#### -live wire-

Joined Dec 22, 2017
902
No knowledgeable designer is even going to consider IGBTs for less than 600 volts and probably not for less than 1 kV. They have been much improved over the years, but FETs even more-so. IGBT switching losses, which are a big player in SMPS efficiency, will be substantially higher. Conduction losses are higher until you get well up in voltage handling requirement. At under a hundred volts they would typically be grossly inferior.
My understanding was that at medium high frequencies and with high currents IGBTs reach a point where they become more efficient then fets. People often tell me that fets are worse when you have higher power (medium voltage high current), and to just use IGBTs. I assumed that their non-idealities would become greatly amplified for switching higher power.
Know what Manganin is?
I have heard that manganin is a copper alloy used for many different electrical things.
Do you even know what poles and zeros are? Do you have any concept of criteria for stability and the objectives for gain, phase and slope?
I am not too knowledgeable when it comes to control systems. I understand the basics, but am not really familiar with most of those terms in this context.

#### nicklus

Joined Sep 11, 2017
8
OK, I took a look through the louvers on the top of my "Epsco NFB 32V/15A" this morning. I was wrong on the rating. It's the same as yours. It's a Variac style variable auto transformer.

You can buy any number of 30VDC/20A CV CC bench supplies from $250 to$400... if that's what you need. If I was going to have only one bench supply in that rating range, I would go for one of those. Which I did. But when the Epsco showed up, I found many uses for it off the bench...electrolysis derusting, foam cutter, sheet acrylic bender, and wind tunnel fan. None require any special regulation, just adjustable voltage with a lot of current.

Ken
Exactly !!! That was my intent from the start. Just a supply for some electrical projects [ nothing electronic] like you have mentioned above. Thank You!!

#### MisterBill2

Joined Jan 23, 2018
4,929
Many years ago I also assembled such a power supply, with a smaller variable transformer feeding a 12 volt step down transformer that drove a diode bridge. It is still useful for a lot of things. if I need a regulated voltage then I use the regulated supply, but for checking solenoids , motors, lights, and stuff the adjustable but not regulated supply works well. Just like most tools, there is a use for each, and they are different.
I had no idea that such a fairly clear question would create such a lot of discussion. And I am wondering if there is a cheaper way than replacing the variable transformer. Something cheap and easy to make work that will deliver adequate performance. That will not be a switching power supply. How about a phase-controlled system? Does anybody have ideas there that DO NOT include an arduino?

#### -live wire-

Joined Dec 22, 2017
902
I had no idea that such a fairly clear question would create such a lot of discussion. And I am wondering if there is a cheaper way than replacing the variable transformer. Something cheap and easy to make work that will deliver adequate performance. That will not be a switching power supply. How about a phase-controlled system? Does anybody have ideas there that DO NOT include an arduino?
I feel like inexperienced electronics enthusiasts often suggest arduino for everything, even when and analog approach is MUCH better, or arduino is completely not suited or overkill for a task. Especially here, the only reason really for one e would be for a display of some sort and maybe supervisory stuff. But no SMPSs?

So what exactly is a phase controlled system? Is it like phase angle firing? And if that is what you are referring to, than I have an idea. Could you phase angle fire it then filter it with some inductors and caps? Would you be able to still acheive a stable sine wave on the output? I know that it would certainly not be linear, and there would obviously be no or poor regulation, but would it at least work?

#### MisterBill2

Joined Jan 23, 2018
4,929
I feel like inexperienced electronics enthusiasts often suggest arduino for everything, even when and analog approach is MUCH better, or arduino is completely not suited or overkill for a task. Especially here, the only reason really for one e would be for a display of some sort and maybe supervisory stuff. But no SMPSs?

So what exactly is a phase controlled system? Is it like phase angle firing? And if that is what you are referring to, than I have an idea. Could you phase angle fire it then filter it with some inductors and caps? Would you be able to still acheive a stable sine wave on the output? I know that it would certainly not be linear, and there would obviously be no or poor regulation, but would it at least work?
Phase angle control is done on the primary side of the step down transformer in rectifier type power supplies. The waveform is no longer a nice sine wave, but since it is feeding rectifiers that is OK. The advantage is that lower current devices can be used for control, which at one time was a cost advantage. Today it may work using a properly sized controller such as is used to speed control induction motors driving fans. Adding feedback gets complicated but is still far simpler than a switching regulator.

#### MisterBill2

Joined Jan 23, 2018
4,929
There have been a few phase controlled power supply designs published over the years, unfortunately some of them used unijunction transistors, which seem to be sort of exotic now.
But typically a phase controlled regulator has a ramp generator synchronized with the AC line frequency, and the difference between a sample of the output voltage and some reference voltage moves the trigger point up or down on the ramp, thus controlling the power delivered into the filter part of the supply. So the response time is based on the line frequency, slower than present day switchers, but still fast enough for many applications. If it is run open loop, then the control just changes where on the ramp the SCR switches on. Not much different than a lamp dimmer, except with greater range and a more linear control.

#### nicklus

Joined Sep 11, 2017
8
Well -live wire-, they do what they do. Right now mine is powering two high performance race car 12V radiator fans connected in series for 30V/15A for a vertical wind tunnel. Yes, I also have an switch mode, constant voltage, constant current, 30VDC/20A supply. But that's overkill for this job. Sometimes heavy metal is all you need...especially if you already have it.

Ken
The Epsco is what I need for what I'm doing right now.

#### KMoffett

Joined Dec 19, 2007
2,784
Nicklus,

Can you take some interior photos to compare with mine to see if any other parts are missing?

Ken

#### nicklus

Joined Sep 11, 2017
8
Nicklus,

Can you take some interior photos to compare with mine to see if any other parts are missing?

Ken
I know it's just the autotransformer missing. I just don't know the value of it.

#### MisterBill2

Joined Jan 23, 2018
4,929
Two parameters define the variable transformer, once you know the brand. They are the input voltage, either 120 or 240 volts, and then the power level, which if the unit that you have has a plate with input voltage and current, or voltage and wattage, listed, then you are all set. Also, the physical size will reduce the number of possibilities as well. But if you decide that the price of a replacement variable transformer is excessive, once you know the voltage and the power it is entirely possible to wind your own multiple tapped autotransformer and you will be able to have a selection of voltages available by adding a tap changing switch. That can work out very well and it has the added advantage that a tap selecting system is more robust than the variable transformer type of system.