Creating a custom voltage on a toroidal transformer

[William Gaatjes]

That does not make a lot of sense. Assuming you have the raw input voltage high enough.

But anyway, I thought the 33V rail (without extensively analyzing it), was only going to use a few milliamps. Higher currents during switching transitions, would hopefully be coped with by decoupling capacitors etc.

Also 30mA seems rather high (assuming no load connected). Surely it should be more like around 12 milliamps ? (4K7) + output ones.

Yeah, it should be 1mA for the voltage set resistor ladder, 2.5mA for the current set ladder and 12mA for the current source and worst case 1mA (Reserved estimate)for the opamps. 16.5mA total current consumption. When i do a load change test on the 33V line from between 10mA and 20mA, the voltage difference on the 33V line is about 23mV. More acceptable.

EDIT: The total current is 16.5mA +10mA and +20mA. Thus 26.5mA to 36.5mA.
I forgot the load current which i used to see the dynamic response of the 33V supply. So, maximum 16.5mA is used by the circuit.

 

[SOFTengCOMPelec]

Yeah, it should be 1mA for the voltage set resistor ladder, 2.5mA for the current set ladder and 12mA for the current source and worst case 1mA (Reserved estimate)for the opamps. 16.5mA total current consumption. When i do a load change test on the 33V line from between 10mA and 20mA, the voltage difference on the 33V line is about 23mV. More acceptable.

EDIT: The total current is 16.5mA +10mA and +20mA. Thus 26.5mA to 36.5mA.
I forgot the load current which i used to see the dynamic response of the 33V supply. So, maximum 16.5mA is used by the circuit.

If it's going from 23mV to 600mV (with a proportionally much lower change in current level), maybe the resistor (4700) is too low, or something. As that is a very non-linear change.

I.e. Something funny is going on (at the higher currents).

 

[William Gaatjes]

If it's going from 23mV to 600mV (with a proportionally much lower change in current level), maybe the resistor (4700) is too low, or something. As that is a very non-linear change.

I.e. Something funny is going on (at the higher currents).

As it turns out, the resistor seems to be too high. I lowered it to 1K and now the voltage delta is 8mV with a 10mA to 20mA load change. And 133mV with a 10mA to 90mA load change.

 

[William Gaatjes]

This sure looks better. With a load change of 5mA, the delta on the 33V line is about 6mV.

Now i am going to add a sine wave voltage source , a diode bridge and a bulk capacitor. To see how the 33V line is stabilized.

 

[William Gaatjes]

The circuit becomes pretty stable now. the delta at the 33V line is now less than 4mV. While the load changes at the output between 0,65A and 4,65A.

I added an electrolytic capacitor to filter out some noise at the base of Q4.

 

[SOFTengCOMPelec]

As it turns out, the resistor seems to be too high. I lowered it to 1K and now the voltage delta is 8mV with a 10mA to 20mA load change. And 133mV with a 10mA to 90mA load change.

Sorry, I meant too low, current wise, and too high, value wise.
tl;dr
I meant too high.

In practice, it would probably give you worse results. Because there will be a fair amount of heating (over time), which would drift the component values. Such as Veb (Voltage Emitter to Base), at any given current.

 

[SOFTengCOMPelec]

Your diagrams look like this:

But they look better, like this:

 

[William Gaatjes]

Sorry, I meant too low, current wise, and too high, value wise.
tl;dr
I meant too high.

In practice, it would probably give you worse results. Because there will be a fair amount of heating (over time), which would drift the component values. Such as Veb (Voltage Emitter to Base), at any given current.

Well, the circuit should sort of compensate for that. Bipolar transistors conduct more at higher temperatures. So, Q4 will attempt to give of a higher voltage but Q6 will also conduct more, lowering the voltage again. So, some equilibrium will be found.

 

[William Gaatjes]

But they look better, like this:

How do you do that ?

EDIT:
Aha :
"~original"

I will add that to all pictures.

 

[SOFTengCOMPelec]

How do you do that ?

EDIT:
Aha :
"~original"

I will add that to all pictures.

View Image = +magnifying glass icon, on the Photobucket page.

Then right click on that (there may be a better way), and open a new tab (or window, if you prefer).

The new tab's URL address, gives a full sized (original) "link".

Alternatively, it seems to just add "~original", to the right of the *.jpg bit.

E.g. gfakjegfajkefgkaegfkjaehfkjaef.jpg becomes gfakjegfajkefgkaegfkjaehfkjaef.jpg~original

But first method is probably best/safest, as Urls can change over time.

 

[William Gaatjes]

View Image = +magnifying glass icon, on the Photobucket page.

Then right click on that (there may be a better way), and open a new tab (or window, if you prefer).

The new tab's URL address, gives a full sized (original) "link".

Alternatively, it seems to just add "~original", to the right of the *.jpg bit.

E.g. gfakjegfajkefgkaegfkjaehfkjaef.jpg becomes gfakjegfajkefgkaegfkjaehfkjaef.jpg~original

But first method is probably best/safest, as Urls can change over time.

Strange that photobucket scales the image. I guess for bandwidth reasons.

 

[SOFTengCOMPelec]

Strange that photobucket scales the image. I guess for bandwidth reasons.

Yes, probably for bandwidth reasons.

The only other thing that limits the bandwidth, is if the AI software at photobucket, notices that there are serious design faults in any circuits.
In which case it limits the resolution, to hopefully limit the damage such faulty circuits create.

So as soon as I fixed your op-amp, controlling its own power supply pins, it became high resolution.

If you believe the above is true, correct and not a joke. Please send me the $2,000 that you owe me, from 6 months ago. Surely you have not forgotten.

 

[William Gaatjes]

Yes, probably for bandwidth reasons.

The only other thing that limits the bandwidth, is if the AI software at photobucket, notices that there are serious design faults in any circuits.
In which case it limits the resolution, to hopefully limit the damage such faulty circuits create.

So as soon as I fixed your op-amp, controlling its own power supply pins, it became high resolution.

If you believe the above is true, correct and not a joke. Please send me the $2,000 that you owe me, from 6 months ago. Surely you have not forgotten.

Very possible. And i did not forget. But i hope that you did not forget that i would pay you 2000 zim dollars.

 

[SOFTengCOMPelec]

Very possible. And i did not forget. But i hope that you did not forget that i would pay you 2000 zim dollars.

Apparently ~2000 Zimbabwean dollars is no longer dealt with by existing banks (possible joke). So at great cost you would have to travel there. This would be extremely dangerous, and you could easily be injured.

So out of the kindness of my heart, I will accept $4,000 American dollars instead.

There is no need to thank me, but if you insist, please add a 25% tip. New total = $5,000.

For $6,000 I will give you completely free advice, on how to avoid paying me the $5,000. Since it is above $5999, you will have to pay VAT as well (possible joke).

New total = $6,000 + $1,200 (20%) = $7,200.

 

[MrTeal]

That's looking pretty good, for a first shot at it.

Ignoring the fact that you are going to alter/improve it anyway. I hope you don't mind if I say a few quick things about it, given the way it is at the moment.

The voltage control potentiometer (R7), can be slightly noisy and intermittent, for a tiny fraction of a second. Usually much less so, if it is a quality, multi-turn one. But even so.
Because you have a wiper, e.g. Metal, in contact with the resistive material. As it moves along it, it can briefly disconnect.

E.g. Think of worn out volume controls on some equipment, where it makes loud pops and noises, as you move the volume control up and down.

This brief (e.g. 3.5 millisecond) disconnection (which if you are lucky does not occur. But it can and might), would cause the output voltage to rise to maximum (ignoring the relay centre tap selector) voltage.

If there were brief 30V (approx), voltage spikes, for a few milliseconds, when it is set to say, +5V for a PIC project. It could easily break your external devices.

In the circuit you have got now, putting in a capacitor to minimise that effect, would mess up the constant current circuitry. But there are ways of doing it, which I will leave you to consider.

Old single turn, carbon potentiometers, were notoriously bad at doing this. So I stand corrected, if modern, quality ones, don't need this capacitor and/or other solutions.

I think expensive equipment in general (VERY top end), may use optical turn detection devices (going WAY over the sophistication of your project, and WAY too expensive/complicated). Which largely eliminates that problem. Since a microcontroller, simply watches the up/down pulses (or whatever), from the optical turn device.
There are other ways of doing it.

Q3 2N5550, may be in danger of having its Emitter/Base junction, damaged by exceeding the 6V maximum, "reverse" voltage, applied to it.

E.g. Imagine the output is peacefully at exactly 15 Volts.

There would be 15 Volts, almost directly connected to Q3's Emitter. So if the Op-amp U1, suddenly turned off (ok, swings towards -5V/negative rail). E.g. The output current, suddenly exceeds the set limit.

U1's output would drop to about -5 volts (rail to rail), which connects to the base (via a resistor) of Q3.

So Q3 would (briefly), be at 21 Volts between the Base and Emitter, in the "wrong" direction. As well as being more than the +6V allowable on the datasheet. It would probably rapidly, or sooner or later, damage/destroy the device.

If it was me, I would probably put in a small signal diode, to eliminate the problem. Being careful to NOT effect other circuit functions.

Even some decent branded lab supplies suffer from this problem. I have a BK Precision 1795 rack mount 0-60V, 0-10A supply that developed the same issue on the coarse voltage pot. In a couple places (including right around 5V) the pot had an issue and the output voltage would jump to 60V.

It always amazed me that a $2000 power supply wouldn't be designed to go to 0V if the pot is open or at full resistance, rather than jumping to full voltage.

 

[SOFTengCOMPelec]

Even some decent branded lab supplies suffer from this problem. I have a BK Precision 1795 rack mount 0-60V, 0-10A supply that developed the same issue on the coarse voltage pot. In a couple places (including right around 5V) the pot had an issue and the output voltage would jump to 60V.

It always amazed me that a $2000 power supply wouldn't be designed to go to 0V if the pot is open or at full resistance, rather than jumping to full voltage.

I had the same idea, but to avoid unnecessarily complicated his PSU project, I kept quiet about it. He could have used an op-amp (actually there may be U4 spare now), to invert it, so that the control voltage would go to 0V, rather than max volts. But I have not properly thought it out.

One solution is to discipline oneself, to only adjust the voltages, when the voltage output switch is in the off/isolate position (if it has this). Also make sure the current limit knob is sensibly set.

An old PSU I have got, has exhibited the same problem (occasional overvoltages), but I can only very vaguely remember about it. So I could be wrong. As well as being old, it is very well used as well. Contact cleaner might help.

There are also types of overvoltage protectors (or similar, inside the PSU, if max output is 30V, probably trips at >= 34V), which better ones can have. The old PSU that I just mentioned, seems to have such a device built in, but I don't have the manuals for the PSU, so don't know the details of it.
But it has sometimes (quite rightly), shut itself down, automatically. Probably overvoltage detection or something. Typically I was over using inductive loads, so an overvoltage trip makes lots of sense.

I hate PSU's with output fuses (but at least it is SAFE, rather old PSUs use to do this, in some cases). I would end up blowing the fuses, like every hour. Until I ran out of fuses. Automatic reseting and/or push to reset ones (breakers) are fine. But all electronic current constant/limiters are best.
But it is good and safe, if the PSU has an overall mains fuse, near the transformer, to avoid/minimize it being a fire/etc hazard.

 

[William Gaatjes]

I had the same idea, but to avoid unnecessarily complicated his PSU project, I kept quiet about it. He could have used an op-amp (actually there may be U4 spare now), to invert it, so that the control voltage would go to 0V, rather than max volts. But I have not properly thought it out.

In the latest iteration of the schematic, just add a 10MOhm resistor from the non inverting input of U1 to ground. It will produce a tiny bit of offset, but that is not important since the supply has panel meters used to set the voltage and a calibrated scale is not used. Whenever the voltage potentiometer would stop functioning, the non inverting would go slowly to ground level and no spikes would occur. Benefit of not using the current source i had before.

I also found the opamp regulating it's own supply in the Horowitz/Hill. It is nothing like my circuit. And i did come up with a reasonable explanation why my circuit does work. I will save that for later, cause i have now a huge cold.

EDIT:
Nevermind, thought of all reasons why it should work (PNP pass element, resistor to start up, LT1491 over the top inputs) except forgot one and only reason why it does not work. Ripple rejection is awful because the pnp pass transistor is always almost saturated.

 

[SOFTengCOMPelec]

I also found the opamp regulating it's own supply in the Horowitz/Hill. It is nothing like my circuit. And i did come up with a reasonable explanation why my circuit does work. I will save that for later, cause i have now a huge cold.

Ignoring your circuit and exact/specific details, to make it easier for me to explain.

So in general.

The Op-amp would need to produce a voltage (at its output), potentially a bit above (depending on the exact circuit), the supply voltage. This is completely impossible.

You have got the diode base/emitter junction (of the transistor), to attempt to lower that voltage requirement, by about 600/700mV. But it is still a very bad idea. I'm not even sure it would work, let alone be a robust and reliable technique.

Any overshoot, interference, high voltage interference on the incoming mains (which sometimes briefly can partly get through, to an extent), would immediately break the datasheet. E.g. Maximum input voltage, common mode range, etc etc . Depending on the exact datasheet op-amp model.
Exceeding the common mode range, and/or other parameters, may latch up, damage or other undesirable effects, could occur.

Once your circuit were to ever exceed the max output voltage of the op-amp plus a few hundred millivolts or so. It appears that the transistor, would "latch up", and permanently be energised (by the too low voltage on the output of the op-amp).
I.e. Although the op-amp is rail to rail, it still has datasheet limitations. Such as voltage drops, when any current is flowing, lack of ability to exceed the positive supply terminal, and max voltage limits, which attempting to exceed, would break the datasheet rules and potentially not work either.

I can see much worse technical problems, which would be likely to stop it from working correctly, but would take ages to explain.
There is tons more I could explain what is wrong with that type of idea, but I will stop now.

To give a quick counter point, to my own stuff above.

It is true that some major, manufactured items, have taken various short cuts (at times), in their electronics. With all manor of "crazy", but actually it works, ideas.

But to end. I would strongly recommend that internal power supplies, for your electronics, is done in a robust, reliable, tried and tested way. By and large there is no need to try wild stuff.

Sometimes clever electronics can be made by exploiting electronics stuff, in unusual ways. (Outside of critical bits, such as internal PSU's)

But I would still recommend that the internal power supply, is done in a somewhat regular way.

tl;dr
Any issues/noise/surges/faults/unreliability in your internal power supplies, will usually/often ruin the project. I.e. It won't work properly.

Just because something works (I'm not sure if that would even be the case), and is almost within the datasheet specifications (I doubt it, without doing a huge analysis job). Does not mean that it is a good idea.

If you are determined to do it. Then why not do it with a separate op-amp (high voltage), which is powered off the incoming voltage.

Get well soon!

 

[SOFTengCOMPelec]

Can I check my understanding of how your op-amp self-regulating of its own psu, works ?

Ignoring arguments of how/why the following happens for now. I need some defined points to mutually agree on how it behaves.

Assuming Incoming (post bridge rectifiers) is fixed at exactly 40 Volts (I know in practice it would have ripple).

Also assuming (somehow, we can argue how, later), the internal Vcc +33V, has ended up, also being at exactly 40 Volts (perhaps just for an instant, because of interference or something).

Finally lets assume/pretend(if necessary) that the op-amp CAN'T ever exceed an output voltage of 39 Volts. Ignoring the datasheet and practice for the second. The datasheet, at a quick glance does not seem to be telling me how the op-amp, handles higher voltages than +15V, above ground (30V from -ve rail). I'm not even sure at higher currents and voltages, that op-amp would even be able to succeed in reliably switching off Q4, because the voltage drop (between the output pin of Q4 and +ve supply pin, could be too big, to reliably switch off the Q4 transistor.

What will happen ?

My quick/brief analysis, seems to indicate that Q4 will now be permently forward biased. I.e. conducting a significant current.

I.e. It will be jammed at close to 40 Volts, even though it should be at 33 Volts.

Because the output of U4 (artificially limited to a maximum of 39 volts by me here. But in practice the datasheet and real life device, will have their own, potentially similar characteristics).

 

[SOFTengCOMPelec]

Ignoring my objections to the fact that it would probably not work and/or be a bad idea.

Can I try and move it into something, which would be somewhat better, anyway.

Your circuit is trying to control it, right at the top end, of the output voltage swing, of the op-amp. It may not even be able to reach that high a voltage, in practice.
E.g. Voltage drops under load, etc etc.

A circuit configuration such as the following (diagram shamelessly linked from elsewhere)

Brings the voltage(s) that the op-amp output is swinging between, to a much more realistic level. (Zener diode 12V).

In their circuit, they HAD to do this, because the 741 is NOT rail to rail on the output.

But the concept, is the same or similar.

 

[William Gaatjes]

Well, i was looking for as few as possible components and i could remember a high voltage regulator but could not come up with the type number.
ignoring all possible solution and going for a simple solution :

I found it. The TL783 will do an input voltage of maximum 125V.
So, i will use that regulator to generate the 33V for the circuit. Then i have what i want for as few components as possible.

http://www.ti.com/product/tl783

I first have to find out what the minimum voltage difference may be between input and output. Well, i have some reading to do later on.

 

[William Gaatjes]

Have my lunch break now,
Been continuing with looking for a solution to get 33V with as few as components as possible bit reliable.
And believe it or not, there exists a 60V version of the LM317. The LM317HV. This model does 60V input. I will use a LM317HV to create the required 33V to power the circuit. And Farnell and Mouser have it available. The circuit with the LT1491 will then be the real bench supply. This way, i only need one adjust potentiometer to set the 33V for the rest of the circuit. Now i got that solved, i can continue focusing on dissipation issues.

 

[SOFTengCOMPelec]

Have my lunch break now,
Been continuing with looking for a solution to get 33V with as few as components as possible bit reliable.
And believe it or not, there exists a 60V version of the LM317. The LM317HV. This model does 60V input. I will use a LM317HV to create the required 33V to power the circuit. And Farnell and Mouser have it available. The circuit with the LT1491 will then be the real bench supply. This way, i only need one adjust potentiometer to set the 33V for the rest of the circuit. Now i got that solved, i can continue focusing on dissipation issues.

That's fine. When I looked at it (Farnell), the LM317AHVT is a lot cheaper.
It's also >=57V (possibly a tiny bit more).

One off price, £0.60, rather than almost three times that.

http://www.farnell.com/datasheets/1949043.pdf

http://uk.farnell.com/fairchild-semiconductor/lm317ahvt/ic-ldo-volt-reg-57v-1-5a-to-220/dp/1564497

 

[William Gaatjes]

I will exchange the protection diodes for BAS31 types. Two diodes in one small sot-23 case.

I will make a schematic with LTspice comprised with the components i am going to use. Unfortunately, since these components are not having a spice model, simulation will not be possible. But at least i nice schematic will be available. Will take some time. I am sick at the moment. Those flu's get more aggressive each few years it seems. I am coughing up cheese from inside my lungs.

 

[SOFTengCOMPelec]

I will exchange the protection diodes for BAS31 types. Two diodes in one small sot-23 case.

I will make a schematic with LTspice comprised with the components i am going to use. Unfortunately, since these components are not having a spice model, simulation will not be possible. But at least i nice schematic will be available. Will take some time. I am sick at the moment. Those flu's get more aggressive each few years it seems. I am coughing up cheese from inside my lungs.