Archive for Power Supply Project

Zener Pre-Regular Revisit

In revisiting my power supply project I also revisited a number of unanswered questions. Chief among them is how on earth do I get rid of the excess voltage from the AC rectification? Previously, I had mentioned this was somewhat of a shock to me as a novice that a 30V AC tap from a transformer can gain 12.6V in the rectification to DC. I know part of that is the combining of the AC waveforms and the bumping up by the mammoth amount of capacitance I have on it to smooth it. I had originally tried some very dodgy and very ugly collection of series power diodes which plain just would not work. They were not only ugly, but ridiculously unsafe and would prevent the proper operation of the circuit at low currents and would be unreliable at high currents. Scrap that. I’ll consign that to the embarrassing fail bin.

The next idea was to pre-regulate the voltage down to a safe level for the downstream regulators and prevent unnecessary power dissipation. I chronicled before some shaky success but discarded that idea after it proved somewhat lacking and prone to smoke. I also thought that now three darlington transistors in the signal path was somehow wrong, there was something about it I didn’t like for some reason.

In poking about again with fresh eyes and a clear mind I decided once again to make a zener pre-regulator, having it control the base of a darlington transistor to set the output voltage to just shy of 30V. Most of the original concept stayed the same with a few little modifications to make it safer and include the proper ratings of components as well as ensuring that no datasheet “Absolute Maximum Ratings” were being flirted with. Schematic below. (please note the caveats at the bottom of this post)


As I had been down this road before, I was tickled to discover that I already had everything I needed in my parts bins and with my 10Ω home made power resistor just completed I set to marrying it all together. Having only a few parts it was rather trivial to assemble it.

I test powered up the AC board as I had not touched it in a year and I got that delightful hum and that crazy 85.2V reading between the positive and negative rails. I kind of freaked at that moment, not only because 85.2 is a lot of volts but I realized I really need to be extra safe with this thing. Also I forgot that it was centre-tapped and i just measured the negative lead (-42.6V) unnecessarily. I cut the power to it and noticed my multimeter barely dipped. I realized that the 10 milliFarads of capacitance I had on the thing to smooth the power is not only extremely dangerous when charged, but would probably take a decade to discharge though the multimeter’s very high input impedance. Rather than touch the positive and negative wires together to discharge the caps instantly (which would have resulted in a very big and dangerous bang) I carefully placed them on a 30Ω power resistor I had to drain the caps quickly and gracefully.

This is why you are always told to never touch capacitors when opening up equipment as they could be charged still. They must be discharged. Smart is using a low value power resistor to “bleed” them dry of charge. Stupid is shorting the terminals with a screwdriver. For safety, I will include such a resistor – a “bleeder resistor” – to discharge the caps when it is switched off.

Anyway, with the AC board working great, it was time to hook up the latest candidate for a pre-regulator and try it with some loads.

The setup. From left to right: my 10Ω home made dummy load resistor, the zener pre-regulator, and the AC board

The setup. From left to right: my 10Ω home made dummy load resistor, the zener pre-regulator, and the AC board

It worked sort of fine though the numbers were of course somewhat off from my simulated circuit. For one thing the 5W 30V zener I was using led to a regulated voltage of 32ish volts which was higher than I wanted it to be. For my stuff to work well I needed it about 26-29V. I needed enough headroom for the eventual voltage regulator to make a nice steady 24V yet as low as possible to reduce the power it will dissipate due to the voltage differential. On a whim I whacked in the 1W 30V zener I had and behold – I got 28-29V. Perfect. Just what I wanted.

I tried a variety of loads including: a 1k resistor, 100Ω power resistor, 30Ω power resistor, and yes – my monster of a 10Ω resistor pictured here in glowing glory as it dissipates something like 90W of power.


Overall I would call it a success with some caveats. I did notice a change in voltage depending on the load I was putting across it. This is not a huge deal as I do not need it to be an accurate voltage regulator, but i do need it to stay under 30V and above 26V, preferably with a bit of padding, no matter what load i draw from it. In the schematic above I added some capacitance to hopefully smooth it up a bit and keep it a bit more stable. I will test this tonight in the lab. I did get the disturbingly low reading of 25.6V (ignore my stupid multimeter it sometimes forgets decimal points) which will definitely need investigation as this is below my absolute minimum of 26V.


Another problem, that I just noticed in fixing up the schematic to post on here, is I probably used the wrong transistor. On it, and from examples I had used to design it, I indicate an NPN darlington to be used and I had probably mistakenly used a PNP one. This worked just fine but I might investigate while i’m down there to see if indeed I did indeed use the TIP147 instead of the TIP142 and what, if any, effect swapping them would do.


Well I just took a poke on the bench and I was indeed using the TIP142 NPN darlington like I was supposed to. I still need to investigate why the voltage dipped and if I can repeat that and take some careful measurements. I understand how to use the darlington as a current regulator, and the dip in voltage would suggest it’s limiting the current (which I do not want it to do at this stage). It makes a basic sort of sense by the 30V zener would net roughly 30V on the output (I guess) but I need to know the why and specifically the calculations involved. CircuitLab showed me that I would get around about 30V regardless of current draw, why this real-life dip I haven’t a clue – yet. I’ll try and repeat the experiment and isolate the conditions under which the voltage dips. I’ll try various other loads too to see if it goes outside the usable window. More to come.

Caveman Dummy Load

Okay, so I thought I’d start with the fun post :). For the truly nerdly, electronics is always entertaining, even if you can’t directly see what’s going on. For everyone, however, the fun bit comes in with a bit of mad science. We want fireworks, smoke, sparks, and flames. So long as we don’t get hurt our burn our house down, we can enjoy a bit of drama and cackle evilly as we dump 3A through something and feel the power coursing through it.

I mentioned in the last post about the need for an immediate, low-tech dummy load that doesn’t involve me sourcing parts, puzzling over schematics, waiting for online deliveries, or rushing to the shops to grab that one part I’m missing (I’m always missing one!). With that in mind I had a problem with my power supply project. I’m busy designing away the various modules, simulating what would happen and punching the math to make sure it doesn’t blow up on me. I need to test the damn thing as I build and although I can simply check the output voltage of each stage I prototype up, it doesn’t tell me anything of what I will do when i put a load on it. Will it be stable? Will it melt into a pool of toxic goo? The only way to know for sure is to find out. I take a calculated risk as everyone does quite literally by simulating my circuits but it doesn’t take into account the real world of component tolerance and the million weird and wonderful little variables that are assumed not to exist in mathematical simulations. Besides, it’s more fun to build stuff up and see it in action :).

My inquisitive searches popped up a number of interesting ideas that don’t involve me making yet another complex project like the electronic dummy load I want to build. I needed something simpler and nothing can be simpler in electronics than a resistor. I mentioned in my last status post that the logical dummy load is a resistor, but of course finding one accurate enough that can dissipate the required power without melting has been challenging. Most power resistors have really wide tolerances, are bloody expensive, and not so easy to find. Keeping a stock of all the required values would then be a challenge even if I could find the right ones. So, like any enterprising maker, I squared my shoulders and proclaimed “I shall make my own!”.

What is a resistor anyway

First, I needed to examine what a resistor is before I could actually build one. Put simply, it is nothing more than a length of wire trimmed to a known resistance. All wires have resistance although it’s usually assumed to be negligible and with good reason – it usually is. This is because the highly conductive copper we usually use for wires are designed to have very low resistance so we get our signals through and don’t waste so much power in our projects. A wire’s resistance is a function of it’s material’s resistivity (a constant that is different for every conductive metal, copper is 1.7 x 10^-8 Ω m by the way) multiplied by it’s length, and all divided by the cross-sectional area of the conductor in question. So it’s easy to see from that statement that the length of the conductor increases the resistance, and the cross-sectional area decreases it. So the thicker the conductor, the lower the resistance, the longer the the conductor the higher.

It seems simple now, doesn’t it? just make it thin and cut it to the desired resistance, done. One problem – the current it’s resisting has to go somewhere, the laws of physics prevent energy from just ceasing to exist. As expected, it’s dissipated as heat. With enough current flowing through it that could be disastrous. Not only would it heat anything touching it, but could melt itself making is a very dangerous thing from both an electrical and “burn down your house” standpoint.

Copper, having such a low resistivity, is hardly ideal to use for a resistor. I would need many kilometres of the stuff to get what I want and/or have it so fine a gauge that it would melt from the current I’m intending to dump through it. As I contemplate this, I toot on my ecig and it hits me – the kanthal wire in the coil could work perfectly. I check a packet of the coil wire I have and sure enough it says clearly 18Ω/m. Perfect.

I laugh to think – all a resistor is a heater.

The build

Seems simple enough, I need a precise 10Ω resistor, I have a meter of kanthal which I know is 18Ω/m, and I know it will take a few amps dumped through it since I do this many times a day with my ecig. All I have to do is cut it and test, cut it and test until I hit 10Ω. Then I can mount it on… what to mount it on… oh jeez. Here I have a length of wire, that’s going to get really hot if I dump more than an amp through it (I intend to dump three) and I have to have some way of holding it down. A live wire with enough current to kill me and enough heat to burn me (and anything around it) very badly is never a great combination. I need something non-conductive yet heat resistant. I look about and grab a pencil, maybe this could work?


As you can see above it did work after a fashion. It measured 10Ω, I was able to put lower currents (<1A) through it without a problem. The above happened when I tried 1.33A from my little power supply and the pencil predictably started to burn. Yeah, that was a dumb idea. Also, my house now reeks of burning pencil.

I rummaged about for anything else I could find that might hold this hot potato and came up with nothing. I was hoping some sort of ceramic something might be floating about somewhere but no dice. In rummaging around I found a part that a friend gave me ages ago. It was a piece of a heater from an old car which is nothing more than a piece of plastic, some sort of heat resistant material and some brass strips. The nichrome heating coils were still on it so I cut them off and fitted on my coiled kanthal wire. Bingo, I had a solution.

In action

With my homemade resistor somewhat safely mounted in something that probably wouldn’t fly apart and melt, it was time for the real deal. Time to use it as my test load. I chose 10Ω as my intention was to test my newly prototyped pre-regulator with it and powers of ten make calculations really easy to do in my head. 30V over 10Ω is 3A or my target max current for my power supply. I will detail the results of the pre-regulator test in a following post about it specifically but the short of it is IT WORKS. Gwahahaha

Pretty isn’t it? that’s 3A of current running through it thereabouts and 30V across it. It was a lovely glow and did indeed give off a lot of heat. My chilly basement lab was quite cozy :). The pre-regulator played nice too and gave me more or less expected results



It just goes to show that even complex problems can still have caveman DIY solutions in this modern age. I built me a dummy load with no controls, no readouts, no fuss, and no real expense. I will still build my electronic dummy load of course, since I do need a constant current sink that is finely adjustable for more accurate measurements and to calibrate my power supply.

Return, re-education, progress

So, after a long hiatus I am back in the lab and back to try and finish at least one damn project before I die. Started innocently enough, was bored on an evening and not knowing what to do with myself I thought I’d catch up on some of Dave Jone’s excellent video blog entries. Naturally, his energy and charisma stoked the flames of interest and had me missing the heady smell of flux and the hum of an energized transformer. I looked over what I had on the go and found with delight that I actually remembered a lot and was able to get back to where I was intuitively in no time at all (read: a couple of weeks of review).


Here I’ll summarize the projects I have on the go before I get to the good stuff – the experiments.

Nixie Clock

Almost finished. Kinda. Sorta. The schematic I reviewed for any errors and made sure I got everything right – no problem. I gave the half-finished board a good look-over and found it just fine, needing only to be populated. I know for certain I will be needing a second board on top of it, which is fine, to house the ridiculous number of high voltage transistors as well as the remainder of the 4017 counter ICs. I mapped out a plan of how i’m going to do my interconnects as well and puzzled over the problem of having something like 30 connections going from one board to the other. This is what happens when you use data that isn’t multiplexed and no microcontroller, you basically end up with a lot of wiring. For the power connectors and a few other things I don’t mind using molex style PCB connectors but I was faced with a challenge of how to route the 16 connector outputs from the bottom board 4017s to the top. I briefly flirted with the idea of keeping them all on the same board but then I would end up with 28 odd connections I would have to take to the second instead of 16. So with that in mind it dawned on me that the old IDC connector and the ribbon cable are ideal for this application! I’ll pick some up at the shops when I finally make it down there. As always, I’m s a few parts short on every project so it’s a worthwhile journey. All this one needs apart from this is a few switches to set the time some way to mount the nixie tubes securely and safely and of course a box to put it all in.

Power Supply

The power supply is, and always was, a beast of a project. Some do their first power supply simply but I wanted something flexible, cool, powerful, and more or less something I will want to use years down the road. The inevitable revisions take forever and the whole project is quite complex by this point. As I’ve mentioned before I’ve broken it into modules to make it easier on myself: AC Power, Pre-regulator, Current Limiter, Voltage Regulator, Control and display.

The AC Power board, as shown from previous posts, is complete and overall I’m quite pleased with it. It does have a rather high DC voltage output of ±42.6V which necessitates the inclusion of the pre-regulator block (mentioned previously) but otherwise performs just fine and will take more than I could possibly throw at it before it dies. Subsequent video watching and research has made me want to add the rather important addition of MOVs (Metal-Oxide Varistors) to add some over-voltage protection to the AC input of the supply though this is a rather trivial addition and simply have to add one each between the hot-ground, neutral-ground, and hot-neutral. I will probably add them to the rectification board or may have them on a separate board or hanging off the terminal block, I have not yet decided. Over-current protection is already present in the 5A hot fuse in the terminal block, as well as redundant case fuses I will employ in the final build. There is also the question of the ground lift. Research has shown me the wisdom of not tying the centre tap of the transformer (the 0V rail) to mains earth so with that in mind I will keep it floating by default with an aircraft switch on the front to enable mains earth referencing should I need it. Easy.

The pre-regulator has seen much progress since the last time I took a poke at it. I had previously included it, discarded it, then included it again in a much more workable form. After much fiddling in circuit lab, I settled on using a zener diode/darlington transistor regulator and crunched all the numbers into a workable solution. I did build it up and test it but will expound on my results in a subsequent blog post. The upshot is I can knock that crazy 42.6V down to much more usable 28-29V and have it work over a variety of loads which is nice. There still is the question of stability and whether or not it will play nice with the rest of the circuit.

The other modules are untouched from last check. It will be a long time before a completed product. I’m still fuzzy on a bunch of things and I’m expecting pitfalls along the way which could be both frustrating and highly amusing. Of course, that is why i’m doing this in the first place – I’m learning, and that is its own reward.

Dummy Load

Of course this ties in closely with the power supply project as I need some practical way to test the thing under working conditions not to mention calibrate it. It has taken a back seat to other projects yet I will have to build it to build my power supply. Projects always lead to more projects. It all started, as mentioned previously, with Dave Jone’s excellent example, but I’ve been further spurred on by the discovery of Martin Lorton’s excellent version which will probably be much more suited to my needs. Naturally, I will add my own modifications to both make it my own and to suit my needs. Martin’s has a 2A cap (I believe he reduced it to 1500mA by the end though) and I need mine to sink 3A to properly test my power supply. It should be a simple matter of selecting the right mosfet and/or using mosfets in parallel.

Given that this project is likely to take me as long as the others and I still have not been able to leave my house to grab the appropriate parts, I went back to why I need this damn thing in the first place. The easiest solution for a dummy load is to of course whack the right value resistor that can handle the power you want to dump through it. This has proven most frustrating since not only do I not have a collection of power resistors on hand, but finding ones with the appropriate tolerance and power dissipation capability has proven to be difficult.

I decided for a more low-tech approach and see if I could make my own power resistor to act as a static dummy load just for now. That is a subject of the next blog post which I will marry with the power supply pre-regulator test. Quick answer: I did, and it not only worked but I’m still alive and my house is still standing.

New Projects

Always something new an shinier on the horizon, isn’t there? This is why I never get anything done.

Milliohm meter

Keeping with the theme that projects beget other projects, the power supply needs a dummy load, the dummy load needs a precise high power low-ohm resistor, I need a way to measure low resistances. It’s commonly known that most DMMs do a woefully awful job of measuring low resistances. One has to dump enough power into it to see a measurable result, and things like the leads now have a nontrivial resistance. So I started building an adaptor for my multimeter that fixes these problems. A post will be written on this also. No, I didn’t finish this either.

eGo Charger

This is merely an idea and a helpful schematic posted by someone on a forum somewhere. One of my eGo chargers for my ecig is malfunctioning and not doing it’s job and I sit here rather nervously waiting for my one functioning one to die and deny me my fix. I’m merely thinking about this one for now, if I get it wrong i’ll have explody batteries on my hands so you can bet I’m going about this one carefully.

Dummy loads

The need to build everything to have something to calibrate against

In my testing of the pre-regulator, it quickly became obvious that stocking a whole bunch of high power resistors just isn’t feasible. In order for that to work, I would have to have on hand a variety of high power resistors to suit various voltages and currents. In addition to that being a pain, it’s also very hard (and expensive) to find accurate resistors at high power and having to acquire a collection of them is just silly. Since my power supply must be capable of delivering up to 3A at 24Vdc, that means a potential power dissipation of 72 watts! Any resistor I find for that specification (which would be 8Ω) would essentially be a giant heating element and probably have a wide tolerance.

The best solution is to build yet-another piece of test gear: the dummy load. I mentioned before the ever-great Dave Jones of EEVBlog fame has a nice quick video on how to build one. Though this appears to work great, the specifications are a bit wimpy for what I need for this application, and limits possible future applications as well. Dave’s design seems to max out at about 1.335A and I would need more than double that to test my power supply. It also marks the re-appearance of another hassle which has plagued me in current measurements, and that’s finding an adequate shunt resistor(s). Dave’s design uses 10x 10Ω 1% resistors which certainly keeps the math easy, but I’m finding it really hard to find any of them at my local electronics shops! I’ve found 1Ω ones, 0.1Ω ones, but no 10Ω ones to parallel up to make 1Ω. Bummer. This is a job for DigiKey for sure.

Higher power design

Naturally, needing something with a bit beefier spec, I hunt through google looking for similar but higher power capability. I found a really nice one done by Paul Renato and he’s used a similar design to Dave’s but has improved on it quite a bit. For one, he’s upped the current sinking capacity to 7A (I had arbitrarily picked 5A for my needs so this is ample). He’s also made use of the two unused opamps in the LM324 to provide some overload and thermal protection. There are a couple of things I’m not clear on, namely how to select the proper MOV, thermistor, and schottky diode from his schematic. I may have to write him on it. Beauty of his design is he’s already gone to the trouble to whack in as much functionality as possible, which is usually something I do when presented with a project schematic. I can’t think of anything I would add to this one. Brilliant.

Here’s the full article and schematic

So yeah, more parts to acquire and more projects to build. This one definitely looks like a winner and not terribly complicated either. I need to go shopping for MOSFETs and some other choice bits. Ah damn those current shunts! I’ll get those too finally.


The diode close to the power supply appears to be a zener, not a schottky (got me symbols mixed), though why that’s there is a bit of a mystery to me. It could be regulating the 12V input since his photos show him using a wall wart to power the thing. It’s the only imaginable purpose I can come up with. The 10:1 voltage divider used to set the current seems to me to have a max of 1.2V in this case (setting 12A) though he indicates 0.7V (7A) should be max. I assume the overload circuitry is designed to catch the over-current, but I may just tweak the input to be a little friendlier as well as substituting the zener with a proper voltage regulator.

Zener Pre-Regulator Build-test

The magic smoke appears but fails to ruin the day!

So after a couple of minutes getting all the bits on the breadboard and blowing the breaker once through my own idiocy, I managed to get the thing together. The results are, well, exactly as expected. In fact, I got a steady 30.0V with no load. I was expecting around 29.3 due to drop-out from the transistor but it makes zero difference.

28 and something volts while pulling 1.3A, not bad. The value kept climbing also.

28 and something volts while pulling 1.3A, not bad. The value kept climbing also.

I left it on, and let it run and no problems whatsoever.

I decided to try a dummy load to see if it catches on fire or melts or something. The only low-value high power resistor I had that wasn’t 1Ω or less was a 22Ω 10W one so I figure I’d give it a try. It ran and did it’s thing for a bit until I saw smoke escape and quickly shut it down. A few more (very careful) power up tests revealed it was the resistor that was smoking! A quick calculation revealed that it was dissipating on the order of 40.9W! Yeeouch that’s hot! I recorded a temperature of 150°C on the thing.

Close up of my pre-regulator

Close up of my pre-regulator

The other parts faired well, the zener and biasing resistor held up fine, though the darlington transistor heated up like crazy. Still not in danger of melting it. It’s designed to run up to 150°C and needless to say, I will have a giant heat sink on it in addition to forced-air cooling.

I count this a success.

Room for improvement

Though it was easy to see that it “worked”, what’s more difficult to tell was how it works over time without that 22Ω resistor becoming lava on my breadboard. Because of this, I was only able to keep it powered up for less than a minute at a time while I feverishly took measurements. I would have ideally liked to be able to safely set the load and forget about it melting on me while I conduct voltage, current, and temperature measurements over time.

It is apparent, I neat an adjustable dummy load. It is the only way I can accurately test and calibrate my power supply without melting anything. Fortunately, Dave Jones did a video on it and I hope to build one soon myself. Valuable piece of kit that. Dave, as usual, is a life-saver.

Zener pre-regulator

The need to drop some volts

I mentioned in a couple of previous posts that my giant toroid transformer that I want to use for my power supply was a little too beefy for my purposes. I mentioned in the most recent blog post that transformers are often rated at just below normal mains voltage to provide a “guaranteed minimum” and that once you combine this fact with your rectifier and filter, you actually end up with a higher voltage than you originally intended. Most of the time, this is great for ensuring you can overcome that pesky dropout voltage (around about 2.5V for most regulators to be on the safe side) but in some cases, it can get dangerously close to frying said regulators by exceeding the max input voltage (maximum voltage differential).

In my case, the chunky toroid, after rectification and filtering with no load gives me a rather beefy 42.6V when on paper it should have been 30V. Given that most common regulators have a maximum input voltage of 35-38V, it becomes obvious that I had to step it down a tad. There were a few options to do this:

  • Unwind the secondary of the transformer a few winds
  • Use series diodes to drop the voltage by 0.7V each
  • Construct a pre-regulator to drop the voltage to safe levels

I have examined each of these in turn and came to to the following conclusions:

Thing of beauty, don't want to wreck it

Thing of beauty, don’t want to wreck it

The toroid is a lovingly packaged beastie, professionally wound in crazy spirals, wrapped in plastic and nicely presented with leads. Since I barely know what I’m doing, it would be unwise in the extreme to mess with it. Why break something to make it work when other solutions can work just as well?

Isn't that gross looking? Embarrassed to say I made that...

Isn’t that gross looking? Embarrassed to say I made that…

I did try the series diodes, and was able to get a voltage drop of 4.2V by bodge soldering 6 of them in series. As you can see from the photo, there are quite ugly and apart from appearances do not reassure me that they will hold together well and 4.2V is not nearly enough of a drop to be useful to me. Essentially, I made a mess of 6 diodes to only get down to 38.4V, barely outside the red danger zone. Also, the legs on the high-power diodes are quite chunky and would be a severe pain in the ass to get into a pre-drilled circuit board and I will not have them floating in the air inside the case in case heat/cracked joint/whatever causes them to fall apart starting a fire. Also, when I’m long dead and someone opens the case to see what I’ve built, do I really want them to see that bodgy mess? No way man.

The third option is one I considered before and after thrashing about with the other ideas, it has been revealed to be the best one. Difficulty is, with voltage regulators having the limitation mentioned above I couldn’t use them. As mentioned in a previous post, I am going to go with a simple, yet high powered idea of using a darlington transistor biased with a zener diode. I was able to acquire some 30V zeners (part: 1N4751A) so these two, plus a couple of biasing resistors is all I need.

Here, I ran up a simulation to illustrate:

Pre-regulator test circuit

Please ignore the 2N3904/3906 transistors, they are mere placeholders for the TIP142/147 darlingtons I intend to use since CircuitLab didn’t have either in their box of parts. Only me building it up would determine whether or not I have to adjust any of these values/parts or not. I like living dangerously.

I chose a 10Ω load resistor to simulate a current output of 3A at around 30V. The actual output voltage doesn’t need to be precise as it will be further regulated later, so long as it’s above 26.5V or so it will be fine.

The tricky bit is biasing the thing. I first assumed that I could use a 10kΩ resistor to power the zener and bias the transistor. I quickly figured out that this isn’t enough, especially when under load. The zener needs a minimum of 5mA to get that nice voltage drop going on and the transistor needs a proper current to regulate the voltage.

Through fiddling about, I settled on what I think is a good compromise. My zeners are 1W so there’s one limitation I have to consider. Though I am well within the max collector/base voltage limit of 100V, the max base current is 500mA, which will be hard to come close to, but worth noting.

Through messing about with the load, I noticed that it had a hard time regulating itself at higher currents and I kept having to drop the biasing resistance to keep it working. This presented a problem of higher currents both for the base and the zener. Though the base can handle it and then some, the zeners were fast approaching their power limit.

The compromise is the 470Ω biasing resistor. It seems, at that level, to keep it’s regulation up to 3A more or less which is what I chose as my maximum current output anyway. Any higher than that my current limiter will drop the voltage to keep it from melting itself.

This also keeps the zener/base current steady at 26.84mA. Though this seems like peanuts, given the voltage drop across the zener the power dissipation would be about 800mW or 80% of it’s maximum.

Though I realize some of that current is going into the base of the darlington transistor, I really don’t want to try lowering the resistance any further. Always better to have some headroom in case Murphy and his law show up. Ohm, and his law, also indicate that the resistor will dissipate a third of a Watt (0.338 to be exact) so 1W 470Ω resistors would be nice to have. As it turns out, I have one.

If I need to get higher power parts, this is also possible. A 30V Zener diode 5W (part: 1N5363B) exists and is cheap, and i’d nead a helluva resistor to go with it, probably a 10W 100Ω. I think I can avoid doing that though.

In a few minutes, I’ll head down to the lab and build it up.

Transformer wiring

So I got my new 2x6V transformer to the lab finally and hooked it up. As expected from a wound coil of wires it works perfectly. I was, however, briefly mystified by the plethora of wires emanating from this chunky beastie and wondered why the hell the secondaries were reading half a volt AC when it should have been much higher.

Transformer wiring can be confusing to the novice (or drinker) and generally you don’t want to mess it up like I did once by plugging the (low resistance) secondaries into the mains blowing the breaker and marring the plug where it melted due to arcing.

Like many transformers, this one was manufactured with two sets of primaries and two secondaries, allowing for a variety of connections. The two primaries are meant to select between US and Euro mains (~120V and ~240V respectively), and the two secondaries meant to deliver the stepped-down voltage to be filtered and regulated.

Each set is nothing more than two terminals and a coil of wire wound around an iron core – an inductor. When current is run through the coil, it generates a magnetic field that inducts a current on the secondary side converting it back into electricity. The voltage on the secondary is function of the primary voltage and the ratio of the number of turns between the primary and secondary.

When presented with two sets of primaries, it’s a bit confusing to figure out how to wire it into a two wire plug from four terminals. The best way I find is to think of it like batteries (though the same rules hold true anywhere in electronics): voltages in series add and currents in parallel add.

Practically, this means that if you wire the two primaries in parallel you can double the current output of the transformer, and if you wire it in series you double the voltage and halve the current.

This particular transformer is meant to convert 117V to 6V, and has two sets of these. Coming from it are 8 wires, four primary and four secondary (two for each coil). This is quite customizable and could be wired in a variety of ways depending on your application.

For 240V euro mains, I would wire the primaries in series, essentially combining the two coils into one allowing a greater step down to still get the desired secondary voltage. For North American mains I would wire them in parallel to keep it at 120V.

On the secondary side, I could keep the two coils independent and get 6V each, or wire them in series to get 12V, or wire them in parallel to get 6V at double the current output.

Here’s where the fun comes in: if I were to wire the primaries in series (as for euro) and keep the secondaries in series, what would I get? Instead of the expected 12V, I would get 6V because the number of effective winds between the primary and secondary is doubled.

What I ended up doing is wiring the primaries in parallel (making two 120V primaries) and wiring the secondaries in series (making a 12V secondary).

What I am glad to note is that many transformers are rated at just below the typical mains voltage on the primary side, so when it receives a typical mains voltage, the secondary voltage is likewise a bit higher. This is helpful considering voltage regulators have a dropout voltage that must be overcome to regulate the desired voltage and takes care of any irregularities or variances in your home mains voltage.

Couple this with the fact that when you rectify and filter (using a bridge rectifier and capacitor) the secondary AC voltage, you end up bumping it up a bit so you will always get higher than the expected voltage which (hopefully) overcomes the dropout voltage of your regulator.

In my case, I wired it so that the 117V -> 12V transformer, when fed my home 120V mains, then rectified and filtered with a 1000µF cap, yielded 18Vdc and change. Wow, half-again my expected voltage! This is hardly a problem and will work great for my application.

Just keep in mind two things:

  • Your rectified and filtered output will be a higher voltage than expected
  • Wire your transformer wrong and you will, at best, get the wrong voltage and at worst, a blown breaker

More goodies

Felt I needed a shopping trip, just because I could. If there is one thing about the electronics hobby you should know, is that you never have enough parts. This time I visited Supremetronic on college street again, if it is still called that since it merged with home hardware.

Though I realize I could probably order the parts online and not have to go anywhere, I brick and mortar shop for two reasons. First, I want to support the few remaining shops still in business. They are having a hard time competing with online retailers. Second, I like the experience. Sometimes it’s damn hard to tell what I’m buying online and it helps to logically find stuff in bins of parts. It’s important also to keep through-hole components and by extension, hobby electronics alive. Radio Shack has become a pale shadow of what it once was, nothing more than a gadget retailer now. Especially in Canada where it has become “The Source” which is just laughable and has nothing I want for too much money. I do not shop there.

Anyway, I was able to get many parts I was missing given the revisions going on in my power supply design as well as my nixie tube clock project which has been on the back burner.

I picked up a 2x6V 12VA transformer, which means I can get 1x12V out of it at 1A. Just what I needed. Smaller and lighter than the 24V ones I was messing with since they are no longer necessary. This will power my relays and output board switching as well as my meters, keeping it nice and independent from my main power board.

As I mentioned before a couple of posts ago, I’m moving the voltage and current reference settings to the main board so that they actually have a common reference. To that end, I was able to find some zener diodes that work perfectly. Picked up 8 or so 30V zeners (didn’t know they existed), a bunch of 12V ones and some 5.1 (i think) volt zeners. This should give me rough voltages to start from and regulate for the references.

As a bonus, I figured out I can use another darlington transistor biased with a zener at it’s base to drop the voltage from the main rectifier (currently at 42.6V) to just shy of 30V so I solved the problem of the voltage being too high as well. This will save me a lot of heat and keep me well away from the voltage differential maximums for my regulators. This now means that I have to use three pairs of darlington transistors: first, as a pre-regulator, second as the current regulator, and third as a pass transistor for the voltage regulator. This seems complicated to me (as I am sure it does to you) but I really want to keep these functions separate if possible. I’m sure I could combine the first two just fine, but I want both to function 100% independently just in case. This is a learning process, so things may change when I breadboard it up and see if it catches on fire.

So I’m homing in on my final design. It’s a bit sad that my control voltage board basically has nothing on it anymore apart from power to the relays and the regulators for the meter supply. I may merge this with the output control board just to save boards. It also occurs to me that I may not have enough headroom on the new transformer for a 7812 regulator to give my output board a stable 12V. I do believe after rectification and smoothing with a 1000µF cap, I should have plenty to overcome the dropout voltage of the 7812.

In addition, I also picked up a few parts I was running low on. Notably, a bunch of dual and quad opamps like the TL084, LM3900 (which I heard was discontinued yet still available), and the NE5532 (one of which I fried by accidentally using it in place of a 555 timer for the nixie supply). Also picked up a bunch of 555 timers and 556 dual timers for various applications. I was thinking of improving my soft start circuit with a more accurate (and more customizable) timer using 555s though this remains to be seen. I may revert to the earlier design anyway. I still haven’t figured out to create a latching cutoff condition during current overload without putting the sense after the relay. A minor, but important detail.

the nixie clock now will also have MPSA42 300V switching transistors for it’s digits. I’ll write on this in a separate post.

Voltage, it’s relative

So yeah, playing about with my new transformers, everything working fine. Until I realized one tiny hiccup. When I switched the control voltage board to it’s own transformer for the sake of the meters, I thought it a good idea to keep the circuit as is and have independent references for Vset and Iset. Though in theory this sounds like a good idea, I had one of those face-palm epiphanies. With two isolated power supplies, they have no common voltage reference. If I try to inject a voltage on the set pin of an LM317 and use the power supply ground as a ground references I get: 0V. It’s interesting to think about what this means. With no common reference, the potential is, effectively, zero. Hilarious! Also, utterly useless for my application.

Fixing this minor issue is quite easy. I have two possible options:

  • connect Vset and Iset to the set pin of the LM317 and the input on the current comparator. I am not sure at all this will work. In fact, I’m pretty sure it won’t.
  • move the 0-24V Vset and 0-3V Iset back to the main power board so they share a common reference. This seems more likely to work.

What a learning experience! When I simulated it up in CircuitLab, it was assumed that ground is ground. Whenever I define a ground reference it’s universal. Sadly not true for reality. Voltage, by definition, is a relative measurement – the difference in potential between any two points in the circuit. With no common reference, it’s meaningless.

What kind of sucks is now my control voltage board has much less functionality. The two transformers I bought, sadly, not for this project. No matter, their values are common enough that they will find definite use elsewhere. Especially since they are both centre-tapped.

So, what’s left? If I subtract the Vset and Iset from the cv board I am left with the relay control circuits and the power for the meters. Not bad that. I can use another, smaller, transformer rated for 12V. As for the voltage references for the current and voltage setting, I can use either fixed linear regulators or zener diodes (probably the former). Not what I had in mind, but not the end of the world either. At least it stopped me from moving things between the two boards, unsure of where they should be.

Back to the lab…

OK so I’ve enjoyed a bit of a respite from pretty much everything except relaxing and revelry as suits the season. I made sure and visited A1parts the other day and picked up some more parts and bits.

There, I was able to find two used transformers, one 25V CT and one (presumably) 28V one in a cool retro hammond case. I bought two since I figured I couldn’t lose by having an extra transformer about in that range, but also wanted to see how big the voltage jump would be after I rectified and filtered it. It was also a major piece of the puzzle for my control voltage board as it requires a separate supply from the main one to give an independent voltage reference as well as power my meters. The plan today is to test their output voltages and build up the basic control voltage circuit. I can then test the power drain of the 12V loads (relays and indicators) and their effect on the control voltages (hopefully minimal). This is probably the shortest path to finishing the next board in my power supply.

I realize I need to do up a new schematic for the cv board given the modifications I previously mentioned, and I will do that soon. I am just too eager to get my hands dirty and breadboard it up.

Other items I managed to grab while I was at A1:

  • Nice pair of side cutters
  • Panel nibbling tool
  • A collection of smaller gauge marettes
  • Some smaller gauge splice crimps
  • Some 0.1Ω 1% 2W resistors

The last item was a nice find and will be useful for experimentation. I could use them in my final design for the current shunts since the max power dissipated would be 900mW, and I may still, but I think I will shop on digikey for those precision current shunts I mentioned in a previous post. It would cut soldering down from 11 components to 2 components for the ammeter. Still, good enough for testing! I can parallel 10 of them to make 0.01Ω. Shame I couldn’t find the same in 10Ω to make the 1Ω shunt. I suppose I could put 10 of them in series but that’s just silly.