I’m trying to select a power transformer for a very simple low current AC→ DC linear converter. I’m looking at using the Triad Magnetics VPP36-140 with the primaries connected in parallel and the secondaries connected in series. The output after rectifiers and filtering should be above 85 VDC and the load current is constant, somewhere in the 20-35 mA range (final value to be determined). The cartoon block diagram is below.
Basic simulation shows the secondary current pulses to be about 70-72 mA(rms) under load. I want to run without fuses in the secondaries because the high series resistance of low current fuses causes unacceptable voltage drop before the rectifier. I think this transformer will work without secondary fuses but I’m not sure.
Questions:
- Can I safely run a VPP-140 without fuses in the secondary? What if there is a hard short in the load?
- How do I select a current rating for the primary side fuse? I haven’t found any suggestions in the Triad documentation yet.
Thanks in advance!
-Chris
Hello Chris,
Great question - this one has me thinking a bit. The purpose of a fuse on the secondary is just as much to mitigate damage when things do go wrong as much as it is to prevent things from going wrong in the first place. Because of that, I personally would steer away from the thoughts of “can we circumvent safety” (especially when thinking about the thermal considerations of a short) and instead look at the problem as “how can we mitigate the issue of voltage dropping over the high series resistance”. I think it may be possible to try to fuse the center tap of the secondary to minimize the resistance, but I’m going to have to take a longer look and do some more research on this. I’ll reach out to some of my colleagues for input and perhaps they’ll be able to lend a bit more insight on this one as well.
Regards,
Klint
The datasheet contains footnotes that may be found pertinent.
Typical practice for small equipment fed by an AC line is to fuse once at the point of power entry to the enclosure, designing the powered equipment such that it’ll never blow the fuse unless/until something’s gone irretrievably wrong and it’s time to worry about not setting the building on fire.
That sorta depends on how the transformer is designed. Note the following from the datasheet:
For transformers at this small a scale, it’s possible/practical to design them such that they can tolerate a shorted secondary indefinitely. This note would suggest that this isn’t the case for the device in question.
Details will vary from case to case, but overloading a transformer that isn’t inherently short-proof typically causes a rapid temperature rise of windings and/or core, eventually leading to insulation damage that results in an exponential worsening of the problem until the transformer becomes a puddle of molten metal or outside forces intervene to shut the party down.
As far as selecting a primary side fuse is concerned, this post might be informative. Basic idea is to leave a bit of headroom between fuse rating (more or less depending on whose rating scheme is used…) and the worst-case continuous load, then evaluate for any peak/transient conditions that may apply.
A person’s then expected to put everything inside an enclosure capable of containing however much glowy-burny sort of activity the chosen fuse value might allow to occur prior to opening. That’s a bit of an art and science both, and in some part why UL94 ratings came to be.
Rick,
The transformer datasheet is where my confusion starts. The footnotes clearly state that secondary fusing is not required. But the safety note says it’s not inherently short circuit proof. Is the safety note referring to the primary side only? Taking the 2 footnotes and the safety note together say to me that secondary fuses are NOT needed but a primary fuse MUST be used. Is this the correct interpretation?
In the case of a primary fault the primary fuse will open. No problem. And any secondary current should be reflected to the primary. So in the case of a secondary fault the primary should see on the order of 1/3 the fault current (neglecting losses). A properly sized primary fuse should also take care of this fault.
Is my reasoning correct? This is inline with your comment regarding standard practice of placing a single fuse at the point of power entry which is also my experience on using larger power transformers with single secondaries.
I’m thinking out loud below. TLDR: Should I just use a 150mA Slo-Blo fuse and be done?
In my application, with nominal secondary current pulses of around +/- 150mA(peak) and 70mA(rms) the primary current should be roughly +/- 50mA peak and 23mA (rms). These are well below the transformer’s ratings. I would think that protecting the primary for the transformer’s max rated primary current would provide the maximum protection against nuisance trips and still be absolutely 100% safe.
But I don’t have enough information about the transformer’s primary or efficiency to know that value. Should I use the VA rating directly? Or should I take into account the the transformer’s efficiency which would result in a higher primary current.
Neglecting efficiency, the max primary current should be 43mA at full load. Assuming maybe 80% efficiency (a guess) the full load primary current should be about 54mA. Either way, my application current would be well below either of those numbers. Should I just fuse the primary assuming the smaller (safer) value of 43mA? Using a typical fudge factor of 300% for currents < 2A would result in a 130mA fuse rating. I would then select the next highest standard value, say Littelfuse 313 Slo-Blo series, @150mA150mA.
BTW, the linked post is useful for selecting a fuse if you know the design’s nominal current limit. In this case I don’t know that. I need more information on the transformer itself. I’ve reached out to Triad directly through their website, but I didn’t receive a response.
Best Regards,
Chris
Klint,
My primary goal is to maintain safety. That’s why I’m asking questions. However, this particular application’s constraints (very low current output) force me to consider options to reduce the series resistance. Other small power transformers that I’ve looked at require secondary fuses. This one explicitly states they are not required.
I’m not sure what you mean by fusing the center tap. How would I do that? Putting a series fuse between the 2 secondary windings is the same as putting it at the output. There is no other way to put a fuse there….there is no connection to the center tap in the application. And regardless, the fuse resistance is still there and it still creates a drop of several volts across the fuse.
Best Regards,
Chris
Hi cwilkson,
I’m not necessarily disagreeing with your thought of only fusing the primary side, as the argument has some merit, but I’m trying to understand the concern about losing a little voltage across the fuse resistance.
As I understand, you’ll be connecting the transformer output in series, for a nominal 36Vac output. Peak rectified output prior to diode drops would be in the range of 50.9V for the transformer’s rated load of 140mA on the output. Since your nominal current will be somewhat less than this, the voltage would be expected to be a bit higher than that. I would expect it to be at least a few volts higher, since your nominal load would be well under 50% of rated load.
According to how I read your application, your nominal load will be somewhere in the 20-35mA range, with transient peaks possibly as high as 150mA. You’ll have two diode drops across the output, so you’ll lose no more than about 1.7V there (probably less, especially if you used Schottky type diodes). This brings the peak down to 49.2V.
You would want to select a fuse which won’t nuisance trip but still protect your system. A slow-blow seems appropriate. As an example, looking at the Littelfuse 218 series 125mA rated 0218.125MXP, the worst-case voltage drop is 1.9V, and would typically be a small fraction of this. That would bring your peak voltage down to 47.3V, a 4% drop. But this would be a worst-case scenario, with typical voltage drop of well under 1V.
With sufficiently-sized caps for C1 and C2, the peak transient currents seen across the fuse and diodes would be minimal, further reducing it’s voltage drop. If they are not sufficiently sized, the ripple will be the dominant factor to your output voltage. You have not stated your application, but this minimal voltage drop due to a secondary-side fuse seems negligible to me, so I’m not clear on why this is of concern to you.
Hi David,
I made a mistake in the peak secondary current. It looks more like 235mA peak transformer current at 35mA load current. So I need to consider series voltage drops at 235mA. Of course all of these numbers are rough estimates at this point with no design margin built in.
You’re correct. The transformer’s peak output will indeed be considerably higher than +/- 50.9V….I’m counting on that. Unloaded it should be almost 64V. But that will fall a lot under the non-linear secondary current. With a 235mA output secondary current pulse it should be about 48 V peak. The rectifier+filter doubles the DC output voltage so I would expect aroudn 98V VDC across the output caps. The (silicon) diodes drop about 1.45V and with 85uF caps (100uF nominal) the ripple across the 2x series caps is about 4V. That leaves me with a minimum output voltage of 92-93 VDC.
At 35mA, I need the supply to be 87V minimum. As you can see I have very little head room. Any variation in the capacitors, the transformer, or the line voltage will cut into this margin. This is why any voltage drop across a secondary fuse is painful.
I was going with the 313 series fuses because they are already used in some legacy designs. The 218 series does have a significantly lower series resistance. Thanks for that suggestion! I will look into making a change there if I can. Still, that’s 1V of drop at 235mA. I would like to eliminate the secondary fuses altogether. It makes life much simpler all around with better output voltage margin, reduced BOM cost, smaller PCB, easier inventory management, lower field maintenence cost, etc, etc.
At this point, the main question is what is the manufacturers recommendation for fusing the primary? Ragardless of whether the secondary is fused or unfused, that needs to be solved.
Best Regards,
Chris
It’s saying that no output fusing is required to meet the voltage/current/power limitations that apply to class 2 or 3 transformers. This link may offer useful info on the concept.
Any appliance that plugs into an AC line source is supposed to have over-current protection capable of interrupting the load before the current rating of the cordage supplying the device is exceeded. The basic idea (looking in either direction) involves limiting the potential for creating a hazardous condition in event of a failure.
Fusing is typically done once at the power entry point for the entire apparatus in which the transformer is installed, and since the transformer may or may not be the only device drawing from the line source, no such recommendation is offered.
The upper end of an allowable fuse value is dictated by the cordage/connectors used to connect the appliance to the AC source, while the lower limit is imposed by the appliance’s expected current draw. Depending on whether one chooses an IEC or UL-spec fuse ~50mA would be a minimum based on the VA rating if used alone, though a higher value would be recommended to avoid nuisance tripping caused by fatigue from turn-on transients that will attend the charging of the output capacitance. Lacking a detailed electrical model, it’s the sort of thing a person’s likely to address using a guesstimate-and-verify sort of process. There’s not much downside to starting with larger values, save for a possible impact on enclosure design.
