• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Dead short fuse trip experiment using an MRBF fuse

I worked on remodeling a lot of old movie theatres back in the day. Some were even old playhouses that had a full stage. The stage lighting systems in some of them had about 20 to 50 variacs connected to the lighting sockets all around the room and on the electric stage rails. I think they even had a sort of high power patch bay for choosing which sockets were on which variac. The lighting control panel had large levers to adjust them. The were mechanically linked to turn the variacs with chains. That all worked fine with a large crew running a live show. But when they ran movies, they didn't want to pay a stage hand or electrician. So there was typically one or two larger variacs located up in the projection room area. Early movies still had 2 people in the booth, so one could dim the lights as the other rolled the show. Later ones had a motor so the one projectionist could just hit the dim button and the lights would slowly dim down.

One thing I learned very early with variacs, is that the NEED to be fused on the OUTPUT side. Many of them were only fused on the input. 20 amps going into a 2,000 watt rated variac seems ok. The problem came up when it was dimmed way down. We turned on the power and no house lights came on. We pulled a couple bulbs and the filaments were good. Even tested one on 120 AC direct and it lit up. Checked the variac output, it was putting out about 10 volts. Maybe just too low to light them. We connected a bulb at the variac, and it lit, dim, but it was on. We turned it up to 20 volts. Bulb in the booth got fairly bright now. And then smoke stared coming out of the conduit. Since the input was only fused at 20 amps, they ran all #12 AWG wire. But with the dimmer variac dialed down to just 20 volts, that input fuse would not blow until the output side was cranking out 120 amps. 1/6th of the voltage means 6 times the current. Lucky for us, the wire from the variac went up to a junction box where it split to 6 feeds, so those didn't burn, only the 20 foot run to the box. An we caught it pretty quick. There were dead shorts on 3 of the runs into the theatre. We found a bunch of broken bulbs and the wires were bent together. At 10 volts, the wire resistance was enough that it didn't make any smoke. At 20 volts, things were getting bad fast. Had we dialed it up to the full 120 volts, it would have hopefully popped the fuse fast. If it was fused at 20 amps on the output, it would have tripped the second we turned on the power at just 10 volts.

I found this image
 
Last edited:
I was in drama club in college in the early 90's. The lighting guy was having issues with the lighting (he was an ME) so asked the EE power prof to come take a look. He condemned the whole system, probably due to similar issues. I was involved only at the tail end helping to rewire. My closest contact with the system was crawling through the attic because one scene had 5 of us climbing down a rope ladder from above.
 
Would you be able to test shorting with a MRBF fuse with just one of those batteries? 12V setup.

Wondering if a MRBF fuse would be sufficient in such a setup or if it still causes molten copper to blow out (boat install).
 
Johan said:
Thanks everybody a lot for the input. Feature requests so far:
  • Create current versus time plot, e.g. using an oscilloscope / clamp meter with inrush (the latter has too low max current).
  • Test more fuse types: Class T (example); bs88; NH; 300A MEGA; CNL; South Bend.
  • Test breakers - cheap amazon etc. Install in wrong polarity too.
  • Test 3D printed shields.
  • Add BMS to circuit, e.g. with serverrack form factor battery, optionally in series with overcurrent device(s).
  • Add inductor coil in circuit as soft start. How to best do this?
  • Test 12V and 24V in addition of 48V.
  • Fundraise and outsource to pros.
  • Apply controlled constant current. Generate DC time-current "trip-curves" for a set of given constant currents.
  • Add wire loop away from rest to cut, cut with better cutter, have disconnect bolt and blade style disconnect
Requesters: @MattiFin @time2roll @HRTKD @740GLE @OzSolar @GXMnow @douglasheld @G8trwood @Hedges @Pi Curio @gotbeans @Skypower @Sennen

Update based on requests - let's start with the first one

Goal: Create a measurement setup for measuring the current versus time, including the maximum short circuit current and the fuse clearing time.
Test "fuse": 22 AWG | copper | 25mm length, see Figure 1.
Experimental setup: Roughly the same as previously, but now expanded with data acquisition: Victron 500A SmartShunt (only using shunt part, not using the Victron measurement circuitry) → Arduino UNO → USB → laptop → Putty → *.txt → Excel. see Figure 2.
Result: The maximum current seems to be roughly 8kA* and the "fuse" clears in roughly half a millisecond as shown in Figure 4. The maximum sampling rate I could squeeze out of the Arduino so far is roughly 7 kHz. To see what happens within the first 0.138ms, faster hardware is required. Also there seemed to be random noise equivalent to 120A. The contactor did not weld closed this time (the same contacts were used as before, filed down a bit after breaking loose).
Conclusion: For data acquisition, the arduino works, but the sampling interval is close to the fuse clearing time. For a higher sampling rate, a better solution like an oscilloscope is probably preferred over the Arduino.

*Measurement accuracy estimation (for currents exceeding 7kA): Roughly 1%, based on Arduino's ATMEGA328P 10bit ADC specs of 2 LSB error . This assumes ADC reference voltage calibration, which was done just before (not during) the experiment.

View attachment 203981
Figure 1: Test "fuse".

View attachment 203982
Figure 2: Test setup.

View attachment 203987
Figure 3: Fuse blows.

View attachment 203988
Figure 4: Short circuit current vs time.
Click to expand...
Curious on the MEGA Fuses and MRBF fuses at 12v.

Did the MEGA fuses get tested?

Thanks again for all your work.
 
OK its a start - more work needed. eg try different gap distances, say start with say 1mm, 2mm ...20mm to see whether arc flash is initiated under what conditions so you have a reference. So copper wire at 22AWG is 0.645mm dia so do some calcs to estimate energy discharged (area under i vs t chart). Need to establish what energy you are breaking with a molten fuse - weight of copper wire vs temp estimate. Try and get a better logging system, borrow a DSO - in other words do some scientific digging before you go to larger diameter wires, also try Zinc strip fuses, pieces of solder, Lots of ground available to cover here if you persevere and publish here - maybe learned members can pitch in and help

FWIW I did a quick excel on graph 4 and assumed a 48V pd and the energy within 0.69s was ca113kJ and this is the amount of heat to raise 0.5L water from 25C to 100C. Somehow this doesnt feel right (2g of TNT equiv explosive power). Anybody comment pls
 
Last edited:
Johan said:
Per request of @Pi Curio, 10 Heschen fuses (sourced on Amazon; HSPV-30; 32A; 1000VDC; I1 = 33kA) were blown while mounted in their DIN-rail fuse holders (sourced on Amazon) in the vertical position.
View attachment 207422View attachment 207423

These fuses are filled with sand, similar to Class T fuses. The resulting current versus time plot is shown below. The legend shows the experiment number. Experiments 1 through 4 blow a single fuse. Experiment 5, 6, and 7 blow two fuses in series just to see if that makes any difference.
View attachment 207421
Here are all seven experiments on video ("Unlisted" on Youtube, so you need to use this link):

Notes:
  • There was no sign of external damage for fuses and holders.
  • Only experiment 5 has measurement data from both the Arduino and oscilloscope for comparison purposes.
  • The negative 25000A "charge current" was completely unexpected and seemed to be artificial oscilloscope nonsense at first, but this current was later reproduced on the Arduino (not shown, separate experiment). This *may* be an effect or "ringing" due to parasitic capacitance in the circuit. Anybody comments on this?
  • Scopes are generally very inaccurate for measuring voltages, so they are also inaccurate for measuring voltage drops across a shunt as was done here to calculate the current.
  • I estimate the Arduino accuracy to be roughly +/- 160A based on specsheets data.
  • The oscilloscope accuracy seems to be roughly +/- 1800A based on specsheets data.
  • See also here if you want to read more about oscilloscope accuracy.
  • When torquing the screws, the side panel of the thin plastic fuse holder housing can easily crack open.
Click to expand...


Hi!
As far as I understand it, these gPV fuses are an excellent alternative to T fuses.
And of course much cheaper.
Can someone tell me what are the disadvantages of using these, instead of T fuses ?
Thanks.
 
Scalabis said:
Can someone tell me what are the disadvantages of using these, instead of T fuses ?
Click to expand...

gPV fuses are especially designed for higher voltages. Using them at lower voltages can mean that the arc-extinguishing may not work correctly. In addition, the fuse element might melt but not fully clear the fault, leading to partial operation or sustained arcing/re-ignition. The fuse resistance is also typically higher, which can lead to losses/voltage drop.
I can't find the datasheet right away for this particular fuse, but they also tend to be slower than a Class T of NH fuse for the purpose (NH00 gBAT for example).
 
Lookup the datasheet for the fuse you are asking about - gPV is a class that can apply to many different fuse body sizes... The ones in the video are the midget fuses.. 50kA interrupt so that is the good. Tops out at 30amps is the bad. But those finger safe holders tend to top out at 30amps as well so that is what you get...

gPV is actually referring to the timer/current curve - saying they fit specific parameters...

In a class T we recommend the JLLN or JJN depending on the vendor --- if you want to use the NH series of fuses the aR curves are faster than the gPV curves... Class T as a rule are round and compact shape and size for the current they carry... the NH are less so.

It comes down to time/current curve, shape, size, voltage, current, AC or DC or both, and the AIC... Larger distances mean higher AIC...

NOTE - they also make the class T in a midget fuse as well... Class T fuses are ceramic or fiberglass bodies that have silica sand around the fuse elements. So when the fuse element burns away the sand fills the void and interrupts the current. The midget fuses as a class are all low current. There are a variety of other shapes and sizes just like there are for the gPV.

Class T fuse descriptions...


1740343593707.png


Scroll farther down to page 38 - now we are talking the NH series of fuses - these are fairly cheap in Europe and asia, but expensive in the US and the Americas. NH describes the physical shape of the fuse in this case

Physically they are larger than the class T and if you look at the data sheets they have similar AIC ratings...

1740343779845.png

 
Thanks for your replies.

I was thinking specifically at this model, that can be found at Amazon US or in Europe:

Solar DC1500V PV Fuse, Photovoltaic Fuses, gPV Type Fuse Link, HSPV-63L, 14 * 85mm, 50A 1500VDC, I1 50kA, for Solar PV System Protection, 5PCS

They are sand filled, 85mm in length (large distance between end caps), up to 50A, with a breaking capacity of 50kA

And much smaller than the NH00 …

Any thoughts about these ?

Thanks.
 
Scalabis said:
Thanks for your replies.

I was thinking specifically at this model, that can be found at Amazon US or in Europe:

Solar DC1500V PV Fuse, Photovoltaic Fuses, gPV Type Fuse Link, HSPV-63L, 14 * 85mm, 50A 1500VDC, I1 50kA, for Solar PV System Protection, 5PCS

They are sand filled, 85mm in length (large distance between end caps), up to 50A, with a breaking capacity of 50kA

And much smaller than the NH00 …

Any thoughts about these ?

Thanks.
Click to expand...

Same issues as what I mentioned in #109. The 85mm length is a bad thing for battery applications. They're just not meant for batteries. If you want something smaller than NH, go with BS88, for example a 125LET.
 
Scalabis said:
Thanks for your replies.

I was thinking specifically at this model, that can be found at Amazon US or in Europe:

Solar DC1500V PV Fuse, Photovoltaic Fuses, gPV Type Fuse Link, HSPV-63L, 14 * 85mm, 50A 1500VDC, I1 50kA, for Solar PV System Protection, 5PCS

They are sand filled, 85mm in length (large distance between end caps), up to 50A, with a breaking capacity of 50kA

And much smaller than the NH00 …

Any thoughts about these ?

Thanks.
Click to expand...
50A is too low for most battery applications. Generally you'd be looking for 125A+.
 
Brucey said:
50A is too low for most battery applications. Generally you'd be looking for 125A+.
Click to expand...
I'm using 125a to 150a but mostly 125a fuses on each battery. The total output I have seen is 300 amps but each battery never see's even 100 amps. The bms on each battery is supposed to limit them to 100 amps each anyways.
 
Crowz said:
I'm using 125a to 150a but mostly 125a fuses on each battery. The total output I have seen is 300 amps but each battery never see's even 100 amps. The bms on each battery is supposed to limit them to 100 amps each anyways.
Click to expand...
yep I use 80 amp class T's on each bank but under 3/4's load I could still run on one bank if needed.
 
Crowz said:
I'm using 125a to 150a but mostly 125a fuses on each battery. The total output I have seen is 300 amps but each battery never see's even 100 amps. The bms on each battery is supposed to limit them to 100 amps each anyways.
Click to expand...

Won't limit current (as an inverter would), rather disconnect. Pushing what it didn't supply onto remaining batteries.
That would save your fuses but system shuts down.
 
OffGridForGood said:
WOW
A Whole Lot of Blue !
Where is @sunshine_eggo !!?? :ROFLMAO:
Click to expand...
www.victronenergy.com

Power lift - Victron Energy

Let’s begin the new year energetically by taking a look at a time-lapse video showing the installation of a super-cool private energy plant in Romania. A lighting manufacturer – Amiras C&L Impex, who are located in Targoviste, Romania – were experiencing frequent power cuts, which prompted them...
www.victronenergy.com www.victronenergy.com

 
Hedges said:
Won't limit current (as an inverter would), rather disconnect. Pushing what it didn't supply onto remaining batteries.
That would save your fuses but system shuts down.
Click to expand...
All good.

I want the magic smoke to stay inside everything from inverters to the batteries. I don't care if it shuts down from it as long as it will come back on after I replace the fuse or fix a problem :)
 
@Johan this is really awesome! Where's the GoFundMe, or PP/VenMo, or KoFi or something so we can all chip in a bit towards you're awesome work?

As a way to help make it long term accessible, perhaps do a github with the description of your test setup, and then store the videos & machine readable source/fuse ratings/type of fuse & raw data captures of the tests. Could end up making for an easy way to put up a interactive page in front that just checks your github, and then as you add data, add new test points, etc it automagically has more data, plots new data points, etc.

I miiiiight be considering volunteering to help, although I don't have a ton of time I could devote to this.
 
You must log in or register to reply here.

Similar threads

wigglybuff
Add another fuse?
Replies
2
Views
142
wigglybuff
wigglybuff
T
Seplos Mason - advice on terminal fuse and DC isolation switch
Replies
6
Views
173
upnorthandpersonal
upnorthandpersonal
S
Class T fuse + MRBF on individual parallel batteries (x4)?
Replies
8
Views
1K
robbob2112
robbob2112
J
Batteries in parallel to busbars or each other?
Replies
8
Views
276
robbob2112
robbob2112
wpns
Bus Bars, MRBF Fuses, and Copper Washers?
Replies
20
Views
634
Rexracer
R

diy solar

diy solar
Back
Top