Counterfeit fuse?

[atgeek]

Some time ago I bought from AliExpress 100 assorted glass fuse sold as quick blow type, from 0.2 to 15 A. I've just tested them with a bench power supply, the 0.5A blowed at 2A after about five seconds, the 6A at about 12A instantly. Is it normal for fuses to have such high discrepancy from the nominal hold current? Or maybe it's just because are poor made?

[PA0PBZ]

Have a look:

http://www.littelfuse.com/~/media/electronics/datasheets/fuses/littelfuse_fuse_217_datasheet.pdf.pdf

[Gyro]

All fuses have a current/time curve. The figures you quoted don't sound that unusual. The fuse rating is the 'carry' current. Depending on the fuse it normally takes around 10s or more to blow at twice rated current and somewhere around 0.1 secs at 10x (for fast blow). Without knowing the time/current curve for the specific fuses it's hard to get closer.

The normal worry with fake fuses is the ceramic body ones which are supposed to be filled with the right grade of silica sand, but in fake ones, often aren't. A glass fuse is low interrupt current (not HRC), so it's no more that a piece of fuse wire in a glass tube, there's no point in faking that. 

P.S. You should be careful not use glass fuses where HRC ceramic are required, eg. Mains use. The can't interrupt the prospective fault current (say, 1000A+) and will explode instead.

[atgeek]

Ok, thanks for the clarification

[RoGeorge]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

As an example, you can easily fry a LED if you set your power supply at 20V/20mA, then connect the LED (without any series resistor). The energy accumulated by a few thousands microfarads at 20V, combined with the low series resistance of the given electrical circuit, is enough to push a spike of tens or even hundreds of Amps through the LED.

[rsjsouza]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

Your statement seems a bit too generalistic, as I wouldn't assume that all bench power supplies fail to take into consideration the output capacitors in the control loop.

As an example, you can easily fry a LED if you set your power supply at 20V/20mA, then connect the LED (without any series resistor). The energy accumulated by a few thousands microfarads at 20V, combined with the low series resistance of the given electrical circuit, is enough to push a spike of tens or even hundreds of Amps through the LED.

Perhaps you are referring to a specific type of LED, since at these ratings you can pretty much fry any of the more ordinary types of LEDs, with or without capacitors on the output. 

[atgeek]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

As an example, you can easily fry a LED if you set your power supply at 20V/20mA, then connect the LED (without any series resistor). The energy accumulated by a few thousands microfarads at 20V, combined with the low series resistance of the given electrical circuit, is enough to push a spike of tens or even hundreds of Amps through the LED.

A solution could be discharge the output capacitor and then connect the load, am I right?

[rs20]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

Your statement seems a bit too generalistic, as I wouldn't assume that all bench power supplies fail to take into consideration the output capacitors in the control loop.

Wait, are you saying that a noteworthy fraction of power supplies detect the high current and actually violently sink charge out of the capacitor faster than the LED can drain it, until it matches the voltage of the LED? That seems to me like a very difficult control systems problem, if not basically impossible.

Seems like it'd be an awful lot easier to make a power supply without output capacitors.

[RoGeorge]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

As an example, you can easily fry a LED if you set your power supply at 20V/20mA, then connect the LED (without any series resistor). The energy accumulated by a few thousands microfarads at 20V, combined with the low series resistance of the given electrical circuit, is enough to push a spike of tens or even hundreds of Amps through the LED.

A solution could be discharge the output capacitor and then connect the load, am I right?

Indeed. In this case the initial current will be lower then the expected Amps setting, because a part of the generated current will be used to charge the output filtering capacitors.

To be more specific, my tests were made with a Rigol DP832, but I know for sure most (if not all) of power sources have a few hundreds or even thousand microfarads placed at their output, after the current sensor. To check if a particular lab source have this caveat, set a very low current and a big voltage, then power the output without any load. You will see (if this caveat is present) that the voltage will raise very slow and linear in time. You can even deduce the capacitance of the output capacitor by timing how long it takes to achieve the requested voltage.

[RoGeorge]

...Seems like it'd be an awful lot easier to make a power supply without output capacitors.

Usually a power supply is used as a voltage sources, so the designer's main goal will be to minimize the output resistance of the voltage source. The low output resistance is a very important parameter, especially for transient loads and AC currents. This might be not so easy to achieve with a series measuring shunt and a control loop with a limited speed. As a compromise, the designer can choose to put a low ESR capacitor right at the output, after the current sensor.

[RoGeorge]

As an example, you can easily fry a LED if you set your power supply at 20V/20mA, then connect the LED (without any series resistor). The energy accumulated by a few thousands microfarads at 20V, combined with the low series resistance of the given electrical circuit, is enough to push a spike of tens or even hundreds of Amps through the LED.

Perhaps you are referring to a specific type of LED, since at these ratings you can pretty much fry any of the more ordinary types of LEDs, with or without capacitors on the output. 

Not if you connect the LED first, then energize the power source output.
https://hackaday.io/project/7590-retardo-davinci/log/25376-rigol-dp832-power-supply-set-for-20-ma-can-kill-a-led

[rsjsouza]

One more note when using a bench power supply as a current source: There are big capacitors at the output of the power supply, near the power supply terminals. These capacitors are usually out of the control loop. As a result, when a load is suddenly connected, the current can go as high as the internal resistance of the output capacitors (and the resistance of the load) allow.

Your statement seems a bit too generalistic, as I wouldn't assume that all bench power supplies fail to take into consideration the output capacitors in the control loop.

Wait, are you saying that a noteworthy fraction of power supplies detect the high current and actually violently sink charge out of the capacitor faster than the LED can drain it, until it matches the voltage of the LED? That seems to me like a very difficult control systems problem, if not basically impossible.

Seems like it'd be an awful lot easier to make a power supply without output capacitors.

When I wrote that I was thinking on the lines of your last sentence but replied something completely nonsensical as you pointed out.

Most of the bench power supplies I have used do not have smoothing capacitors directly on the output.

[macboy]

At least your fuses blew at some reasonable current. The closer the current is to the nominal current, the long it takes, up to several minutes. At 2x the nominal current, most fuses will blow in some tens of seconds or so (less than a minute). At 3x the nominal current, it should take only a few seconds. At 10x the nominal current, any fuse should blow very quickly. (alway check the specs, these are just very rough estimates).

THIS is a counterfeit fuse:
https://www.eevblog.com/forum/chat/counterfeit-chinese-fuses/

I put 33 A through a "3 A" fuse long enough to completely melt the thing, but it kept on conducting. Any fus should have blown nearly instantly at 11x the rated current. Apparently, this is one of those newfangled Chinese-spec "no-blow" fuses. I guess they are used (by idiots) where the inrush current will blow even slow-blow fuses. Just don't count on them preventing a fire in case of trouble. 

[Neilm]

I have posted this before, but it shows a UK plug with a fake fuse trying to stop a short circuit.

[691175002]

I have posted this before, but it shows a UK plug with a fake fuse trying to stop a short circuit.

Shouldn't the circuit breaker open in that situation?  Is this just a problem with fake fuses, or glass fuses in general?

[Mechatrommer]

I have posted this before, but it shows a UK plug with a fake fuse trying to stop a short circuit.

or a fake video? the fuse must be like 100A or more during the blast. the enough energy to propel a heavy mains plug out of its socket is suspicious. a fake fuse with very thick copper is unlikely imho. or either they put The Black Sand in it in the demo video instead of silica sand just for the fun of it.

[rs20]

Seems like a pretty good fuse to me; it certainly cut power to the faulty device, isn't that what fuses are for?

[amyk]

That reminds me of the explosive disconnects used on HV mains transmission lines.

[rs20]

Yeah, and it's easy to identify which plug has the blown fuse in it too!

[Gyro]

That reminds me of the explosive disconnects used on HV mains transmission lines.

BMW use something similar on their rear mounted car batteries. I think they call it a pyroelectric disconnect (it has a separate trigger wire).

[joeqsmith]

I have posted this before, but it shows a UK plug with a fake fuse trying to stop a short circuit.

or a fake video? the fuse must be like 100A or more during the blast. the enough energy to propel a heavy mains plug out of its socket is suspicious. a fake fuse with very thick copper is unlikely imho. or either they put The Black Sand in it in the demo video instead of silica sand just for the fun of it.

May be a controlled test of sorts.  Personally I can believe it.  My own fuse powered rocket.  Who needs Black Sand.
https://youtu.be/xLDok9Sm07Q

[Mechatrommer]

May be a controlled test of sorts.  Personally I can believe it.  My own fuse powered rocket.  Who needs Black Sand.
https://youtu.be/xLDok9Sm07Q

some "practical" achievement there congratulation. but that rely on air expansion on a tiny sponge, try again with the open air plug like that. i'm sure the fake fuse manufacturer need to spend sometime to get the effect, i bet more expensive cost than the original fuse.

[joeqsmith]

some "practical" achievement there congratulation. but that rely on air expansion on a tiny sponge, try again with the open air plug like that. i'm sure the fake fuse manufacturer need to spend sometime to get the effect, i bet more expensive cost than the original fuse.

Forgot to say, I would guess your comment about the cost to set it up and film it was a more expensive than the cost of the original fuse was an understatement! 

I'm sure the fuse holder made a good enclosure.   I started out with fuses that were not contained. Like the Black Sand, you need to enclose it.


RANDOMTRONICS and I attempt to run a fuse inside an enclosed fuse holder.  This was my fuse tests to go along with his.

[Mechatrommer]

I would guess your comment about the cost to set it up and film it was a more expensive than the cost of the original fuse was an understatement! 

to be fair, what i meant is the cost of making a fake fuse alone thats capable of propelling a quality 3 pronged plug, compared to the cost of making one original fuse, and i meants the production cost in factory, not the cost of home diy. the cost of making a demo video, the hi-speed videocam, and damaged socket/plug is another story for you to waste. btw, lucky you no pcb projectile in the 2 later videos. i hope you are not the one who made the fake fuse fake video plug projectile above. if yes then you can elaborate, not extrapolate from your own setup.

[Neilm]

I have posted this before, but it shows a UK plug with a fake fuse trying to stop a short circuit.

or a fake video? the fuse must be like 100A or more during the blast. the enough energy to propel a heavy mains plug out of its socket is suspicious. a fake fuse with very thick copper is unlikely imho. or either they put The Black Sand in it in the demo video instead of silica sand just for the fun of it.

Produced by a testing house called ERA who were taken over by Intertek some years ago. It was part of an investigation into fires from faulty power lead extensions. As part of the testing they investigated the fuses - they also found similar fuses on sale in market stalls.

[joeqsmith]

Appreciate your concern about safety and my wastefulness.  Fortunately, I am capable of assessing my own risks and have resources to waste.  No I was not involved at all with making what you are considering a fake fuse fake video.  Only showing examples of what a little copper expansion looks like and suggesting that it may not be fake but part of a controlled test.   

I would guess your comment about the cost to set it up and film it was a more expensive than the cost of the original fuse was an understatement! 

to be fair, what i meant is the cost of making a fake fuse alone thats capable of propelling a quality 3 pronged plug, compared to the cost of making one original fuse, and i meants the production cost in factory, not the cost of home diy. the cost of making a demo video, the hi-speed videocam, and damaged socket/plug is another story for you to waste. btw, lucky you no pcb projectile in the 2 later videos. i hope you are not the one who made the fake fuse fake video plug projectile above. if yes then you can elaborate, not extrapolate from your own setup.

[joeqsmith]

Produced by a testing house called ERA who were taken over by Intertek some years ago. It was part of an investigation into fires from faulty power lead extensions. As part of the testing they investigated the fuses - they also found similar fuses on sale in market stalls.

I should have read one more down.  Thanks for the info.  Did they provide and details about the test?

[Mechatrommer]

showing examples of what a little copper expansion looks like

copper only expand very little. there are lot of other high speed youtube videos showing how it behaves during fuse failure. do that again in the open air, you'll get just a little pop. what expanded tremendously is the closed air volume due to sudden temperature rise.

Produced by a testing house called ERA who were taken over by Intertek some years ago. It was part of an investigation into fires from faulty power lead extensions. As part of the testing they investigated the fuses - they also found similar fuses on sale in market stalls.

I should have read one more down.  Thanks for the info.  Did they provide and details about the test?

this. otherwise its just hearsay nobody obliged to the belief. dont take this personally i just trying to become academic here over rated fake fuse is also possible though.

[joeqsmith]

copper only expand very little. there are lot of other high speed youtube videos showing how it behaves during fuse failure. do that again in the open air, you'll get just a little pop. what expanded tremendously is the closed air volume due to sudden temperature rise.

this. otherwise its just hearsay nobody obliged to the belief. dont take this personally i just trying to become academic here over rated fake fuse is also possible though.

I have ran the copper wire in open air and provided you with the link.   "copper only expand very little", what's academic about that?  Compared with what?  The black sand you mentioned?   Telling you the pressure ratio from copper converting from a solid is no good as you stated, you are not "obliged to the belief" or to do the research.  Of course, I doubt you would get very far in academics without doing your own research.

[Mechatrommer]

not sure what you mean, but here, for kids to learn...


here some more, barely moves anything, have fun...

[SeanB]

Fake a fuse, using a ceramic ( or glass) that is made very thin, and with end caps that are just a light press fit on there. Then make the element out of aluminium foil strip, and use one that fits the inner diameter, so there will be no problems with poor contact with those poor caps. Very cheap as you do not need to do any incoming QC checks or testing for conformance aside from the final part looking similar to the real one.

Then enclose in a typical UK plugtop, where the fuse is in a holder accessible from the underside, and make the pins of the plug from brass coated steel to cut cost further, and at the top use a crimped hollow section for the wires, then mould it in cheap plastic. As UK sockets are typically fed with a 32A breaker, and the supply impedance will in almost all cases be well under 100milliohms the fault current this fake must break is very high.

This current is plenty enough to turn the foil into metal vapour, and this shatters the fuse housing, blowing out the door of the holder, which is between the plugtop and the socket. This will blow it out the wall, as seen in the video. Yes, the breaker in the consumer unit will trip, but it will still take at least 2 mains cycles to completely open and break the arc inside it safely, which is 2 cycles at the short circuit current dissipated in the arc of the failing fuse, and a single cycle of the breaker arc disruptors clearing this arc. You might have the supply fuse itself blow, but this will be rare, as there typically is a 100A fuse at the point of supply, and a 60A or 80A breaker at the metering point, which are there for cable protection, so are designed to be slower acting, to clear shorted or overloaded cabling before it heats up enough to catch fire.

By me the fuse is a 400A unit, at the substation 50m away across the park, though the cable run is 500m in length. Really unlikely for that fuse to go for any fault downstream of the meter short of you going to the cable right by the meter with an axe.

[joeqsmith]

not sure what you mean, but here, for kids to learn...

here some more, barely moves anything, have fun...

Last I knew, these AC plugs were not used with DC battery chargers or small bench supplies but I have no problem with you believing that compares with the AC mains.  At least the last link shows a 120V fault.  Still a nice low voltage.  Do you believe that the devices rated for 10KA and up fault currents have no purpose? 

Very cheap as you do not need to do any incoming QC checks or testing for conformance aside from the final part looking similar to the real one.

   

As UK sockets are typically fed with a 32A breaker, and the supply impedance will in almost all cases be well under 100milliohms the fault current this fake must break is very high.

To be clear, you are saying that common household outlets are 220V with a 32A breaker?  If so, wow! 

[Gyro]

To be clear, you are saying that common household outlets are 220V with a 32A breaker?  If so, wow! 

The 32A breaker is protecting a whole ring of Sockets. In the UK we use ring circuits in houses rather than the radial circuits used in the US. The 32A is to protect the ring wiring from damage due to shorts or gross overload, by high loads on too many sockets at the same time. As all UK mains plugs contain fuses (3A, 5A or 13A), these are used to protect wiring to the individual appliances... unless people start buying fake fuses of course!

P.S. It's 230V

[joeqsmith]

Does the average home use multiple rings?  How many outlets per ring?  Does the home have another breaker up from the 32A?    Just curious how different it is.

[crazyguy]

Does the average home use multiple rings?  How many outlets per ring?  Does the home have another breaker up from the 32A?    Just curious how different it is.

use multiple rings : depends on floor/room area, cable size, length, $$$ .... 2.5mm2  cable ring main for 32A socket outlet

How many outlets per ring : no limit, if my memory is correct

another breaker up from the 32A : Yes, separate circuit breaker for lighting, air conditioner, shower, cooker ....

[Mechatrommer]

its your house you can put however many rings you like, your money. here in our resident, the developer used 6A breaker for lights, fans appliances, and 16A breaker for plug outlets, around 4 - 6 plugs per radial, this number needs to use some common sense. for our house extension, my brother advised me to use ring for the kitchen (heavier current requirements, heaters, cookers, refrigerators etc), so he did, iirc around 6 plugs (we dont have enough wall to put 8 ), we use 20A breaker, any objection? my ring is safer than the regulations, the breaker will trip before any projectiles.

[Gyro]

Does the average home use multiple rings?  How many outlets per ring?  Does the home have another breaker up from the 32A?    Just curious how different it is.

Yes, normally 2 rings for a typical house, one upstairs one downstairs, each with a 32A breaker. With solid floors downstairs it is also common to split the rings LHS and RHS of the house - less cabling length. Convenient for having easilly accessible power when one circuit is out, but requires a bit more care in checking that sockets are actually dead. I can't remember if there's a limit to the number of sockets directly on a ring (there probably is somewhere). Most sockets are duals so the maximum possible load at one location is 26A (2 x 13A fused plugs).

It is also possible to have a few radial spurs off the ring with a maximum of 2 sockets (normally a dual), a single length of 2.5mm2 will take 26A. Triple outlets are also available, but these always include a 13A fuse in the faceplate to avoid any possibility of locally overloading the ring.

Lighting is also normally on rings (6A breakers) one for upstairs, one for downstairs.

Dedicated High load appliances (electric shower, cooker, water heater) are run on separate radial circuits on 4mm2 and 32A breaker (occasionally 6mm2 for high powered electric showers I think). Other stuff (gas heating, alarms - for convinience,  outbuilding etc would also be on spurs).


P.S. The incoming utility supply cable for a typical domestic installation is normally fused at around 80A.

[SeanB]

South Africa, where I live, uses exclusively radial connections, and these are generally fitted with a 20A breaker, and use 2.5mm cable ( approx 14G USA cabling), with lighting generally using a 10A breaker and 1.5mm ( 16G USA approx) cabling, and the regulations state there must be a ground conductor in each cable and it must be connected to all metal parts of the fitting other than screws. Water heating, cookers, ovens and outbuildings via a dedicated breaker, either 20A for water heaters, 40A ( and 6mm cable) for cookers, and for outbuildings the cable size and length determines the breaker size, it could be from 20A to 40A.

Incoming supply is almost always via a meter on the edge of the property ( for ease of reading manually) and will have typically an 80A breaker on it. Supply cable will have a 100A fuse on it, and will be 16mm cable.

You can also have 3 phase supply for a big house, with a 16mm 4 core cable, 3 phase 80A breaker and meter, and then your distribution board after it. If you need bigger you are a business, and the cable will be 25mm, with a 200A fuse on the supply side, and 200A breakers on your side that are rated for 20kA breaking capacity. Lesser breakers tend to blow apart at this prospective fault level, and you really want to have that blast resistant steel panel to keep the flash in there.

[Deodand2014]

John Ward back in 2015 found someone selling BS1362 15 amp fuses, which are obvious counterfeits as BS1362 fuses cannot be rated higher than 13 amps. The odd thing was the fuses were reasonably well constructed, conformed (externally) to BS1362 and actually had a sand filling. Under testing the '15 amp' fuses spewed smoke and discolored before failing, while the legitimate BD1362 13 amp fuses just failed. Of course the big unanswered question was why someone was trying to counterfeit a non-existent product.

[amyk]

15A is a common fuse rating (likely more common than 13A), and BS1362 is a common size of fuse. It's not hard to see some manufacturer producing that combination.

(The ink turning black might be a feature to show that the fuse has gotten too hot, even if it remained closed. The smoke is less featureful, but perhaps the sand had some impurities in it...)

14A fuses apparently also exist, but it's hard to explain why, given the tolerance that fuses have --- see video above, where even the genuine ones are specified to take anywhere from less than a second to almost 2 minutes to blow at double their rated current (a range of 3 orders of magnitude!) Thus the 15% difference between 15A and 13A fuse wire may not be significant until you have a very, very large sample size.

[joeqsmith]

Voltlog posted this video about some Bussmann fuses.  I suggested in the comments some ideas about why their test setup behaves the way it does.  Waiting to a follow up.

[joeqsmith]

A video I made about some SIBA branded fuses that appear to be counterfeit.



I also tested some of the fuses from the kit Dave sells that did not match up with the data sheet.  Dave had a different datasheet than the one I had.  Dave contacted the manufacture and it turned out there was yet a third datasheet that was the correct one.  Pretty crazy stuff. 

[IanMacdonald]

Fuses are not all that good at protecting equipment anyway. As the OP pointed out, short term rupture current is about double the rated current. However you can't reduce the rated current because that would lead to gradual degradation and eventual nuisance blowing. That's for enclosed fuses; for open wire jobs the margin can be even greater.

It's quite common for a UK house supply to have the 32A MCB trip on a short, whilst the (genuine) plug fuse of 13A or even 5A rating remains intact. That shows how much better MCBs are are detecting overcurrent and dealing with it rapidly. The proviso is that MCBs do have a limit to the s/c current they can interrupt, typically it's 6kA or 10kA. Unlikely to be exceeded on domestic circuits but may be a reason to use fuses on higher current stuff.

The other problem with fuses is that in the primary side of small transformers like wall warts, a fuse light enough to protect the transformer is likely to suffer nuisance blowing. That is why they tend to use thermal trips instead.

In the audio field, users were forever protesting that amplifiers shouldn't have failed when the speakers were shorted, because after all they have fuses. I would point out that the fuse is a safety device and as such it's not necessarily intended to save the equipment, only to prevent a hazardous situation from arising.

[Deodand2014]

John Ward has been sent another set of 'BS'1362 fuses, these ones were credited to a manufacturer called Kingguang and rated at 20 Amps. Sadly he does not test the fuses to see what happens when subjected to an overload. Otherwise they seem to be like the 15 amp fuses featured in a video I posted earlier in this thread, they seem well made, have a sand filling and the correct dimensions for BS1362, but the wire again seems to be a flat bar shape rather than a round wire.

Some of the comments to the video suggest these fuses are being made for the types of multimeter that use BS1362 fuses.

[Gyro]

One small clue might be that BS1362 fuses aren't made above 13A, the rating of a UK mains plug. 

If it's above 13A then it's not covered by BS1362.

[supercilious]

I have encountered fake 13A BS1362 fuses before (with bussman branding, no less). They looked identical to the real thing with no obvious giveaway that they're fake other than the fact that they didn't blow in time to protect a device from meltdown. Don't just rely on the markings being ridiculous to indicate a fake.

[Deodand2014]

I have encountered fake 13A BS1362 fuses before (with bussman branding, no less). They looked identical to the real thing with no obvious giveaway that they're fake other than the fact that they didn't blow in time to protect a device from meltdown. Don't just rely on the markings being ridiculous to indicate a fake.

Sound's like the experience of this guy, the fuses no matter what the marking apparently broke around 18amps. He only shows one test, but that's enough, since the fuse was rated 1 amp.

[rs20]

Fortunately these things are very easy to test. If you're cheaping out on fuses, just buy a few extra and characterise them.

[Deodand2014]

John Ward has been sent another batch of 'BS'1362 fuses, this time marked for 13amps. This time whoever made them was not following the BS1362 specification as to dimensions/color. Nor as to performance.