Welcome to Laser Pointer Forums - discuss green laser pointers, blue laser pointers, and all types of lasers

Buy Site Supporter Role (remove some ads) | LPF Donations

Links below open in new window

FrozenGate by Avery

Bench power supply and laser diodes..

PLT5 450B single mode laser from above-picture.
an retailaluhousing with an alubacktrigger with an cobber-module inside and radial heatsink at front.

clocking in at 197mA at tailcap DMM on mA-setting and 0.169A on (10A) Amp-setting //2xAAA. (often puzzeles me why multimeters-values differs from when on Amp (10A) to when on mA (<600ma) setting and with so much and on relative accurrate meters..
(Le2max X822 campictures, plain no fog or likewise)

55020d1488676130-bench-power-supply-laser-diodes-single-m-case-flash.jpg

55021d1488676130-bench-power-supply-laser-diodes-single-mode-flash.jpg

55022d1488676130-bench-power-supply-laser-diodes-singlemstand.jpg

55023d1488676871-bench-power-supply-laser-diodes-singlm-casing.jpg

55028d1488683895-bench-power-supply-laser-diodes-sigkm.jpg

55024d1488676871-bench-power-supply-laser-diodes-singlm-case.jpg

55030d1488686293-bench-power-supply-laser-diodes-awt.jpg

55025d1488676871-bench-power-supply-laser-diodes-victor.jpg
54886d1488246440-bench-power-supply-laser-diodes-img_20170228_001920.jpg
55029d1488686293-bench-power-supply-laser-diodes-img.jpg

Btw, the 60US'labby on the handheld rack, is an NDB7775 2.5w with driver and all inside the heatsink-housing, that seem to be flooded with siliconerubber :-/ but its ideel to be run of RC 3cell lipo packs with an XT60 to jack adapter.. its clocking in around 1.20amp on attched RC'ampmeter (though its <12v as to the 1.2A)
 

Attachments

  • single mode flash.jpg
    single mode flash.jpg
    125.9 KB · Views: 218
  • singleMstand.jpg
    singleMstand.jpg
    161.1 KB · Views: 236
  • single M case flash.jpg
    single M case flash.jpg
    208.3 KB · Views: 260
  • singlM casing.jpg
    singlM casing.jpg
    130.5 KB · Views: 225
  • victor.jpg
    victor.jpg
    76.7 KB · Views: 239
  • sigkM.jpg
    sigkM.jpg
    133.4 KB · Views: 232
  • img.jpg
    img.jpg
    120 KB · Views: 208
  • awt.jpg
    awt.jpg
    159.9 KB · Views: 348
Last edited:





For the first question, I install a dummy load across my
adjustable power supplies. I use a 2.2K ohm to keep a
small load in the regulator.
HM
 
I just watched the EEVBlog testing of the Korad KA3005P...
I wouldn't use that PS for LD testing. Check out the Fail
starting at 23:52.


Jerry
 
Last edited:
I just watched the EEVBlog testing of the Korad KA3005P...
I wouldn't use that PS for LD testing. Check out the Fail
starting at 23:52.


Jerry all of these issues have been adressed by Korad with stronger trans in newer batches...(was down to an lacking 220v trans(not the aussie model), being tortured by the max'amp on a hefty 240v aussie juice)

on EEVblog the Korad seems to be hands-own the best budget psu with the price in mind and actually consist of inner parts that for an extreme cheap liniar-psu, would be difficult to manage in private DIY-world with an eye on price, - in the Korad-thread on the EEVforum, these Korad PSUs gets feedback that are pretty solid, but there are multiple models and the first batch of KA3005D suffered according to EEVblog.

I got both the KA3003P (62US delivered) model and the EU KD3005P model (75EU delivered), and there is quite a difference, but i belive the KD model is of newer heritage and this model i purchase from Reichhelt GE in Germany, and with all the bells and whistle as to EU legi & regi and sadly also insunk bananaplugs.

one significant difference is the noise, the KD3005P model is extremely quite when in use' as to the KA3003P and I also cherise that you got the double joystick click-features on the two-knobs V & A, but I do missing an specifik load button on the KD-model front, but still you got all the features from the KA-front panels also on the KD model if you run the KA-software on the KD-model, then you have all the memory-features and protection and a seperate loadbutton as software..(the models need to have an P as the end sign to have rs232 and usb interface so you can datalog and graph it out, and all that jazz of graphcolorfont and graph'grid systems and you can also program both model to run timely-sequence-setup' so at specified timeintervals it runs a set valueagenda of volt&amp-values at numerous intervals)
software:
Top Quality Power Supply Manufacturer in China - SHENZHEN KORAD TECHNOLOGY CO., LTD.
https://www.reichelt.de/index.html

EEVblog has only tested the old KA3005 model (first batch) and a limited old 220vtrans'model not suited for the high-240veffect in Australia, and are also stating in the details in the videos that these issues have been adressed from Korad in the newer batches and EEVblog linking to this partically video in there info-spec in the fail-vid above...


here is an teardown, but its the first batch and where it come short..(literally)



Here EEVblog showing the new batches of KA-model and whats have been changed..


check the conclusion from 2:20 where he sums it up as its an fantastic power supply now, and your getting an amazing bargain for the extreme low ridicalus price and its the best bang for buck out there and also relates to the positive-conclusion in the followup video above, where another users are testing & torturing it as an arc welder, and the wicked thing Korad has actually make more upgrades since that, and afaik' the new KA models got 3 caps on the board from only 1 on the old batch and still only 1 on the new batch, to having 3 caps in the newest ones and a regulated fan, but at the end of the day, it aint the KA3005 model i got, its only the KA3003P and they are not identical, its an lesser trans in the 3amp version (36v/3a/220v spec trans, you can see it on the weigt pictures in page1) and the board are also different in mine, so cant conclude or salute to all the good things he states about it, as mine KA-model is only the 3amp version, while the KD model I got is the 5amp, but that one is likely the same since that has regulated fan also.(but all the video and the EEVblog on the korad-forum-thread give an one hell of an good insigt into this adjustable & programable psu from multiple EEVusers testing it. :)
--
Here is the forum-EEVblog-thread where the users testing there own with oscillioscope and different tool for spiking and general pro and cons, (and with the price in mind' as not that relavant to compare a budget model like this against an fx 500-1000'US modell) overall verdict imo' pretty solid adjustable programable liniar-150w PSU.
EEVblog #315 - Korad KA3005P Review/FAIL - Page 13

It was actually a minor mistake i ended up with two, but they both got there pro and cons, as to different models' so thats a plus.. - i had first checked aliExpress for an cheap PSU and saw and officiel store from Korad selling an KA3003Pmodel (30v/3A/90w) way under all the other thirdparty-sellers, and this offer was supprinsingly with free shipping, and took the plunge on that to 62US, but soon got the impression when it wasnt processed after 10days and the item had been pulled for any further purchases, - that it likely was an error-listing, and contacted the officiel Korad site as an buyer that wanted to purchase directly, and what was the cheapest they could deliver a programable KA-model for, and they stated some before-shipping-prices there where around aliExpress thirdparty-level, and the cheapest way of shipping was 70US for shipping alone when buying direcly from Korad, and not thrue AliE, so it would summon to around 140US delivered..

then i could conclude that the item to 62US i have purchased and there was just hanging and was not active anymore, likely was an mistake and would not get shipped, as to "error-listing" and lack of shippingcost not included..(these PSUs quite big in casing and packing, so ewen if they are from China quite pricey to ship)

I then started to look regionally here in EU, since i also risk an siginificant tax, here from the worlds most taxed country DK, and a item like this could easyly double when it ran into massive-tax + a mandatory taxhandling-fee..

I then saw an offer on Reichhelt (huge German-tool & electronic seller) for the new EU KD3005P model, that was supprinsingly cheap with EU-price-level in mind and with only 5'EU in shipping to DK Copenhagen, and since purchases in EU got no tax, it was quite ideel, and 2 days delivery-time' so took the plunge on that KD3005P (30v/5A/150w) and at the same time i could go bananas in numerous-different-resistor-diodes to be shipped along same time, in the chase to make an dummyload for an rookie like me. :can:

and all was good, the KD3005P arrive and it was great, but then suddenly the KA3003P suddenly also arrived at my doorstep, and the AliE Korad-storeseller had been so straight to actually processed the order, ewen thow just shippingcost on the DHL fastshipping-packagemarker on the package seems to origin more then what i paid for it all together, and the partically seller has also been so nice to put an value on the parcel at 9.99US that i stated at check-out, so i shouldnt pay any danishimportfee, as to an <11US danish value-item-limit.. (topnudge' to experience chinese-sellers like that)

For the first question, I install a dummy load across my
adjustable power supplies. I use a 2.2K ohm to keep a
small load in the regulator.
HM
thx for the feedback Mike, what adjustable PSU do you have...?
Im a rookie when it comes to electronic, and havent taken any step to mod the psu, or cross the pads with an dummyload, simply from lack of knowledge, i do cross the +/-pads before i put a diode on it, to make sure aint any juice left in the system and short it out with an little amp in play..
I have checked for spiking but a very basic test, with an DMM at MAX to see if it delivers higher then ask for, and no spike-signs there, but dont got any O-scope in house, still waiting for my 13US shipped DIY oscilliscope from GB- :whistle:.(and certainly dont got high hopes for that one, but will give some knowledge, for relative little amount)

as when it comes to dummyload i tryed some of the schematics there are here on LPF, most basic ones, as to the different voltage needed for different LDdiodes with appropiate 4001 diodes and 1ohm and 0.7v each, and then read the current from both sides of the 1ohm, and take it from there, but ewen a phrase like mv = ma, got me to wonder, hmm i need to be in the mv setting on the dmm when reading current..(go figure..)
but for the time being trying to get some insigt into resistor-diodes and testing with some of the bigger resistor diodes, since the 4001 is pretty ampweak, i also purchased a lot of different ones, and just testing at a breadboard how they stand out (with some thick breadboardwires, and what load they are synonym with in series, wthout any fragile-driverparts in between them and the psu and nothing to f..k up..(but as a start i likely need to take a moment to read the manual from the PSU, so i can get a better understanding of the jump between constant current and constant voltage that takes place and not just take it all on the fly
at some diffuse-level as to actual comprihention..)
 
Last edited:
the new double threaded-screw-in-diodesink works good :beer: and makes it a lot easyer to change diodes with wth the use of RC-plug to the LDpins. (high C'rated lipo-plug & silicone wires)...
Housing a PLTB450B
P8MAX cam
55162d1489126970-bench-power-supply-laser-diodes-sink-med.jpg
55164d1489126970-bench-power-supply-laser-diodes-sink-off-clo.jpg
55163d1489126970-bench-power-supply-laser-diodes-simk-off-dist.jpg

LE2M cam.
55161d1489125744-bench-power-supply-laser-diodes-sinked-pltb450b.jpg
55160d1489125744-bench-power-supply-laser-diodes-sinked-mid-pltb450b.jpg
55159d1489125744-bench-power-supply-laser-diodes-sinked-pltb450b-low.jpg


Looks wicked how deepblue the PLTB450B diode delivers at relative little effect, the beam allmost looks like its painted on the picture (no fogh or likewise and picture is taken in the middle of the day with curtains on with Huawei P8MAX mobile on std sett., a mobile that got 4 colored sensor RGB+W (was socalled worlds first, the implement of White to the standard RBG camsensor-spektre (RGBW) to sum it up' a promogimmick, but it does got its forces..
55178d1489146994-bench-power-supply-laser-diodes-blue-pltb450b-2.jpg
55179d1489146994-bench-power-supply-laser-diodes-blue-pltb450b-3.jpg

Gotta get myself a budget LPM, Its certainly missed :-/
 

Attachments

  • sinked pltb450b low.jpg
    sinked pltb450b low.jpg
    82.4 KB · Views: 180
  • sinked mid pltb450b.jpg
    sinked mid pltb450b.jpg
    92 KB · Views: 321
  • sinked pltb450b.jpg
    sinked pltb450b.jpg
    82.4 KB · Views: 177
  • simk off dist.jpg
    simk off dist.jpg
    76.3 KB · Views: 172
  • sink off clo.jpg
    sink off clo.jpg
    74.3 KB · Views: 184
  • sink med.jpg
    sink med.jpg
    77.2 KB · Views: 193
  • blue pltb450b 2.jpg
    blue pltb450b 2.jpg
    101.6 KB · Views: 183
  • blue pltb450b 3.jpg
    blue pltb450b 3.jpg
    110.1 KB · Views: 314
Last edited:
That blue light of massless particles, looks wicked and like its painted on..
another Osram PLTB450B build driven conservative, where im fooling around with different screw-on-LDsinks attachements and testing the use of indium, copper'grains and solderballs in trying to close as much of the "diode in thermal conductive-materiel and see if any gain of thermal management is achived by the use of any of these and perhaps in the c-operation with some good shin-etsu thermal nano-compound..

(P8MAX cam)
55456d1490434637-bench-power-supply-laser-diodes-dark-beam.jpg

55457d1490434637-bench-power-supply-laser-diodes-img_20170312_132417-crop.jpg

55459d1490434637-bench-power-supply-laser-diodes-img_20170312_132247.jpg

55458d1490434637-bench-power-supply-laser-diodes-img_20170312_133620-cr.jpg

55460d1490435668-bench-power-supply-laser-diodes-img_20170312_133810.jpg

55461d1490435719-bench-power-supply-laser-diodes-long-cropped-not-downres.jpg
 

Attachments

  • dark beam.jpg
    dark beam.jpg
    100.8 KB · Views: 183
  • IMG_20170312_132417 crop.jpg
    IMG_20170312_132417 crop.jpg
    67.2 KB · Views: 170
  • IMG_20170312_133620 cr.jpg
    IMG_20170312_133620 cr.jpg
    108.7 KB · Views: 174
  • IMG_20170312_132247.jpg
    IMG_20170312_132247.jpg
    64.5 KB · Views: 170
  • IMG_20170312_133810.jpg
    IMG_20170312_133810.jpg
    84.4 KB · Views: 339
  • long cropped not downres.jpg
    long cropped not downres.jpg
    54.3 KB · Views: 174
Last edited:
Nice build indeed.

Too bad the camera makes the beam look like that, but that's a problem with bright monochromatic light in general - it's not something that commonly occurs so they don't optmize camera's to render well.

Oddly this makes real images look badly photoshopped, a problem for sellers i'm sure.
 
After seeing the EEVBlog update on the PS...
I'd get one for the shop since they fixed the
problem.

Jerry
 
Very nice build Enchilada. :gj: Lovely cat BTW. :p +rep :beer:

agree!!

Found this very helpful... some here may want to start using a laser driver every time--even if the own a $$$$ Bench power supply -the authors are with a company that makes/sells PS so if anything, they would be biased and not be advising us to use a driver, even a cheap one EVERY TIME__

_________________



Powering Lasers: Evaluating Bench Power Supplies
Photonics Handbook
Supplying power to a laser is delicate work because lasers are extremely sensitive and easily damaged. A single pulse of excessive current, lasting only a few microseconds, can overheat and damage the light-emitting facets of a laser diode. While a laser diode driver is the safest and most effective method for powering a laser, in some cases, a bench power supply may be an option. The answer depends on the application and on the performance of the specific bench supply in question.

PAUL CORR AND PATRICK KLIMA, ARROYO INSTRUMENTS LLC

A bench power supply should be thoroughly tested and characterized before it is used to power a laser. The quality of the controls and safety measures incorporated into a bench supply will need to be evaluated to determine whether it is suitable.

The issues of bench power supplies

1) Dangerous current control

A laser diode is inherently a current-sensitive device, requiring constant current to function properly.

Figure 1 shows that a small change in voltage can result in large changes in current. When using a bench supply that is voltage controlled (only output voltage is adjustable), small adjustments can quickly produce unexpectedly large changes in current and easily drive the laser beyond its maximum ratings. Never use a voltage-controlled power supply.

Voltage-current performance of a hypothetical laser diode.

Figure 1. Voltage-current performance of a hypothetical laser diode.

2) Unpredictable power-up

A bench power supply is unpredictable at startup (Figure 2). Even with the voltage set point at zero, initial power-up can produce a brief, but large, current spike. This spike, which can reach very high currents, may only last a few microseconds or milliseconds but can critically damage the laser.

3) Slow response time

In response to issue two, some users may be tempted to turn on the bench supply at zero volts and let it stabilize before connecting the laser. In this circumstance, the circuit is open before the laser is connected. Even though the set point is at zero, there will typically be a small voltage present. As the connection to the laser is completed, the current may spike rapidly as the supply reacts to the nearly zero impedance load. The bench supply will respond by bringing the voltage back to the nearly zero set point indicated, but that response will not be fast enough to protect the laser.

The current response for a hypothetical bench power supply at power-on.

Figure 2. The current response for a hypothetical bench power supply at power-on.

4) Limited voltage and current safeguards

The only remaining option would be to connect the laser to the bench supply with the voltage and current set points at zero. Assuming that the bench supply did not allow a current spike at power-up, the subsequent increase in voltage and current still have an inherent risk. A bench supply is not typically designed to make the transition from voltage control to current control with the necessary speed and may fluctuate enough to damage the laser. Even bench supplies with current limits will typically not be sufficiently fast to protect the laser in an overcurrent situation.

5) Other missing safeguards

Other safeguards that will almost certainly be missing from bench supplies include:

• Poor contact protection: Many bench supplies will go well beyond their current limits when the connection is restored.

• Overvoltage protection: Shutdown of the drive takes place when voltage exceeds a preset limit.

• Shorting protection: Mechanical and electrical short of the laser when the supply is off.

• Other engineered protections, such as safe shutdown on AC power failure and electrical isolation.

With all of these potential difficulties, the only way to verify whether a specific bench supply will be suitable for an application is to fully characterize its performance.

Characterizing the performance of a bench power supply

Characterizing a bench power supply requires equipment capable of measuring fast voltage and current transients while minimizing the impact on the supply performance during the measurement process. A common practice for current measurement is to place a noninductive current-sense resistor in series with the supply and measure the voltage across the resistor to derive the current. However, adding a series resistor can change the dynamics of the loop and produce a different – and inaccurate – picture of the supply performance.

Hypothetical performance of a voltage-controlled bench power supply with a current limit.

Figure 3. Hypothetical performance of a voltage-controlled bench power supply with a current limit.
While it may not always be practical, a better choice is to use a noncontact DC current probe to measure current flow, which will not change the dynamics of the control loop and therefore not influence the measurement. In either case, an oscilloscope of at least 300-MHz bandwidth will be needed to capture any transient events. Using a current sense resistor also requires an understanding of the grounding scheme used by the power supply and oscilloscope. The ground connection of an oscilloscope is typically connected to earth ground. If the supply’s outputs are also connected to earth ground, a current loop could be created and damage the equipment. Floating the oscilloscope by disconnecting earth ground using a ground-lift plug should only be done if the operator is fully aware of the risks in doing so.

Another consideration is the load to be used during testing. Obviously, the laser itself should not be used – the whole point of characterization is to first ensure the supply is safe to use with the laser. A laser can be approximated by using high-speed rectifier diodes, with a sufficient number of diodes in series to develop the same operating voltage as the laser. Make sure to select rectifier diodes with sufficient current handling, and heat sink the devices as necessary.

The various operating conditions under which the bench supply will respond will then need to be simulated. To begin, connect a load to the power supply, and measure the current performance in the following scenarios, along with any others that may be relevant:

• Turn the power supply on and off several times.

• Start the circuit in an open condition, and then close the circuit.

• Move from one voltage and current set point to the next.

• Test an intermittent connection, such as that caused by a loose cable.

• If the power supply accepts a function generator input, test the performance of the bench supply while connected to the function generator.

• If the bench supply performs adequately under these circumstances, it may be suitable for the laser application.

When are bench supplies appropriate for lasers?

A bench power supply will typically be more successful in a high-power application (for example, with a 60-A bar laser) where power can be distributed across several light-producing facets and is therefore more tolerant of power spikes. Low-power lasers are very easily damaged and far less tolerant of spikes and poor power delivery.

In any case, careful characterization of the power supply is of the utmost importance to reduce the risk of laser damage. Ultimately, the best option may be to use a laser diode driver. A laser driver will include a variety of safeguards specific to operating a laser. Even the most inexpensive laser driver will offer superior protection, higher accuracy and better performance than a bench power supply.
Voltage-current performance of a hypothetical laser diode.
Figure 1. Voltage-current performance of a hypothetical laser diode.

2) Unpredictable power-up



*** PS
Any with questions about why diodes are NOT returnable... READ.. so many things can go wrong in a hurry.. good luck to all--hak



more or same??

having probs with the illustrations

http://www.arroyoinstruments.com/whitepapers/Arroyo-UsingBenchPowerSupplies.pdf

click to see diagrams post them for me if you can---TY
 

Attachments

  • PoweringLasers_Arroyo_Figure2.jpg
    PoweringLasers_Arroyo_Figure2.jpg
    7 KB · Views: 14
  • PoweringLasers_Arroyo_Figure3.jpg
    PoweringLasers_Arroyo_Figure3.jpg
    12.7 KB · Views: 15
  • PoweringLasers_Arroyo_Figure1.jpg
    PoweringLasers_Arroyo_Figure1.jpg
    11.7 KB · Views: 15
Last edited:
If you are contemplating using a voltage controlled power supply to test a laser diode you won't like the results. I use CV/CC power supplies that I have checked out thoroughly with a scope before using it for this purpose. It just isn't possible to use a laser driver unless it is one that is completely adjustable to test laser diodes, so I do use my PSs to do this. I have binned many dozens of diodes by wavelength this way. Jordan, DTR, has likely used his power supplies to test more diodes than I have, so if you know what you are doing and take steps to insure the supply won't damage your diodes with spurious spikes, it can be done.
 
OK --somehow I just do not feel 'lucky'

what would you do if you found a diode and had no clue which it was?? .. thus needing to determine that before you go too far.

Also as a seller would you replace/refund a buyers diode for a killed one when NOT using any driver?
 
Last edited:
If you are contemplating using a voltage controlled power supply to test a laser diode you won't like the results. I use CV/CC power supplies that I have checked out thoroughly with a scope before using it for this purpose. It just isn't possible to use a laser driver unless it is one that is completely adjustable to test laser diodes, so I do use my PSs to do this. I have binned many dozens of diodes by wavelength this way. Jordan, DTR, has likely used his power supplies to test more diodes than I have, so if you know what you are doing and take steps to insure the supply won't damage your diodes with spurious spikes, it can be done.

I fully agree.

Never use any power supply that has no current limiting feature for this - powering a laser diode from a fixed voltage is a bad idea even if the power supply is perfect due to thermal runaway problems: As a (laser) diode gets hotter, it will draw a larger current at a given forward voltage. This can cause runaway even if you power supply will keep producing exactly say 4.900 volts regardless of load current.

What you want to test laser diodes is to operate in constant current mode, and just read off the voltage as a result of current, not the other way around. You set it to 1.000 or 1.500, 3.000 amps or whatever, then turn up the voltage until it switches from CV to CC mode and leave it there.

This really is fine on most power supplies, given that you connect the laser diode before turning on the power supply. Even good power supplies often have a small buffer capacitor right at the output terminals - which is fine in most applications as it allows it to deal with inductive / rf loads better. If you just set a current and let the voltage float up this capacitor can have enough energy to kill a laser diode when you connect one after turning it on though, and similar with intermittent connections (solder the things, don't much around with croc clips).

Last thing could be power-up output spikes and current regulation overshoot. Not a problem with most power supplies, but check with a scope before risking any expensive laser.
 
In all the diodes I have tested I have only found two that were bad out of the box. This just happens sometimes and is not caused by a good working CC supply. When you have 20 diodes to bin by wavelength, it is impractical to use the drivers we have for powering these diodes to test them. I would never use alligator clips, but I do have transistor sockets that I use for the connection. This might not work well for 5 amp diodes, but for these new Sharp diodes it works just fine. I also use cube shaped module made out of copper or brass to heat sink diodes that I am testing. It doesn't require pressing the diode as it has a back that holds the diode in and transfers heat very well. I am considering adding a TEC to the modules with a feedback loop to insure the temperature I want the diode to stay at.
 
Last edited:
the circuit is open before the laser is connected. Even though the set point is at zero, there will typically be a small voltage present. As the connection to the laser is completed, the current may spike rapidly as the supply reacts to the nearly zero impedance load. The bench supply will respond by bringing the voltage back to the nearly zero set point indicated, but that response will not be fast enough to protect the laser.

What idiot wrote that? Yes, there's usually some residual voltage with a zero voltage setpoint, but it's on the order of millivolts. maybe up to around 700mV if it's really bad. But it doesn't matter how hard you try - You cannot kill a laser with millivolts, no matter the current behind it.

Many bench supplies will go well beyond their current limits when the connection is restored.

ALL current sources (with nonzero output capacitance) "go well beyond their current limit when the connection is restored" - not just bench power supplies.
 
Last edited:


Back
Top