Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Saturday January 28 2023, @05:37AM   Printer-friendly

Stick with a traditional thermal paste from a reputable brand:

[Source] Editor's take: Hardcore enthusiasts are constantly on the hunt for new techniques and methods to shave a degree or two off their operating temperatures. Most efforts bear little to no fruit, and instead serve as cautionary tales for what not to do. This is one of those examples.

A ComputerBase community member recently put several thermal pads and pastes to the test using an old Radeon R7 240 graphics card. The GPU is normally passively cooled but a fan was added to expedite the testing process. Some unconventional alternatives were thrown in to spice things up, and that's more of what we are interested in here.

The card is clocked at 780MHz @ 1.15V by default and has multiple levels of throttling that automatically dial back the clock speed and voltage supplied. If things get too toasty, it shuts down entirely.

First up is ketchup and after a five-minute torture test in Furmark, the GPU registered a temperature of 71c. Among the wacky alternatives, this was the best performer.

Standard toothpaste reached 90c after a similar run while diaper rash cream, a potato thin and a cheese slice all hit 105c and were thermally throttled.

What's your best bet in pursuit of lower temperatures? For starters, stick with an actual thermal paste. Tried and true options like Arctic MX-4 and Corsair TM30 were among the top performers in the test with temperatures of 49c and 54c, respectively. How you install the paste can also impact temps and some even claim additives like salt can help but we wouldn't recommend trying that.

Can anyone in our community describe bizarre or unlikely solutions to problems that actually work? If you are not too ashamed to admit it - any failures that you might have tried too?


Original Submission

This discussion was created by janrinok (52) for logged-in users only, but now has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 2) by driverless on Saturday January 28 2023, @07:50AM

    by driverless (4770) on Saturday January 28 2023, @07:50AM (#1289051)

    Can anyone in our community describe bizarre or unlikely solutions to problems that actually work?

    "Kill them and keep killing them until they stop coming back" solves an awful lot of problems.

    And not just at work.

  • (Score: 2) by looorg on Saturday January 28 2023, @11:27AM (4 children)

    by looorg (578) on Saturday January 28 2023, @11:27AM (#1289064)

    I'm fairly certain that nobody thought that Ketchup, Cheese and a few other odd objects were going to be doing great or the next hidden replacement for thermal paste. It's like the burger-thermal-paste test. Won't fly but could be delicious. But this was clearly made in jest. If you just take Ketchup as an example it was never going to be a good replacement -- it's slightly acidic, contains a lot of sugar and bits of various things such as tomatoes or chili or whatever you have in there and then add the usual amount of preservatives and spices and such. It's not very clean or pure in that regard. Nobody would be surprised that slices of cheese and potato didn't last very long either since they'll both be turned very charred. Toothpaste have the same issues as ketchup except less sugar in it.

    How would adding salt into the thermal paste help? Salt (NaCl) doesn't make things cooler as far as I know. It's not like they are building some kind of molten salt reactor. As the comments for the article over at the other place goes I'm more surprised they didn't just try to use more human fluids, at least they do bind a lot of water but that will boil of quickly so it will be horrible in the end and smelly.

    Come to think of it it would probably have been better if he had just squeezed the entire burger in there as a heatsink.

    • (Score: 2) by Reziac on Sunday January 29 2023, @04:00AM (3 children)

      by Reziac (2489) on Sunday January 29 2023, @04:00AM (#1289156) Homepage

      When this notion first went around the Interwebs, one of the better contenders was... Vegemite.

      --
      And there is no Alkibiades to come back and save us from ourselves.
      • (Score: 1) by anubi on Sunday January 29 2023, @11:23AM (2 children)

        by anubi (2828) on Sunday January 29 2023, @11:23AM (#1289172) Journal

        Vitameatavegamin.

        Anything that made Lucy make a face like that has Got to be good for something.

        https://m.youtube.com/watch?v=cwkzNclZR4s [youtube.com]

        ( You gotta be an old fart to remember this one )

        --
        "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
        • (Score: 2) by Reziac on Sunday January 29 2023, @03:47PM (1 child)

          by Reziac (2489) on Sunday January 29 2023, @03:47PM (#1289181) Homepage

          LOL, that'll make your CPU happy :D I think that's the face I made too! one taste, and I knew why people were trying to use it as lubricant and cleaning fluid...

          [Don't remember this one in particular, but I sure do remember I Love Lucy...]

          --
          And there is no Alkibiades to come back and save us from ourselves.
          • (Score: 0) by Anonymous Coward on Monday January 30 2023, @11:57AM

            by Anonymous Coward on Monday January 30 2023, @11:57AM (#1289270)

            I thought it might be usable for keeping my neighbor's dog from peeing on my tomato bush.

  • (Score: 2) by crafoo on Saturday January 28 2023, @11:57AM (2 children)

    by crafoo (6639) on Saturday January 28 2023, @11:57AM (#1289066)

    I knew a nice fellow that made a very pretty and functional machette from a leaf spring, angle grinder, torch, and bucket of old oil. he must have had some nice sharpening stones somewhere too. it was very very sharp with a durable edge.

    I rigged up a diesel fuel filtering system from spares and whatever I could get in a fairly remote resort town, off season.

    most people who work in the field, in commercial construction in remote areas, or who do any kind of transportation in remote areas - they all consistently have the most interesting daily life stories.

    • (Score: 2, Funny) by redback on Saturday January 28 2023, @02:29PM

      by redback (1011) on Saturday January 28 2023, @02:29PM (#1289075)

      are you lost?

    • (Score: 3, Insightful) by looorg on Saturday January 28 2023, @02:41PM

      by looorg (578) on Saturday January 28 2023, @02:41PM (#1289076)

      Sounds like some general junkyard forge technique. The leaf spring should be good steel and be able to hold an edge, form/cut it into shape with the grinder, heat what will be the edge with the torch, quench it in the old oil and sharpen it with the grinder again. You can sharp it with the grinder if need be, after all it's a machette so it's not supposed to be a razor but use it's weight to do the cutting.

      It's not that people that live out in the boonies have the more interesting stories, they just have different stories about how they use every day items to do thing or repair things or do general MacGyvor things. City-people have those stories to, they just do different things. Perhaps not as essential things but fun things. Just look at all the IKEA-furniture-hacking etc.

  • (Score: 3, Interesting) by anubi on Saturday January 28 2023, @12:00PM (4 children)

    by anubi (2828) on Saturday January 28 2023, @12:00PM (#1289067) Journal

    I have been using this stuff for years to mount 10W, 30W, and 100W led chips to heavy metal things for years to create unique (usually outdoor) lighting fixtures. Things like worn disk brakes, old cooking pots and pans, structural members, old farm implements, whatever looks arty and would hide the emitter from line of sight, and usually powered from a 12 volt input, adjustable current output DC boost converters. As a bonus, I often can throw in a float battery charger and a lead acid truck battery that can still power the lighting for several hours should the power fail. A restaurant had commissioned me to build one for them, as it also acts as an emergency light, yet kinda free battery option as the owner uses his old truck battery to run it. Its iffy starting the truck but runs the lights just fine. So the truck gets the new battery, the lights get the old one.

    I run the LED at maybe 20% of rated power, LED glued to whatever massive metal thing, with additional clear silicone sealant gooped all over the top of the LED for both insulation and light diffusal. If I ever have to change the LED out, it will be a hammer and chisel job.

    I just use lots of chips, and big ones, and drive them very conservatively. They are hard to get to once the light fixture is in place. It would not be fun to fix it once it's up.

    Its been working 24/7 for over five years now without a single failure. Occasionally, we pull the plug on the battery charger. It will run the rest of the night on the old truck battery just fine.

    Incidentally, when mounting to old iron stuff, those little neodymium magnets make great clamps to hold the LED in place while the first dab of silicone between the LED and the mounting surface sets. No drilling.

    I am not confident this GE RTV stuff ( The stuff WalMart sells ) has any sort of exceptional thermal characteristics, but so far I haven't had any problems with LED failure due to thermal cycling causing mechanical mounting failure. And it's kept the water out of my solder joints to the LED power tab wiring.

    If anyone else has observations on using silicone sealant for it's thermal properties, please feel free to chime in. When I build things, I want them to last damn near forever. Way too much work to maintain poorly designed crap, which I already see way too much of.

    --
    "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
    • (Score: 4, Informative) by VLM on Saturday January 28 2023, @03:48PM (1 child)

      by VLM (445) Subscriber Badge on Saturday January 28 2023, @03:48PM (#1289081)

      No disagreement with anything you wrote, but chiming in that you can buy acid cure and neutral cure silicone.

      The acid cure stuff is about as acidic as vinegar (it don't dissolve your fingers, but it will eventually corrode electronics) and helps it stick better to kitchen and bathroom tile type stuff.

      As long as you don't buy "kitchen and bathroom" branded stuff it'll be OK for electronics.

      I've had mixed results removing old silicone with wd-40 (presumably is the kerosene) and isopropyl alcohol. Will not melt it off but it'll help clean the surface so try such that sealant application 2 will stick as well as attempt 1

      For various safety reasons as this is up on an ham radio antenna the iso is less slippery than wd-40 when you're working on a roof and climbing ladders and trying not to drop stuff, which may or may not be relevant to your led lights. I think if you're fixing something on the ground maybe wd-40 helps wipe silicone off better than iso but its not a huge diff.

      My experience outdoors is silicone looks pretty bad after a decade or so protecting antenna stuff, but maybe OK for your LEDs. If you build your own microwave ham radio amp or preamp you can mast mount it up at the antenna in a box thats 99% sealed except a hole at the lowest point for condensation to drip out if any gets in. It works "pretty well" in that application.

      • (Score: 3, Informative) by RS3 on Saturday January 28 2023, @06:20PM

        by RS3 (6367) on Saturday January 28 2023, @06:20PM (#1289096)

        Super awesome info! I've always liked silicone "RTV" adhesive / sealants.

        I've always noticed one silicone RTV sealant smells like acetic acid, and the other maybe ammonia? The acetic-acid stuff seems to last forever in the tube. I might be embarrassed to say how old a couple of tubes are that I have- 40 years or more, and it still flows and cures. I wasn't aware it would remain acidic. Yikes. Good to know, thanks.

        The much more common ammonia-smelling stuff is "use it or lose it". It may still flow after the exp. date on the tube, but it's probably not going to cure- it'll remain goo. Whenever I need some, and find I have a tube that's past the date on the tube, I'll run a test bead and let it sit overnight to see if it'll cure. Normally the stuff cures to the touch within a few hours (not in cold temps), but sometimes if it's a bit old it'll still cure but slower.

        I wonder which variety weathers better?

        I've dismantled several things that were siliconed up and outside for very many years, including antenna preamplifiers. It looked weathered, but held strong. I forget how I dealt with it- probably various scrapers and X-acto knives. I'm not aware of anything that will attack it nor dissolve it. Another good thing I like for cleaning things is brake cleaner, but there are many types / formulas and as my mom would tell us, do it outdoors or somewhere very well ventilated. ISO or denatured is your friend otherwise. And yes, WD-40, which IIRC is mostly kero?

        It withstands very high temperatures and doesn't seem to burn. IIRC I think it was developed by or for NASA in the 60s.

        Silicone RTV sealants are extremely good electrical insulators. I've hit them with 30KV and they're's no conduction or arc-through.

    • (Score: 3, Informative) by RS3 on Saturday January 28 2023, @06:32PM (1 child)

      by RS3 (6367) on Saturday January 28 2023, @06:32PM (#1289098)

      Wow, cool creative ingenuity. I especially like the idea of using the silicone as a light diffuser.

      Only thing I'll add: silicone RTV sealant is a very good insulator, both electrical and thermal, so your LEDs could get hot if run near their design limits, but running them at low currents is your friend here.

      I've noticed many (most?) LED bulbs are being run HOT which greatly shortens their lives (right?) Manufacturers are trying to get the most lumens for the lowest cost and don't care about product life. Some brands have fairly complicated heat sink systems and external finned heat radiator / convectors- some of which will burn you pretty badly.

      Dimming them a little bit should make them last the 20+ years people talk about. For truly non-dimmable ones (constant-current regulator) it should be possible to get in and change a resistor, but you'd have to have a schematic, reverse-engineer, or play around. Your method of simply starting with the LEDs and separate power supply is spot-on.

      • (Score: 1) by anubi on Sunday January 29 2023, @10:29AM

        by anubi (2828) on Sunday January 29 2023, @10:29AM (#1289168) Journal

        Thanks...I was afraid RTV might not be the best thermal conductor, but like every other design I have ever done, everything is a tradeoff one way or another. Hence using the hermetically sealed led assemblies and boost converters ( boost converters are natively a current source anyway, and failure of the switching transistor will result in something I can easily fix, instead of a blown led array hanging in a really hard to get to place! )

        I have noticed too how many led lighting fixtures seem to be designed by business school graduates, who have never had to fix stuff at the customer site, outside, usually in wet, cold weather. That was the first thing Chevron ( the oil company ) taught me as a newbie engineer. They let me make a mess, then I had to go fix it. Taught me so much about the real cost of pennypinching.

        What came out of it is I still like to make stuff, that may look like junk ( the art thing ), but designed to be easily maintainable, but should never need any. It should run damn near forever.

        If anything, give it the old truck battery, and give the one it had for the core charge. I've had old lead acid batteries on constant float run my garden shed for 10 years, albeit they couldn't perform anyway near the capacity of a new battery, but they did light up the shed, from battery, as I am paranoid about having line AC voltage anywhere near that steel shed, especially with the Santa Ana winds, rain, leaves, and other ignitables nearby. 400mA at 9 volts ( PTC + series ballast ) powers up four 10w rated ( but driven at 0.9 watt each ) lights up the shed nicely

        The LEDs are simply glued onto the central steel rail that holds the roof up.

        A lot of stuff I see looks good, but doesn't last worth a hoot, and nigh near impossible to maintain. It saddens me to see someone put business signage up, and in no time, it blows out, and someone has to get out there with a cherry picker to service it

        With all the component parts so available these days, it is so do-able to be arty and make all sorts of custom lighting, all run off a 12V car battery on a float charger, so that everything is intrinsically safe, yet very robust and also function with 100W PV panels so that the illumination functions, ( even USB type power ports ) are easily implemented. I never know when the power is going to fail, and I don't like pitch black surprises.

        Even if I am not using it for direct illumination, I still like certain things on all night just for navigational markers, so if everything goes pitch black, I can still recognize my "beacons" and orient myself.

        --
        "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
  • (Score: 4, Informative) by VLM on Saturday January 28 2023, @03:39PM (8 children)

    by VLM (445) Subscriber Badge on Saturday January 28 2023, @03:39PM (#1289079)

    Can anyone in our community describe bizarre or unlikely solutions to problems that actually work

    Their experiment would have been more easily reproducible using a power resistor.

    Their experiment would have been simpler with a positive temp coeff thermistor. Its bizarre and unlikely in that nobody talks about it much but its a simple analog feedback loop where power dumped in the thermistor makes it heat up, making the resistance increase, making the current drop, making the wattage drop, making it cool down... and its usually underdamped enough that sticking one on a quartz crystal (electronic, not woo woo crystal) will stabilize the temp. Its esoteric in that there's an underground of mostly ham radio guys who know this one weird model at Digikey is exactly perfect size and stable temperature for standard -49 size crystal osc at 13.8V in and around 60C, vs those metal DIP can oscillators work better (IMHO) with disk shaped PTCs. Its all very esoteric. I have a bag in a drawer that must be a quarter century old so I have no current (oh the pun) idea, but I'm sure this is still state of the art for homemade xtal ovens for stability.

    Anyway slap that PTC thermistor on, send in a stabilized 13.8V or stable 5V or WTF as long as its stable, do it indoors with no wind and the current the PTC is dumping will indicate how good the thermal bond is to the heatsink.

    More "professional" EEs will make xtal ovens using a forest of FETs controlled by op amp comparators and strange temp references, which when done competently will be more temp stable AND stabilize faster AND have all the advantages of a full PID loop. Less competent EEs will F it up worse than slapping a PTC on it. Most applications, frankly, are NOT demanding, and as such you just slap a PTC on the xtal and call it good nuff for govt work.

    I'd like my SDR dongle to drift less. Not trying to build a NIST traceable melting point apparatus tester, LOL.

    If you really want to cheap-ass it you don't need to generate the heat energy electrically, just fill a soda can with boiling water and graph the decline of temp depending on compound, as long as the baro pressure is constant-ish the BP of water will be constant-ish and you can use a fixed mass by filling the cup to the brim each time.

    • (Score: 2) by RS3 on Saturday January 28 2023, @06:36PM (3 children)

      by RS3 (6367) on Saturday January 28 2023, @06:36PM (#1289100)

      Okay- nerd test here: "Can I use a 5.6V zener diode to measure temperature?"

      :-}

      • (Score: 2, Funny) by anubi on Sunday January 29 2023, @10:47AM (2 children)

        by anubi (2828) on Sunday January 29 2023, @10:47AM (#1289169) Journal

        Did you pick that voltage on purpose?

        How about a nice game of voltage reference?

        --
        "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
        • (Score: 3, Informative) by RS3 on Monday January 30 2023, @12:26AM (1 child)

          by RS3 (6367) on Monday January 30 2023, @12:26AM (#1289221)

          It was a little quiz. I was hoping someone would know. It may vary with specific zener manufacturing, but right around 5.6 V zeners become temperature stable. IE, there's almost no voltage change over reasonable (non-destructive) range of temperatures. So using a 5.6V zener with a simple NPN makes a very nice simple and stable 5V regulator, which I believe is why 5V was chosen for TTL, and is still a very strong and common power standard (USB, for ex.)

          Search down a bit to 5.6 you'll get some better explanation:

          https://en.wikipedia.org/wiki/Zener_diode [wikipedia.org]

          • (Score: 2, Interesting) by anubi on Tuesday January 31 2023, @01:04AM

            by anubi (2828) on Tuesday January 31 2023, @01:04AM (#1289406) Journal

            A lot of my 60's and 70's stuff was 7400 TTL.

            It needed clean 5 volts DC.

            I had plenty of power available .. at about 8 - 15 volts DC. With Nasty voltage spikes on it. Inductive load kickbacks. Alternator load dumps. Yup, the Truck Battery

            All that TTL added up to a lot of current.

            And the 7805, which was just coming out, could not supply much current, when it had generally 8 to 9 volts across it, series dropping 13 - 14 volts to 5.

            But we had lots of nice big 2N3055 NPN TO-3 silicon power transistors that had just come out. I seem to remember Texas Instruments made them. I ordered cases of them. Often had one 2N3055 driving a dozen more 2N3055. I could get 20 amps of 5 volts, reliably, out such a contraption, after adding the requisite base ferrite beads and compensation capacitors to keep the thing from self oscillation in the hundreds of KHz/low MHz region. If it oscillated, my trusty little Triplett 630 would still say the DC was fine, but my logic became most, shall I say, illogical.

            Yup...used 6.2 volt zeners to set base voltage at the Darlington array, and I think it was a 4.7 volt in a potentiometer circuit that would set the drive to trip off the overvoltage SCR shunt.

            I used fuses on each emitter, not just as fuse, but their resistance also acted to force the dozen or so 2N3055 to equally share current. If the overcurrenr tripped, I got to change all the fuses

            But that was a helluva lot easier than changing all the logic IC's!

            Those were the days!

            Today, it's as simple as an off the shelf switcher supply, but my whole 7400 logic thingies would be 4000 CMOS and an Arduino, running off an 18650 cell.

            --
            "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
    • (Score: 3, Informative) by RS3 on Saturday January 28 2023, @07:14PM (3 children)

      by RS3 (6367) on Saturday January 28 2023, @07:14PM (#1289107)

      Their experiment would have been more easily reproducible using a power resistor.

      I'll suggest a power BJT (bipolar junction transistor (NPN, PNP)- not FET). Measuring the forward bias voltage across a semiconductor PN junction is a very good way to measure temperature. You can easily know the power into the transistor (collector amps X voltage across collector to emitter), measure BE volts, and know the temp.

      It's important to know the BE vs. temperature curve and watch it during the test so you don't fry the DUT (device under test). Maybe a simple circuit with some opamps (one used as a comparator) and a temp. stable reference voltage would limit base current and prevent destruction.

      • (Score: 2, Informative) by anubi on Sunday January 29 2023, @11:12AM (2 children)

        by anubi (2828) on Sunday January 29 2023, @11:12AM (#1289170) Journal

        Stephen Woodward

        PTAT

        My paper books still work, but it's not showing up for me. Electronic Design magazine, also EDN magazine

        Beautiful circuit. Accurate and cheap. Easily expandable.

        Can't seem to find it anymore?

        --
        "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
        • (Score: 2, Interesting) by anubi on Sunday January 29 2023, @11:50AM

          by anubi (2828) on Sunday January 29 2023, @11:50AM (#1289173) Journal

          Also a nice way of making a precision temperature reference using the power transistor as a heater to keep, say, a crystal, at a certain temperature... As already mentioned.

          --
          "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
        • (Score: 2) by RS3 on Monday January 30 2023, @12:30AM

          by RS3 (6367) on Monday January 30 2023, @12:30AM (#1289224)

          That's awesome, thanks. I don't remember if I ever heard of the PTAT circuit, but I find lots of references online. If I'm understanding it from a quick glance, it's a current mirror that has one side sensing temperature, so it pretty greatly amplifies the variation. In other words, it's a nice simple and stable instrumentation amplifier.

  • (Score: 5, Interesting) by RS3 on Saturday January 28 2023, @07:03PM (1 child)

    by RS3 (6367) on Saturday January 28 2023, @07:03PM (#1289104)

    Can anyone in our community describe bizarre or unlikely solutions to problems that actually work?

    MS Windows? Oh wait, you said "actually work". Nevermind.

    But seriously, it's interesting to watch the world grab onto EE-type things, like heatsinks and heatsink compounds, and not really fully learn and understand a few of the most critical concepts.

    The jobs of gaskets, sealants, and heatsink compounds is to fill the microscopic voids between two surfaces that ideally would be as nearly smooth and perfectly flat as possible. In some of the highest pressure and vacuum systems, surfaces are flattened (often lapped), polished, and clamped together with nothing needed. In fact, at extreme pressures any kind of gasket, sealant, or O-ring would simply blow out anyway.

    I didn't see anything in the article about thickness. Heatsink compound is not as good a thermal conductor as the things you're trying to thermally join- usually copper-based or aluminum. The point is: thinner is better. Recommended thickness for heatsink compound is 100 microns (0.004")- about the thickness of notebook paper.

    I can't tell you how many CPU, GPU, chipset, etc. heatsinks I've removed to find a super thick layer of heatsink compound, sometimes with voids (air gaps).

    Some heatsink compounds are electrically conductive (Artic Silver, etc.). They're great thermal conductors, but when they get onto the motherboard, often cause shorts and small board-killing fires. Some CPUs used to have electrically connected capacitors soldered on the top surface (AMD?) and some people would slop the compound all over, or just use too much and it would ooze out and get on the caps and yep, smoke, fire, destruction.

    Some heatsink compounds are much too thick and dry and they don't flow out. Again, the goal is the thinnest film possible between the two surfaces. I've sometimes blended thick ones with some soupy new stuff I have and get a nice consistency paste.

    I always pull the two things apart and look to see if contact is full, even, and very critical: there is nothing that would prevent the two things from coming into full contact. Sometimes the tiniest bit of grit gets in. I always rub both surfaces with my finger, and if I feel anything, any kind of grain of grit, wipe it off and start over.

    So again, TFA doesn't mention thickness, but it's a very critical factor in these considerations and tests.

    • (Score: 2) by RS3 on Monday January 30 2023, @12:35AM

      by RS3 (6367) on Monday January 30 2023, @12:35AM (#1289225)

      TL; DR:

      1) Very clean flat 2 surfaces. Clean with alcohol or other solvent if needed.
      2) Clean heatsink compound. I use a finger to apply. If I feel any grit, wipe clean, start over.
      3) Press / clamp together, take apart, see that intersecting surfaces are fully covered, no voids, and very obviously very thin when clamped: no clumps / high spots.
      4) Wipe a bit again, if clean, clamp in place, you're done.

(1)