[HN Gopher] Replacing the bad flyback transformer in Apple's Stu...
___________________________________________________________________
Replacing the bad flyback transformer in Apple's Studio Display 17"
Author : riveducha
Score : 56 points
Date : 2023-08-05 18:41 UTC (4 hours ago)
(HTM) web link (www.riveducha.com)
(TXT) w3m dump (www.riveducha.com)
| aurizon wrote:
| That 'snap' is the arc and the voltage = zero until it gets going
| again - repeat. This is common on old monitors as dust builds up
| and often a fix can be done by a careful cleaning plus
| application of 'corona dope' along spark paths. Turn it on in a
| dark room and watch/listen for the spark. Dope that path and
| alternate paths and dry with hair dryer. Test. There are
| universal repair flybacks - select what seems to fit. Usually the
| existing controller will serve - computer monitors are far
| smaller than most TV's of that era. Hardest thing is finding new
| old stock. To save the nation new flybacks can be wound - complex
| but doable. Hardest problem is adapting the space and circuit.
| The flyback can be wired point to point with hookup wire. Only
| the HV wire needs to be HV wire. Then you can zip tie the flyback
| in whatever space and dress the HV with respect.
| ivraatiems wrote:
| I have a fully functional third-gen iMac (the "jelly bean") that
| needed its inner bezel replaced. Sourcing a non-broken bezel was
| one big job, but even after I did, finding someone with the
| knowledge and skill to repair it safely was a task. I wasn't
| willing to do the repair myself, it's just too dangerous for an
| amateur. These machines are supposed to have components that
| automatically discharge the screen after they're removed from
| wall power, but on a 20+ year old machine, no way to know whether
| that was working. You have to touch the screen and fasten it to
| the inner bezel, so repairing it without proper discharge was a
| non-starter.
|
| I called around to various TV repair shops and was generally told
| they couldn't do it, or that it'd be prohibitively expensive.
| Finally, an electronics surplus store in my area referred me to a
| gentleman of the old-school AV hippy variety who knew how to
| repair all sorts of old equipment. He was able to construct a
| tool to ground the machine while working on it and we did the
| repair successfully together. It was an adventure!
| Scoundreller wrote:
| > The focus knobs should be tuned so that the screen image is
| sharp. I don't really have a good set of steps for this, so just
| play around with them until the monitor is as sharp as can be.
|
| This explains my methodology for a lot of things in life
| nyanpasu64 wrote:
| I haven't tried focusing a CRT with two focus knobs (horizontal
| and vertical?), but from my general experience with focusing a
| CRT oscilloscope and tuning VGA signal sharpness on CRT and
| LCD, one good way to focus is to create three B&W images of
| vertical stripes, horizontal stripes, and a checkerboard. When
| showing the vertical stripe image, tune horizontal focus until
| as much of the screen as possible is in-focus. Then repeat with
| the horizontal stripe image and vertical focus, then double-
| check on a checkerboard and regular desktop usage.
| vondur wrote:
| We had a bunch of these monitors back in the day, don't remember
| any of them dying early. I was always afraid of taking apart CRTs
| thanks to those transformers in them. People used to overclock
| monitors back in the day to increase the refresh frequency.
| mhb wrote:
| The capacitors are the things to be concerned about.
| wkat4242 wrote:
| Nice job and write-up.
|
| As the author says CRTs are dangerous. Even when they're off they
| can still carry insanely high voltages. The flyback transformer
| is one of the more dangerous components of them.
|
| I'm great at electronics repair and I tend to avoid them. Luckily
| these days that's easy :)
|
| If you don't know what you're doing, you shouldn't attempt it.
| Just get a more technical friend to do it.
|
| If you do know what you're doing you probably wouldn't need this
| guide in the first place :)
| glhaynes wrote:
| I've never worked inside a CRT (but may need to soon) and am
| _not_ great at electronics repair, but I 'm a little surprised
| you describe yourself as having a lot of skill but still tend
| to avoid them. From my understanding, it's straightforward to
| reliably discharge them and make them safe to work on. Do you
| think it's easier to get this wrong than I've understood or do
| you just feel uneasy about it in general? Just curious!
| rasz wrote:
| Im an EE that never worked on CRT and declined plenty of
| requests. I also like to think Im quite mechanically
| inclined, but I would _never_ go anywhere close to a split
| rim being serviced. There are things just not worth the risk
| in life, like jumping head down into unknown stream, or
| skating down handrails.
| toast0 wrote:
| One problem is there are a lot of adjustments that are really
| best made while the CRT is on. Another problem is CRT circuit
| boards experience a lot of heat, so repair work can be
| frustrating; it's hard to replace components and what not
| when traces have lifted.
| wkat4242 wrote:
| It's just the risk. Most components in a modern TV are very
| safe to work on. Even the PSU is not that bad. CCFL backlight
| had voltages only in the hundreds of volts, and modern LED
| backlight is even lower. A CRT is totally different business.
| The author of the article pointed out another thing, wire
| insulation properties at such high voltages.
|
| There's just a lot to consider, and extremely high voltage
| stuff isn't really my area. I tend to do more low voltage
| stuff. When things go wrong with high voltage they tend to go
| wrong pretty spectacularly, and can cause serious bodily harm
| too. I just prefer not to take the risk. I'm similarly
| cautious with Li-ion and Li-Po battery charging circuitry. I
| don't like to mess around with those and if I repair them
| it's purely a like-for-like repair.
|
| Another thing is that when I started with electronics CRTs
| were still very common and I wasn't skilled then, as such I
| had a lot of respect for them and a lot of caution. As my
| skills grew, CRTs became obsolete so I never really got
| comfortable with them and the apprehensive feeling remained.
|
| But I think I have skills with electronics yes, I repair a
| lot of electronics, designed some of my own and I also built
| and modded some radios, and came up with some of my own mods
| (I have a ham license too).
| sneak wrote:
| Yeah, this webpage really should have a big prominent warning
| label at the top.
|
| The flyback transformer can fucking kill you, and anyone who
| needs this webpage or is interested in the instructions (ie
| someone not working professionally in electronics repair and
| aware of such hazards) is less likely to know that.
| amelius wrote:
| I've always wondered why modern multimeters don't have a
| "discharge" option.
|
| I.e., you select the discharge mode, put the probes on a
| capacitor, and watch the voltage go down, while an internal
| current limiter makes sure that it happens at a safe rate.
| nickt wrote:
| The probes aren't big enough. That's why it's often a big
| screwdriver and jump lead to earth.
|
| The better option is the Fluke 80K 40.
|
| https://www.fluke.com/en-
| us/product/accessories/probes/fluke...
| Dylan16807 wrote:
| Not big enough in what way? A normal probe should be able
| to handle the distancing and the small current fine. Maybe
| it would lack insulation but I wouldn't use the word "big"
| for level of insulation.
| kens wrote:
| Typical probes are rated for 600 volts (CAT III), so
| you'd have insulation rated for 600 volts between your
| fingers and 30,000 volts, which is kind of hazardous.
|
| Here's a recent Twitter video showing what happens when
| you combine a flyback transformer and faulty insulation:
| https://twitter.com/zh1nu/status/1673487720780529664
| tgsovlerkhgsel wrote:
| For lower voltages, some do ("low impedance measurement" or
| LoZ). Not specifically meant for discharging capacitors but
| should work for that too.
|
| But this only works up to the voltage the multimeter is rated
| for (i.e. usually 300 or 600 V). If you wanted to make it
| safe to put 30 kV across the multimeter, it'd have to be
| designed with sufficient insulation etc. for that.
| formerly_proven wrote:
| An electrician's voltage tester (the not-fake ones) can be
| used for this, for off-line filter caps anyway. 30 kV, not so
| much.
| the_third_wave wrote:
| Ah, electronics hobbyists these days... I repaired my first TV
| at 16yo, it was a Nordmende B&W set with a resistor directly on
| the output of the flyback transformer. That resistor was broken
| which led to a nice corona discharge lighting up the inside of
| the set, looking at it was like looking at the set of a 30's
| Frankenstein movie. This set me up for a 'career' of repairing
| televisions and (later) monitors while at university where I
| came across all sorts of interesting light/sound/smoke effects
| from those 35kV flyback transformers with cascade circuits
| (called 'tripplers' since they were used to raise the output
| voltage) on large-tube sets.
|
| I never got zapped which is a good thing since yes, that can be
| unpleasant and - depending on when and where it happens -
| dangerous. Discharge those tubes and cascades before you fiddle
| with them, use a screwdriver connected to a ground wire which
| is connected to the metal tube frame. If you want to do it
| 'professionally' you'll want to put a HT resistor in that wire,
| otherwise just zap the thing by sticking the screwdriver under
| the plastic cap on the high tension connector at the tube.
| dougg3 wrote:
| Thanks for sharing! I had a brand new one of these displays in
| 1999 when they first came out and it developed the same problem a
| year later. The replacement from MacMall's warranty company
| (which was graphite instead of blue) also eventually developed
| the issue too. Good to know there's a solution if I ever decide
| to fix it.
|
| I think Apple should have recalled these.
| riveducha wrote:
| The 21" version was supposed to be solid but I have never seen
| anyone with one or seen it for sale. The price tag was probably
| too steep compared to the faulty 17" one. Sad to hear that the
| graphite model was just as bad though.
| metadat wrote:
| I love this kind hardcore repair hacking to keep hardware going.
| There is no logical reason to spend time on this, yet someone
| did. Big respect to the author, you are inspiring!
| veave wrote:
| I find your accent soothing.
| andyjohnson0 wrote:
| From what I understand of this, the danger comes from charge
| stored in large capacitors. So what is the longest that these
| capacitors can hold a significant charge if the crt is left
| disconnected from mains power? Hours/days/weeks? Is simply
| waiting long enough a reliable way to render a crt safe to work
| on?
|
| (Irrespective of the answer to this, there is no chance at all
| that I would ever mess with crt circuitry. But I am curious.)
| ComputerGuru wrote:
| I got a really nasty shock from a capacitor in a flash (as in
| camera) that was powered down for weeks, back in the 90s.
| fragmede wrote:
| The sad reality of externality-removed capitalism is why would
| you do that? In our disposable economy, a new monitor can be had
| for _cheap_. Repairing things is for the poor or nerdy. which I
| resemble the latter pay. Incentivizing fixing things, rather than
| buying a new shiny flashy thing is not a thing that we do well.
| Oh well. Hopefully room temperature superconductors and solar and
| batteries can lead is to a place of victory for the environment
| in place of coal.
| jrmg wrote:
| Even with the externalities accounted for, if you must use a
| monitor, it might be better to buy a new one than repair the
| old one.
|
| The replacement parts have a footprint, and a newer monitor
| might use so much less power in operation that it's more
| efficient to make and use it than to continue to use the older
| CRT.
|
| I really wish there were an easy way to work out if this was
| where the 'efficiency crossover happens in X years of use'
| point was. For example, I have a perfectly usable Thunderbolt 1
| [edit: Thunderbolt 1 _dock_ ] that meets my needs. But it uses
| 5-10W of power more than a new Thunderbolt 3 or 4 one would.
| Should I continue to use it or switch to a newer one?
| JonathonW wrote:
| It'll depend on your usage pattern and local cost of
| electricity-- if you're running that device that uses 10W
| more power 24/7, that works out to 87.6 kWh of additional
| power usage over the course of a year, or about $15 at the
| current average cost of electricity in the US ($0.17/kWh as
| of June 2023). If that's a monitor, you'd have to run it for
| a pretty long time before your break-even point on purchase
| cost alone (10+ years on a lower-end monitor, and probably
| not in your lifetime on something from Apple).
|
| If your usage isn't continuous (which it probably isn't),
| that'll reduce your electricity cost and extend your break-
| even point even further; if electricity costs more in your
| region, it'll shorten it (average EU energy prices are around
| double that of the US IIRC, and some countries are much
| higher). But it's still going to take a while to break even
| if the efficiency improvements you're considering are in the
| 5-10W ballpark; that's not all that much power in the grand
| scheme of things.
| xur17 wrote:
| A rule of thumb I've always used is that 100W for a month
| costs roughly $10.
| toast0 wrote:
| A new monitor can be had for cheap, yes. But a new CRT monitor
| cannot be had for cheap. If you have reason to prefer a CRT,
| repairing a monitor you have is probably the least cost option
| towards getting a working CRT.
|
| Many people don't have reason to prefer a CRT, and that's fine
| too. Lots of non-CRT options out there to use.
| Hizonner wrote:
| > Opening up CRTs is dangerous. Do this at your own risk.
| Considering hiring a professional instead.
|
| Hire a professional.
|
| To fix a quarter-century-old monitor that was nothing special
| when it was new.
|
| Why? I mean, I wouldn't bother fixing the thing anyway, but
| hiring somebody is going to cost you more than a much better
| brand new monitor.
___________________________________________________________________
(page generated 2023-08-05 23:00 UTC)