> Stick a heatsink on it. Job done.
It's already got a goodly sized heatsink on it. In fact, I couldn't
find anything bigger in the TO-220 category. Operating temps turn out
not to be as bad as I thought--about 124F at most. Perhaps the
original regulator was just defective. I'll never know.
I will probably just run a fan from the other, less loaded regulator
and suspend it over the circuit board. That would be a lot less work
than milling some scrap heatsinks from old PC power supplies down to
fit.
William
> Why would you care so much about some Korean POS?
Well, because I can? Because I like 'em? Because the internal design
is pretty clean and straightforward? Because I like their sound and
think they're a pretty damn good modern receiver? Maybe because it's
just one other thing that won't end up in the landfill? Has the
purpose of this group changed since I was last here?
That's not the question I asked, or the information I wanted. Pulling
out my IR thermometer reveals that the regulator really isn't getting
as hot as I thought it was...about 124F or so at its hottest point.
I'll not worry about it further, though I probably will add a fan
powered by the lesser loaded regulator.
William
Sherwood makes receivers for many others, and there is a common problem the
past several years with cheap fixed regulators failing. Mostly 79 series 12
and 15 volt regs but also 78 series positive types. They are heatsinked, but
fail anyway.
Mark Z.
Your purpose in coming here seemed quite quixotic, not to say
masochistic, and therefore provoked curiosity:
"Suffice it to say that I've wanted to get my hands on a failed
Sherwood RX-4105 or RX-4109 stereo receiver to see just what it is
that kills them...abuse, misuse, weak parts, bad engineering/quality
control or something else. "
Hard to imagine anything other than frequent failure would provoke
this quest. A likely parallel in the automotive world would be a
mission to isolate failure modes of the Yugo.
Regarding the purpose of the group: Generally people come here to get
information to help them repair various pieces of electronic
equipment, often because their livelihood depends on it. Instead, you
want advice so you can redesign this run of the mill receiver for
greater reliability.
>
> That's not the question I asked, or the information I wanted. Pulling
> out my IR thermometer reveals that the regulator really isn't getting
> as hot as I thought it was...about 124F or so at its hottest point.
> I'll not worry about it further, though I probably will add a fan
> powered by the lesser loaded regulator.
>
Why clutter up such a clean and straightforward internal design with
adequate cooling?
(snip)
>What I really want to know, though, is whether or not a drop-in
>replacement with more current delivering capability than the LM340 or
>78xx series exists. I've looked halfheartedly over the years but never
>found anything. I could always build a more capable regulator board
>and hack it in there, but I don't really feel like doing that. A fan
>would be easier and faster.
Alternative 3-terminal linear regs will still dissipate the same
amount of heat regardless of their current capability.
If the problem is in the load (i.e. faulty components) then that
should be where you focus your attention.
If the real problem is the load current and the consequent thermal
issues (i.e marginal/poor design vs SOA) then measure the actual load
current and consider a drop-in switcher such as the (TI) PT78T112V or
(RECOM) R-7812.
Actually, it sounds more like they blew the design by having too
much input voltage for the output voltage and current.
As a modest example of what I mean:
18V input 12V output at 1 amp. This is 6 watts dissipation at
the regulator package.
24V input 12V output at 1 amp. This is 12 watts dissipation
at the regulator package.
<http://www.datasheetcatalog.org/datasheets2/52/529144_1.pdf>
Doesn't specifically give a voltage drop vs output current
derating curve. But I would think that adding a bit of series
resistance to the input side and letting some of the waste
heat go away there would go a long way towards extending the
life of the regulator package.
Jeff
--
"Everything from Crackers to Coffins"
"William R. Walsh" <wm_w...@hotmail.com> wrote in message
news:32e0bfb3-bb79-4af3...@j19g2000vbn.googlegroups.com...
Oops, should have read your post properly, it was a long hard day yesterday
and today I have aches and pains to remind me.
Gareth.
Gareth.
>
>What I really want to know, though, is whether or not a drop-in
>replacement with more current delivering capability than the LM340 or
>78xx series exists. I've looked halfheartedly over the years but never
>found anything. I could always build a more capable regulator board
>and hack it in there, but I don't really feel like doing that. A fan
>would be easier and faster.
These will go up to 3A.
http://search.digikey.com/us/en/products/MIC29300-12WT/576-1118-ND/771587
Bottom line is that it's the heatsink and the voltage in that will
determine whether you'll actually get 3A.
> Your purpose in coming here seemed quite quixotic, not to
> say masochistic, and therefore provoked curiosity:
I've never been described in quite that way before. ;-)
> Hard to imagine anything other than frequent failure would
> provoke this quest.
To the contrary...I have a number of these sets used in various
applications. All are well treated and used under reasonable
conditions. I've never had a failure, even when making heavy demands
from the set. The one time I accidentally shorted a pair of speaker
wires, the set's protection functioned just as it should and there was
no lasting damage.
Yet I look around on the 'net and see people saying "well, the
Sherwood unit blew up, blew sparks and smoke, shuts off past a certain
volume level or just shut down one day, never to turn on again". What
few repairs I've seen done suggested that blown final transistors or
bad filter caps were the cause. My experiences don't align with what
people were saying.
I have no particularly good reason for wanting to know other than to
say I am eternally curious. "Why?" is one of my favorite questions. As
proof that I have no shortage of obnoxious opinions, troubleshooting
and repairing this receiver was arguably more worthwhile than watching
what passes for television these days.
> Regarding the purpose of the group: Generally people come here to get
> information to help them repair various pieces of electronic
> equipment, often because their livelihood depends on it. Instead, you
> want advice so you can redesign this run of the mill receiver for
> greater reliability.
Not totally. I wanted to share what I found, in the case that other
people here, who might someday be faced with the same situation, could
potentially find a solution. Though I've not been here for a while, I
am not new to the group or its purpose. Your initial response was
taken to mean more of a "why would I bother repairing that piece of
junk".
If I think that something I've repaired could break in the same way,
making a better repair or improvement to the design is something I'd
go ahead and do, provided it does not involve a massive reworking. As
I have never found any simple linear regulator better spec'd than the
78xx series, I thought I'd throw the question in while I was here.
I don't do this professionally, nor do I have the audacity to claim
that I am a professional. I'm just someone who knows a modest amount
and takes an interest from the sidelines.
> Why clutter up such a clean and straightforward internal design with
> adequate cooling?
No doubt the unit was designed to a price. Who could say what the
original designers had in mind...or maybe they just made a mistake.
It has to be said that Sherwood has been (mostly) helpful along the
course of this repair, much more so than the competion (Sony and
TEAC).
William
> Alternative 3-terminal linear regs will still dissipate the same
> amount of heat regardless of their current capability.
Okay...if you don't mind my asking, why is that the case? It would
seem to me that a more capable part asked to do the same work as a
less capable part would have an easier time of it.
> If the problem is in the load (i.e. faulty components) then that
> should be where you focus your attention.
There does not appear to be any problem with anything that is powered
by the regulator's output (though I do have some suspicious caps to
replace elsewhere).
William
> Actually, it sounds more like they blew the design by having too
> much input voltage for the output voltage and current.
The regulator is supplied with 25 volts. While I hadn't really looked
at the curves and figures, what they are doing did not sit well with
me from the outset. Why ask the regulator to do that much work?
> But I would think that adding a bit of series
> resistance to the input side and letting some of the waste
> heat go away there would go a long way towards extending the
> life of the regulator package.
I had not thought of this and might just sit down and do the math to
see what I'd need. What I do will depend upon whether or not there is
room to put something in place without going to extremes.
William
> They are heatsinked, but fail anyway.
The regulator I replaced had a fully plastic body, and I wonder if
this did not help to accelerate its demise. The replacement part is a
Motorola/OnSemi LM340 with a metal tab. It seems to me like this would
work better at shedding heat than a fully plastic body.
William
>> Alternative 3-terminal linear regs will still dissipate the same
>> amount of heat regardless of their current capability.
> Okay...if you don't mind my asking, why is that the case?
Because that's how linear regulators work -- they behave like a sort of
intelligent resistor. If you regulate 28V down to 12V with a linear
regulator, the regulator acts like the right value of resistor to get
rid of 16V as heat.
You can do better with more modern regulators, but you need a switching
regulator rather than a linear one. National's "Simple Switcher"
regulators are good for this sort of thing; you pretty much just add an
external capacitor and inductor.
--
Adam Sampson <a...@offog.org> <http://offog.org/>
>Because that's how linear regulators work -- they behave like a sort of
>intelligent resistor. If you regulate 28V down to 12V with a linear
>regulator, the regulator acts like the right value of resistor to get
>rid of 16V as heat.
>
>You can do better with more modern regulators, but you need a switching
>regulator rather than a linear one. National's "Simple Switcher"
>regulators are good for this sort of thing; you pretty much just add an
>external capacitor and inductor.
Murata now has some interesting little switching regulators which
"mimic" a 7805 - they have the same pinout, same current capacity, and
are built on a small PC board which is roughly the size and shape of a
TO-220.
I recently installed one in an old MFJ packet-terminal node controller
(TNC). THe original TNC design uses a 7805, with a small clip-on
heatsink, and a thoroughly-incompetent attempt to couple the heatsink
to the case for additional heat transfer (they gooped the heatsink
surface with silicone compound, but did *not* actually fasten the
heatsink to the case!). Even with the clip-on heatsink, the 7805 was
running quite hot... not too hot to touch, but warmer than I like.
The Murata part dropped right in, requires no heatsink, works nicely,
and it's barely warm to the touch.
As with any decision to substitute a switching regulator for a linear
regulator, you'll need to consider the output noise - it might be too
much for an audio amplifier to tolerate. Or, one or more additional
stages of filtering may be required.
These Murata parts are about $4.50 each in onesies, through Mouser.
--
Dave Platt <dpl...@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
>Hi!
>
>> Alternative 3-terminal linear regs will still dissipate the same
>> amount of heat regardless of their current capability.
>
>Okay...if you don't mind my asking, why is that the case? It would
>seem to me that a more capable part asked to do the same work as a
>less capable part would have an easier time of it.
It WILL have an easier time of it but OTOH it will still require
dissipation of the same amount of heat.
>In article <y2afwiw...@cartman.at.offog.org>,
>Adam Sampson <a...@offog.org> wrote:
>
>>Because that's how linear regulators work -- they behave like a sort of
>>intelligent resistor. If you regulate 28V down to 12V with a linear
>>regulator, the regulator acts like the right value of resistor to get
>>rid of 16V as heat.
>>
>>You can do better with more modern regulators, but you need a switching
>>regulator rather than a linear one. National's "Simple Switcher"
>>regulators are good for this sort of thing; you pretty much just add an
>>external capacitor and inductor.
>
>Murata now has some interesting little switching regulators which
>"mimic" a 7805 - they have the same pinout, same current capacity, and
>are built on a small PC board which is roughly the size and shape of a
>TO-220.
Although I didn't say so explicitly, the two devices I mentioned are
also TO-220 "drop-in" switchers. It is certainly an emerging market
area.
>>Murata now has some interesting little switching regulators which
>>"mimic" a 7805 - they have the same pinout, same current capacity, and
>>are built on a small PC board which is roughly the size and shape of a
>>TO-220.
>
>Although I didn't say so explicitly, the two devices I mentioned are
>also TO-220 "drop-in" switchers. It is certainly an emerging market
>area.
The "Simple Switcher" (tm) parts I have seen up until now, have the
switcher controller, but you need to install external components
(inductor, etc.)., so their pinout isn't directly 7805-compatible.
The Murata parts have all of that on their own PCB - no additional
external components are required, if you can live with their output
levels.
I would expect to see other manufacturers come up with these sorts of
complete drop-in-ready modules.
>> But I would think that adding a bit of series
>> resistance to the input side and letting some of the waste
>> heat go away there would go a long way towards extending the
>> life of the regulator package.
>
>I had not thought of this and might just sit down and do the math to
>see what I'd need. What I do will depend upon whether or not there is
>room to put something in place without going to extremes.
I agree that's the simplest solution. In fact I've done similar things
in lots of gear, including a 586 motherboard.
- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.