I will be replacing the mains fan as 10 years of much use but any other
observations from anyone familiar with them ?
Dusty inside but not blocking any air throughput and no reported changes in
fan noise and currently seems the correct sort of air flow for 14W fan.
Design fault concerning airodynamics? There are 2 full width flutes of the
heatsink that go up to the fan face. I would expect these bits of aluminium
to be milled out normally in such situations.
--
Diverse Devices, Southampton, England
electronic hints and repair briefs , schematics/manuals list on
http://home.graffiti.net/diverse:graffiti.net/
N_Cook wrote:
> In for intermittant loss of one channel but this query is about overheating.
> There are 5/8 inch high feet on the base but unless the owners prop up on
> blocks the whole base gets noticeably hot.
>
> I will be replacing the mains fan as 10 years of much use but any other
> observations from anyone familiar with them ?
> Dusty inside but not blocking any air throughput and no reported changes in
> fan noise and currently seems the correct sort of air flow for 14W fan.
> Design fault concerning airodynamics? There are 2 full width flutes of the
> heatsink that go up to the fan face. I would expect these bits of aluminium
> to be milled out normally in such situations.
Is that the model that has a fan mounted horizontally blowing down onto a skimpy
heatsink ?
Graham
Blowing up into an adequate , for fan blown, heatsink with a bank of 12
motorola 70483080 TO3 . As far as I can see the 2 uncut-down heatsink vanes
that go across the fan do little functionally for cooling and probably
disrupt the air flow.
The air has to come in horizontally under the base , up the one inch of fan
axis and then have to abruptly divert horizontally again down the 2
(separated by these vanes) relatively small channels in the heatsink. I'll
take a pic perhaps
Not to the extent of tripping any of the thermal switches but enough to
concern the owner into trying to improve airflow - admirable thinking for a
non-techie.
Just giving greater clearance to the airflow under an amp should not
normally affect the amp temperature though. I would suggest that it is due
to forcing the air to follow a tortuous path so a generic design flaw. I've
heated up the fan motor, etc and there seems nothing wrong with this fan as
well as no change of fan noise reported by the owner.
What say if an amplifier was designed in such a way that the thermal
protection was intended to operate at some given device temperature, say
the safe working temp of the output transistors, and the cooling
arrangments designed accordingly? Output transistors get hot, heatsink
gets hot, thermal devices trip, amplifier shuts down or whatever it`s
designed to do.
But, by increasing the airflow from the fan[1] or improving the cooling
to the heatsink in some other way, results in the heatsink remaining
cooler when the actual output devices themselves are still running at a
high temperature. Bearing in mind that the temp sensors may not be
directly monitoring the devices temp, but some other area on the
heatsink, that might lead to the thermal protection failing to operate.
By changing the design of the cooling, you could be blowing cool air
over the temp sensors and not the output devices. Result = a trip to
silicon heaven.
I have seen Peavey mixer amps shut down in hot weather when placed in
direct sunlight and driven hard, that`s the protection doing what it`s
designed to do tho. Generally speaking, if a decent quality amplifier is
overheating, it`s operator error, not a design fault. (IMO)
Ron(UK)
[1] This is hypothetical of course, I realise you havent changed the fan
or it`s cooling path in this instance, but I think it`s something to be
aware of. Maybe someone with experience in 'designing' power amplifiers
could comment?
s.e.d added
All very true but you seem to be missing the point.
An amp that works perfectly all right should not suddenly run noticeably
cooler doing exactly the same job in the same gig but just placing fag
packets, lighters etc , whatever was at hand, under each of the feet on the
base.
2 th sw on 2 of the o/p TO3 and one on the heatsink in this case and yes the
air blows over the those 2 th sw as much as the cans of the TO3, in fact
they are directly in the ducted flow of air probably reducing the passing
volume/area by a 1/3 at each end of one of the heatsink ducts, compared to
no th sw in the path.
N_Cook wrote:
> Eeyore <rabbitsfriend...@hotmail.com> wrote in message
> >
> > Is that the model that has a fan mounted horizontally blowing down onto a
> > skimpy heatsink ?
>
>
> Blowing up into an adequate , for fan blown, heatsink with a bank of 12
> motorola 70483080 TO3
How do you define adequate ? I design my own heatsinks for amps you see and most
others I come across are very inadequate, notably recent Peaveys.
Graham
N_Cook wrote:
> Perhaps the sensible approach to this is extend the rubber feet by half an
> inch as this is a known cure.
A simple solution at least.
> There is no obvious place to put a secoind fan at the air outlet.
You CAN 'gang' 2 fans in series to increase the air pressure but I've never
needed to do it.
Also, fitting a 'thicker' fan will increase the air pressure. You may find 12,
19 and 25mm versions of certain axial fans. The thicker they get, the more
pressure they generate and that helps achieve good airflow. The CFM figures are
largely just a guidline since there are with zero pressure gradient.
Graham
Meat Plow wrote:
> N_Cook wrote:
>
> > Perhaps the sensible approach to this is extend the rubber feet by half an
> > inch as this is a known cure. There is no obvious place to put a secoind fan
> > at the air outlet.
>
> So every XR 886 overheats?
Would depend how they're operated, load impedance and the style of music
(especially so). I have researched this area extensively and the variations are
enormous.Stnadardised tests such as 1/8 or 1/3 power noise are especially useless.
Graham
N_Cook wrote:
> Just giving greater clearance to the airflow under an amp should not
> normally affect the amp temperature though.
Yes it will by reducing the back pressure.
Graham
N_Cook wrote:
> I would suggest that it is due
> to forcing the air to follow a tortuous path so a generic design flaw.
That's what causes the high back pressure (and hence low air flow). For good
cooling the airflow path should be in a striaght line with air vents in and out
as large as possible (consistent with safety regs) and there are some other cute
aspects which are even less well known too.
Graham
For fan blown o/p devices the heatsink surely does not have to be much more
that a metal matrix for holding the devices to. Does not have to be
multifluted or with those chisseled multi-vanes etc (sorry don't know the
generic term , end up curved ) or a great mass of metal for convection
cooling
Agreed , I'll take a photo later to show how tortuous this path is.
Even more dramatic was an Ashdown amp I had to repair recently for another
fault.
Supposedly fan assisted cooling but the fan "design" location was just
circulating air inside the casing as no baffles etc to guide air in and out
the single "vent" port.
Owner much appreciated the amp running much cooler and being able to feel
air coming out of the vent.
s.e.r added as design issue
"Ron(UK)" wrote:
> Maybe someone with experience in 'designing' power amplifiers
> could comment?
I have done !
There's something worth adding though.
To optmise air flow you need a high 'static pressure'. There's a huge variation
in this characteristic. The best examples I've found are NMB (formerly Boxer) and
Panasonic's Panaflow fans. Be careful to order the high pressure version. They
make versions with various motors speeds. In the case of NMB, the 'best' ones
used to have 'B50' as part of the part number. I forget Panasonic's method of
marking.
DO NOT use cheap 'off brand' fans for applications like this.
I see Farnell have NMB but only as a special US import. You didn't say what size
it is but the popular 80mm size would be for example ....
http://uk.farnell.com/8235414/electrical-circuit-protection/product.us0?sku=nmb-3112kl-05w-b50-e00
Read the data sheet here
http://www.eminebea.com/content/pdf/TMP00123/I/3112kl.pdf
to see just how radically. the airflow varies with the Bxx model number.
Graham
N_Cook wrote:
> An amp that works perfectly all right should not suddenly run noticeably
> cooler doing exactly the same job in the same gig but just placing fag
> packets, lighters etc , whatever was at hand, under each of the feet on the
> base.
You don't understand equipment cooling adequately. It's PERFECTLY obvious to me.
It's all to do with the effect of back pressure on air flow.
Adding more space underneath the unit reduces the back pressure. It's a truly
cruddy design that they did it that way though.
Graham
N_Cook wrote:
> For fan blown o/p devices the heatsink surely does not have to be much more
> that a metal matrix for holding the devices to.
The larger area the better.
> Does not have to be multifluted
That certainly helps.
> or with those chisseled multi-vanes etc (sorry don't know the
> generic term , end up curved )
Those increase surface area which is a GOOD THING.
> or a great mass of metal for convection cooling
It most certainly DOES NOT need to have great weight. But almost none of the
colling in SR amp is classic convection anyway. What typically helps is maximum
surface area of the fins/vanes to create the best opportunity to let the heat
transfer to the air flow, but you can improve even that by 'slotting' the
'vanes' to create a slightly turbulent air flow.
Trust me I could write a lecture on the subject.
Graham
I recall many years ago that there was a 'manufacturers mod' for some
power amp or other ( dont recall which) where you had to bend all the
vanes of the heatsinks - which were like combs - away from the vertical
in various different directions to improve the marginal cooling just
that tiny bit to prevent the thermal protection tripping. Looked damn
ugly but did the trick.
What works best Graham, a thick ally heatsink with many thick fins, or a
thick ally plate with many convoluted thin sheet fins?
Ron
Anyone know the generic name for this type of what I assume is
chisel-action-cut vaned heatsink.
Is it made using a shaper type machine with some sort of chisel or plane
blade?
http://home.graffiti.net/diverse:graffiti.net/chisel_vane_heatsink.jpg
I think that the worst part of that desgn IIRC is that the board is
mounted inverted with the sinks beneath, thus any heat which tried to
escape by convection has to pass through the pcb, heating the other
components. Am I correct in that assumption?
Ron
"Ron(UK)" wrote:
> Eeyore wrote:
> > N_Cook wrote:
> >
> >> For fan blown o/p devices the heatsink surely does not have to be much more
> >> that a metal matrix for holding the devices to.
> >
> > The larger area the better.
> >
> >> Does not have to be multifluted
> >
> > That certainly helps.
> >
> >
> >> or with those chisseled multi-vanes etc (sorry don't know the
> >> generic term , end up curved )
> >
> > Those increase surface area which is a GOOD THING.
> >
> >
> >> or a great mass of metal for convection cooling
> >
> > It most certainly DOES NOT need to have great weight. But almost none of the
> > cooling in SR amps is classic convection anyway. What typically helps is maximum
> > surface area of the fins/vanes to create the best opportunity to let the heat
> > transfer to the air flow, but you can improve even that by 'slotting' the
> > 'vanes' to create a slightly turbulent air flow.
You can btw see this in many 'CPU cooler' designs. However it can cause dust and
fluff to collect.
> > Trust me I could write a lecture on the subject.
>
> I recall many years ago that there was a 'manufacturers mod' for some
> power amp or other ( dont recall which) where you had to bend all the
> vanes of the heatsinks - which were like combs - away from the vertical
> in various different directions to improve the marginal cooling just
> that tiny bit to prevent the thermal protection tripping. Looked damn
> ugly but did the trick.
Yes I can believe that. It would help cause the 'turbulent airflow' effect. I first
learnt about this around 30 years ago from data supplied by the IERC heatsink company
and have widely used the technique ever since with measurable success.
You see with purely vertical or horizontal fins widely spaced, the air going 'up the
middle' between them never actually extracts much useful heat. You need to get the
airflow to 'mix' to get the best effect.
> What works best Graham, a thick ally heatsink with many thick fins, or a
> thick ally plate with many convoluted thin sheet fins?
Within reason, the fins don't need to be that thick at all. Their minimum width is
normally a limitation of the extrusion technique. What's best it have lots of them,
consistent with not creating a dust trap by having them pitched too fine..
Don't make the fins TOO convoluted as this will actually impede airflow. I have my
own 'proprietary' technique that avoids any need to twist the fins which most people
would find counter-intuitive but in a calibrated test produced a 20% improvement in
thermal resistance with a lower weight of aluminium in 2 (superficially to the casual
viewer) identical designs.
As it seems no-one has yet twigged my method, I'm reluctant to give away my 'secret'.
As for the 'plate' its needs to be thickest near the devices and can taper further
out. Thermal resistance calculation are little different to electrical resistance
after all.
Graham
N_Cook wrote:
> Anyone know the generic name for this type of what I assume is
> chisel-action-cut vaned heatsink.
> Is it made using a shaper type machine with some sort of chisel or plane
> blade?
>
> http://home.graffiti.net/diverse:graffiti.net/chisel_vane_heatsink.jpg
I'm afraid I found the pic too unclear to be sure what I was looking at.
Is it formed from sheet aluminium ?
Graham
N_Cook wrote:
> This is the amp in question
> http://home.graffiti.net/diverse:graffiti.net/peavey.jpg
That is about the worst excuse for a 'heatsink' I have seen in my entire life.
It's utterly pathetic. In fact it's completely WRONG in just about every way
possible.
I have seen another that came close to being that bad but even so at least there
were a few more vanes on it (C-Audio btw). The company I was with was
considering taking the C-Audio design as a OEM product and re-branding it, but
my tests showed a heatsink temp of up to 150C with a 40C ambient on music test
signal. This was enough to cause several of the output mosfets to fail
open-circuit and I only realised that when later bench testing it and finding
asymmetrical clipping.
After declining their offer of the product I later found they'd added more
heatsinking but it was a kludge.
Just for good measure, C-Audio also exhausted the searingly hot air right over
the PSU electrolytics ! These were well known to need regular replacement which
C-Audio justified as they said they were a 'service item'.
My own designs (including one in the local venue) are known for working day in,
day out for years on end. I think the one in there is getting on for 10 years
old now and is used every night although not every night is a 'loud' band. It
has had a 'blow through' or two of the heatsinks and 2 new level pots (only one
went crackly but it's best to replace both whilst you're at it, and yes, they're
NOT pcb mounted so it's an easy job too. Compare that with a QSC RMX 2450 I also
had to do the same to (labour time approx 4:1 ratio).
I never fail to be amazed at the number of amps that illustrate that the
designers are completely incompetent when it comes to thermal management.
And don't even get me started on power transformer design !
Graham
N_Cook wrote:
> The view of this amp
> is upside down as the fan is at the base. So drawing air horizonatally
> theough a "designed" 5/8 inch space under the amp.
Hahahahahahahaa !
What LOONIES. That's STALLING the fan blades. NO wonder putting bigger feet on
helps.
How well would a propellor aircraft fly if you put an obstruction in front of
the prop ?
Graham
The heatsink is fitted onto the pcb that caries nothing on that area but
holds a lot of ceramic W/W seen either side on the pic, usefully, I suppose,
in the airflow.
I was rushing around a didn't focus on the right distance.
Will have another go closer up , showing the chisel? cuts. Could not find an
image of one out there.
Often seen in Sony domestic amps of the 70s/80s TA..., STR...all convection
cooled
The cutting action must be like those butter or cheese knives/scoops that
curl the material into whirls
Expressing it like that highlights the flaw admirably.
Voltages in the UK don`t generally sag much if at all
Ron
Our mains supply is 240 volts
AFAIK the lowest legal voltage the electricity company can supply in the UK
is 10% down (from 240) which is 216v, the highest legal voltage is 254v,
less that 6% high. Suffice to say you won't normally get those kind of
variations at the mains intake at the venue.
I don't know how much voltage drop you might experience due to the clubs
wiring before it was deemed unsafe or illegal though.
Gareth.
"Ron(UK)" wrote:
You beat me to it.
Graham
BobG wrote:
> On May 29, 2:35 pm, "Ron(UK)" <r...@lunevalleyaudio.com> wrote:
> > Voltages in the UK don`t generally sag much if at all
> =============================================
> Is that in percent or in volts?
Whatever you like
Graham
gareth magennis wrote:
> "Ron(UK)" wrote
> > BobG wrote:
>
> >> That ancient Peavey transformer might be hotter (thermally! not
> >> wattagely!) on 50hz than on 60hz, and class AB amps dissipate worst
> >> case power at about 1/3rd max power.... its where the load line runs
> >> right across the hyperbola on the IV curve... and if the voltage in
> >> the club sags when the drink cooler compressor turns on, its a perfect
> >> storm of overheat factors.
> >
> > Voltages in the UK don`t generally sag much if at all
>
>
> AFAIK the lowest legal voltage the electricity company can supply in the UK
> is 10% down (from 240) which is 216v, the highest legal voltage is 254v,
> less that 6% high.
It won't be based on 240V any more because we're officially (on paper) 230V.
Maybe 230 +/- 10% but that's just an EU standard - it's NEVER as low as that.
> Suffice to say you won't normally get those kind of
> variations at the mains intake at the venue.
>
> I don't know how much voltage drop you might experience due to the clubs
> wiring before it was deemed unsafe or illegal though.
Never seen any worth talking about. Have you ?
Graham
On paper maybe, but in practice its still pretty much 240v.
>
>> Suffice to say you won't normally get those kind of
>> variations at the mains intake at the venue.
>>
>> I don't know how much voltage drop you might experience due to the clubs
>> wiring before it was deemed unsafe or illegal though.
>
> Never seen any worth talking about. Have you ?
>
> Graham
>
No, I put that down to good wiring regulations.
(I remember when I was in a function band and we used to play the Forces
bases in Germany with dodgy old 220v domestic-at-best wiring. We only had
about a 3K system but often had to carefully switch the amps on one at a
time to stop the breakers tripping, then all the house lights and all the
gears' illumination would dim in time to the kick drum. Frightening really.
None of my keyboards ever crashed though.)
So what would be the maximum voltage drop you might expect from a non
overloaded mains system?
Gareth.
gareth magennis wrote:
> So what would be the maximum voltage drop you might expect from a non
> overloaded mains system?
Crikey. I know where I'd test with a highish power PA rig, the local venue that
has some 5kW ? or so of amps but last year I had a separate 32A circuit run in
just for the amp rack so I doubt it would be more than a few volts.
Graham
I was just curious, knowing nothing really about wiring regulations.
If the regulations require heavier cable for longer runs to prevent voltage
drop, then is there some kind of formula with voltage drop limits? Or can
you just run a very long cable (in a large building) capable of carrying the
current required and have to live with the voltage drop. Or is any voltage
drop just a few volts as you suspect, so the point is mute.
Kind of thing.
Gareth.
gareth magennis wrote:
> "Eeyore" wrote
> > gareth magennis wrote:
> >
> >> So what would be the maximum voltage drop you might expect from a non
> >> overloaded mains system?
> >
> > Crikey. I know where I'd test with a highish power PA rig, the local venue
> > that has some 5kW ? or so of amps but last year I had a separate 32A circuit
>
> > run in just for the amp rack so I doubt it would be more than a few volts.
>
>
> I was just curious, knowing nothing really about wiring regulations.
>
> If the regulations require heavier cable for longer runs to prevent voltage
> drop, then is there some kind of formula with voltage drop limits? Or can
> you just run a very long cable (in a large building) capable of carrying the
> current required and have to live with the voltage drop. Or is any voltage
> drop just a few volts as you suspect, so the point is mute.
Well, a 'spur' rated at 32A requires 6mm2 cable. You could calculate a voltage
drop along that based on resistance data. There is AFAIK no limit on length
aside from practical limitations ! I don't know how the incoming feed from the
electric company to the breaker box might drop with load though.
Graham
In the US, it's no less than 5% drop from the source to the load at full
load.. conductor must be rated to supply a min of 95%.
for example, 10 awg = 0.999/M or 3.28/kg ohms
of course, all conductors are based from copper, copper being 100% of
the scale. All others are based from that.
etc..
I haven't looked it up properly but in the design of fixed installations
you are not supposed to have more than a 4% voltage drop from the point
of supply. So whilst 6mm cable will easily carry >32A, once you go over
a certain length you would need to go up to 10mm to comply with the regs.
http://www.tlc-direct.co.uk/Technical/Charts/VoltageDrop.html
--
Tim Phipps
replace "invalid" with "uk" to reply by email
Thanks. Then from my calculations the maximum legal drop for any
installation should be 13.6% from 240v - 207v and a bit, not including
your own power distro.
Gareth.
Um, actually I think that's a bit innacurate as the 4% will be calculated
for the power loss at 240v or even 254v? Close enough though.
Gareth.
gareth magennis wrote:
> "Tim Phipps" wrote
> > gareth magennis wrote:
>
> >>
> >> I was just curious, knowing nothing really about wiring regulations.
> >>
> >> If the regulations require heavier cable for longer runs to prevent
> >> voltage drop, then is there some kind of formula with voltage drop
> >> limits? Or can you just run a very long cable (in a large building)
> >> capable of carrying the current required and have to live with the
> >> voltage drop. Or is any voltage drop just a few volts as you suspect, so
> >> the point is mute.
> >>
>
> > I haven't looked it up properly but in the design of fixed installations
> > you are not supposed to have more than a 4% voltage drop from the point of
> > supply. So whilst 6mm cable will easily carry >32A, once you go over a
> > certain length you would need to go up to 10mm to comply with the regs.
> >
> > http://www.tlc-direct.co.uk/Technical/Charts/VoltageDrop.html
>
>
> Thanks. Then from my calculations the maximum legal drop for any
> installation should be 13.6% from 240v - 207v and a bit, not including
> your own power distro.
Hah ! 230V -10%. That's what I test kit at when doing low voltage dropout tests.
Graham