Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Josh Billings]

I'm looking to get a powersoft amp for subs and a PLD4.5 for mids / highs but I noticed the airflow on them is different. Is this an issue? I know ideally they'd be all the same, but just wondering how much this matters.

[John Roberts {JR}]

If the front and back of the racks are open and free flowing it doesn't much matter. If front or back is closed in, one direction could exhaust hot air into the other.

Still might work but not ideal.

JR

[Chris Grimshaw]

I'd avoid it if possible. Whichever way you rack them, the top amp is still sucking up the rising hot air from the bottom amp.

FWIW, when I came across this problem, I reversed the fans on some of the amps. To be 100% perfectly clear, THIS IS NOT A GOOD IDEA, unless you've done the research to make sure it'll be okay. Do not go turning fans around because you feel like it.

In my case, I was fortunate - someone had done the same modification before me, and measured temperatures before and after and found the output devices to stay within 1C before/after, while a small regulator in the power supply actually ran cooler after.

I got lucky with mine. Might be worth talking to QSC about yours.

An alternative would be to add a small USB fan (4-5") at the front and back of the rack to disperse any hot air before the top amp can suck it up.


Of course, we're talking about engineering this to the Nth degree. Chances are, unless you run everything into limit all night (and the amps are acting like hairdryers), you'll be fine.

Chris

[Chris Hindle]

I'm looking to get a powersoft amp for subs and a PLD4.5 for mids / highs but I noticed the airflow on them is different. Is this an issue? I know ideally they'd be all the same, but just wondering how much this matters.

Last century I was called to a club to figure out why their newly installed system was thermalling out.
Rack of 12 or so amps. QSC and Crown.
Looked pretty as heck.
They stacked them QSC, Crown, QSC, Crown, QSC... etc.
so, we had an innie, an ouitie, an innie, an outie.
Some installers.......

The rack was open in the front, solid walls, and full grill on the back.
I re-racked everything. Innies together, outies together and slid a sheet of masonite in to split the rack into 2 sections.
Worked like a charm. All good.
You really don't want suckers and blowers just recycling each other's air.
Chris.

[John Roberts {JR}]

I'd avoid it if possible. Whichever way you rack them, the top amp is still sucking up the rising hot air from the bottom amp.

FWIW, when I came across this problem, I reversed the fans on some of the amps. To be 100% perfectly clear, THIS IS NOT A GOOD IDEA, unless you've done the research to make sure it'll be okay. Do not go turning fans around because you feel like it.

Not just a bad idea, but DO NOT DO THAT...

In my case, I was fortunate - someone had done the same modification before me, and measured temperatures before and after and found the output devices to stay within 1C before/after, while a small regulator in the power supply actually ran cooler after.

A little curious how you measured that?

One of my patents at Peavey was for an amp heatsink design (6,515,859 Roberts , et al. February 4, 2003)

In forced air cooling, direction makes a bigger difference that you might think. Observe the difference between sticking your hand in front of a fan, and behind that same fan. In front you will feel numerous air collisions from the directed forced air. Behind that same fan, the same amount of air is moving, but the direction of air flow is diffuse and less effective for cooling a target.

I got lucky with mine. Might be worth talking to QSC about yours.

An alternative would be to add a small USB fan (4-5") at the front and back of the rack to disperse any hot air before the top amp can suck it up.


Of course, we're talking about engineering this to the Nth degree. Chances are, unless you run everything into limit all night (and the amps are acting like hairdryers), you'll be fine.

Chris

While not pretty some amp racks have threaded rails in the back so innie and outie amps could be realigned to agree.

I have even seen some forced air amp cooling paths that exhausted out the sides of the amp chassis (not a very good strategy for playing well with others). 

This is not rocket science but it is science.

JR

[Chris Grimshaw]

A little curious how you measured that?

I didn't. The guy that did used an IR camera. The amps were some of the bigger iNukes where the cooling strategy is simply bulk air flow through the case. Such techniques aren't particularly efficient, but if the cooling needs aren't extreme then it does seem to work.


There's far less of a difference in front/behind a fan when the fan is in a tunnel. The difference then comes down to low vs high pressure cooling.

The desktop PC guys get really into this stuff, so a lot of info can be gleaned from their forums.

Chris

[Frank Koenig]

In my case, I was fortunate - someone had done the same modification before me, and measured temperatures before and after and found the output devices to stay within 1C before/after, while a small regulator in the power supply actually ran cooler after.

I'll reinforce what JR said. Don't do it. Engineers spend weeks working on the thermal design and weeks more verifying that it works. It's not just the temperature of the semiconductor devices, although that's clearly important. Electrolytic capacitors are perhaps the shortest-lived and most temperature sensitive components. If you take the lid off a Powersoft amp you'll see a sea of big electrolytics that are all on the air-intake side of the chassis. The capacitors don't give off significant heat themselves so there is no cost in passing the air over them first. Were you to reverse the airflow, the (considerable) heat coming off the power semiconductors would cook the capacitors in short order.

It's not just the power supply and output devices that need cooling. Modern DSP amplifiers contain computing chips that give off a lot of heat and need proper cooling as well. If the amplifier uses air filters, those clearly need to be on the intake side to do any good. In short, don't mess with the thermal design.

As for the problem at hand, maybe use a simple baffle to ensure that every amp gets to pull cool air, or flip the offending amp end-for-end in the rack?

--Frank

[John Roberts {JR}]

I didn't. The guy that did used an IR camera.

Inside a closed steel chassis?

Back in the day design engineers would attach thermocouples to power devices to log temperature. Modern technology has improved...

Back last century I wanted to use a digital IR camera on power amp burn in stations to detect abnormal operation from spot temperature readings comparing readings from a good assembly. The technology was not quite ready for prime time back then.

I have seen some prototype amps built with clear plexiglas top covers so the packaging engineer could study air flow inside, using cigarette smoke as his smoke source. I guess if IR thermometers existed back then he could have shot temperature readings through the plexi. 

The amps were some of the bigger iNukes where the cooling strategy is simply bulk air flow through the case. Such techniques aren't particularly efficient, but if the cooling needs aren't extreme then it does seem to work.

I don't recognize "bulk air flow" as an actual cooling strategy. Sounds like the design was add a fan and see what happens.

There's far less of a difference in front/behind a fan when the fan is in a tunnel. The difference then comes down to low vs high pressure cooling.

Huh? are you saying Behringer used a heat sink tunnel... I really doubt they did. These are expensive and work better with lots of power devices. I haven't seen one of those used inside an audio amp for several decades. Modern class D amplifiers use few power devices (and no heat sink tunnels).

Again I do not recognize low or high pressure cooling as general full system descriptive terms. Within any forced air system there will be regions of higher pressure (more collisions with cooling air) and lower pressure (for less cooling). Reversing fan direction will alter where those high and low pressure regions occur.

Coincidentally my heat sink patent was for older analog amplifier technology so involved multiple power devices attached to two long heatsinks. The invention was to angle the two heat sink fins facing each other, closer together at the outlet end. This toe in resulted in more collisions with cooling air and more heat transfer at the hot end. This was desirable because by the time air reached the outlet end it was already warmed by cooling the earlier devices. This increased heat transfer to the warmer cooling air helped keep the hot end vs cool end device temperature differences smaller (a good thing because the amp is only as reliable as its one hottest device.)

Reversing the fan direction for my heat sink approach would completely defeat the design. 

The desktop PC guys get really into this stuff, so a lot of info can be gleaned from their forums.

Chris

Actually there might be some parallels between computer chip cooling and class D audio amps as both have heat concentrated in a small sized footprint. 

I did some work with Peltier devices and used a computer chip cooler fan/heat sink assembly, to move the heat out of a modest sized P device.

JR

[Nathan Riddle]

Have any of yall looked at an iNuke's internals?
You can definitely reverse the fan's direction.
Why?
Because they don't have heatsinks or order to their internal components

Other amps. Yeah, what ^ they said (unless the manufacturer says you can).

[John Roberts {JR}]

I'll reinforce what JR said. Don't do it. Engineers spend weeks working on the thermal design and weeks more verifying that it works. It's not just the temperature of the semiconductor devices, although that's clearly important. Electrolytic capacitors are perhaps the shortest-lived and most temperature sensitive components.

Everything is affected by temperature. They make high temperature rated capacitors but they cost more.  In solid state reliability engineering a popular rule of thumb is that every 10'C temp increase MTBF (mean time between failure) drops in half. Electrolytic caps can have extra sensitivity to electrolyte loss.

If you take the lid off a Powersoft amp you'll see a sea of big electrolytics that are all on the air-intake side of the chassis. The capacitors don't give off significant heat themselves so there is no cost in passing the air over them first. Were you to reverse the airflow, the (considerable) heat coming off the power semiconductors would cook the capacitors in short order.

cool end/ hot end...  reversing fan direction reverses that...

It's not just the power supply and output devices that need cooling. Modern DSP amplifiers contain computing chips that give off a lot of heat and need proper cooling as well. If the amplifier uses air filters, those clearly need to be on the intake side to do any good. In short, don't mess with the thermal design.

As for the problem at hand, maybe use a simple baffle to ensure that every amp gets to pull cool air, or flip the offending amp end-for-end in the rack?

--Frank

Or read the amplifier owners manual for advice.

This topic has been discussed here multiple times over the decades.

JR

[ProSoundWeb Community]