Modified Zalman 300A PSU - tests

[Steerpike]

I just replaced the stock fan in the Zalman 300 PSU with a Panaflo 12L, and ran the cable outside the case so I could control it with a zalman fanmate controller. While I was in there, I also hot-melt glued a thermistor to the big heatsink (next to the built-in thermistor) and ran a cable from it outside too. Hooked that to my digital doc 5 temp/fan controller. So with the fan on slow (but above minimum), the thermistor is reading 37 C. A second thermistor is measuring the air temp right at the exhaust point of the PSU. It's reading 34 C. My CPU (as indicated by 'hardware sensor monitor' prog) is 36.5, and the mobo is 35. Dtemp is recording the disk temp at 35. Ambient is 23.4 (ambient under the desk, where the PC is). Case is fully closed.

Summary:
PSU heatsink: ___ 37
PSU exhaust: ___ 34
CPU: __________ 37
Mobo: _________ 35
Dtemp (HD): ____ 35
ambient: _______ 23.4

Dropping the PSU fan to abs. minimum, (inaudible!)
PSU heatsink: ____ 48
PSU exhaust: _____ 40
CPU ____________ 40
mobo: ___________ 38
Dtemp (HD): ______ 38

This may be too hot for the PSU, long term?

I cranked up the power a touch on the PSU fan, to the 'threshhold' where the noise seems to really rise (unfortunately, the panaflo doesn't have an rpm monitor so I can't say what the speed is. Maybe I can now proceed to experiment with heatsinking the PSU, running it outside the case?

[SungHyun7]

how's it going now? psu temperature seems to be a mystery... i don't think anyone really knows what's max imho.

i'm thinking of drilling bunch of holes in my zalman psu heatsink and putting copper water tubing over it and cool the whole system with water...

probably a good idea to get a digital camera beforehand

[Steerpike]

Well I'm not thrilled with the Panaflo; it's a bit noisy. I asked for recommendations, cost being no issue, and was recommended the panaflo, but it seems to 'whine' a bit at low speed.

My PSU is definitely the biggest source of noise now. I think I'll remove it from the case, put a very low-flow (low voltage) (fan just to vent the case) in it's place.

[MikeC]

psu temperature seems to be a mystery

Yes they are because there are SO many variables, not only within the PSU but also in the case/system it is used with. But those numbers posted look very low to me. Perfectly safe, I'd say.

cschofie: who recommended the Panaflo?

[Steerpike]

You did ! In this topic http://forums.silentpcreview.com/forums ... php?t=2329, in your post November 16.

Panaflo 80 "L". There is a cost: higher temp in the PSU. maybe in the system. I do it mine but am willing to accept possibly shortened life (of PSU) and use expensive CPU HS to ensure cooling. I think the fan in the Zalman may be soldered to the circuit board, which means cutting leads and soldering...

.

I'm also willing to believe it's just one noise surfacing over another - remove one noise, and another previously inaudible noise now becomes a source of annoyance. The current noise just happens to be more high-frequency, and annoying to me.

[MikeC]

I did!?! okay...

You know, I have been playing around with the original fan in the Zalman. The MGA8012HB by Protechnic Electric -- I presume that's the one you removed. Well, I have come to feel it is the quietest higher airflow fan I've yet seen. It has a very smooth sound & ramps down very nicely, yet at full 12V, can move a LOT of air. You might try putting it back in with a resistor to bring the default voltage level down to 4.3~4.6V. It might actually be quieter than the Panaflo at 5.5V (the min fan voltage in the Zalman). It will certainly be cooler when you push the PSU/system hard.

I'll do some quick experments and let you know how much resistance you need. I figure since you unsoldered the fan, you can handle soldering a resistor in line?

[Steerpike]

Actually, I simply clipped the wires, and 'sealed' off the existing wires. I then wired the replacement fan (Panaflo) to a zalman fanmate, powered from a standard PSU connector, outside of the PSU itself.

What I found is that if I slow the Panaflo down to it's minimum, I get great 'sound' performance, but ... the PSU gets a bit warm (48C, as noted in this topic). I've been slowly increasing the voltage (via the fanmate) to find the point where the temp drop is sufficient. Currently, I've got the PSU heatsink running at 35C or so, which I think is fine - but the fan is a tad noisy.

I could easily wire up the Zalman fan again. Note, I just purchased a 'fanstorm' device, which let's you monitor and control a fan based on temp (posted in this forum). It will take a few weeks to arrive, though!

Of course, the zalman PSU itself controls it's own fan in a variable, temp-controlled manner - I'm just assuming that I can control it 'better' - weighted towards sound, not longevity.

So you are thinking of inserting a small resistor in series with the stock fan, wired to the PSU internal (original) fan source, so that it's still controlled by the PSU itself, but also slowed down a bit to improve noise performance. Good thought!

I'll try hooking the Zalman stock fan separately.

Question - I guess I can measure the voltage currently 'used' (set) by the zalan fanmate to see what voltage is being supplied to the panaflo at it's current 'somehwat optimum' temperature performance? Will I need to keep the fan connected (does the zalman fanmate supply a constant voltage regardless of load?).

[MikeC]

Question - I guess I can measure the voltage currently 'used' (set) by the zalan fanmate to see what voltage is being supplied to the panaflo at it's current 'somehwat optimum' temperature performance? Will I need to keep the fan connected (does the zalman fanmate supply a constant voltage regardless of load?).

Yes to 1st Q, no to second. Just be careful not to short the + and - leads. Easy to do with the closely spaced terminals.

Also you have bypassed the smart thermal circuitry on the Zalman that speed the fan up when the unit gets hot. Not a terrible thing, but not quite as cool, for sure. If you are ok with soldering, I'd recommend solding the orignal fan + resistor on to those clipped wires. Will probably give you the best of both worlds -- quiet & cool. I have some time now, give me half an hour, I'll see what resistance you need.

[MikeC]

OK, som quick measurements & experiments:

1. The Protechnic Electric fan will start as low as ~3.5V, but this is probably too low.

2. To get ~4V, add a 25-30 ohm resistor to the fan voltage line (in PSU). The PE is quieter at 4V than the Panaflo at 5V.

3. If you want the thermal fan voltage circuitry in the Zalman NOT to speed the fan up as quickly, here what you do: The thermistor wire comes up from the big bunch of wires and to the top of a HS. See the pic below.



It's screwed & glued on the HS. Remove the screw, gently pry off the thermistor. Now it is dangling there. Bend the wire upwards sliightly so that when the cover is put back on, the thermistor might push up against it. Put the screw back where it was before (so you don;t lose it), close cover.

What that will do is keep the fan thermal circuit operative but not speed up the fan as much as before because the thermistor won't get as not, not being in direct contact with the HS. It will still get plenty hot when the PSU is pushed, but in the low/mid power range where the PSU usually runs, the fan should stay running slower more constantly.

Disclaimer: if you kill yourself with some big zap, don't come complaining to me, ok? Tips: unplug from AC, touch only what you have to, and use surgical rubber gloves if you don't trust yourself.[/url]

[Steerpike]

Thanks! What wattage would the resistor need to be? The PE fan is rated at 0.24 amps, so 12Vx0.24A = 2.88 watts - but I guess the wattage of the resistor is based on the voltage drop across it, which would be (12-4) = 8V? 8Vx0.24A=1.9 Watts - but again, I presume with that reduced voltage, the max. current is reduced. Bottom line is, can I get away with an off-the-shelf half-watt resistor?

Note - the PE fan is rated at 0.24A; the Panaflo is rated at 0.1A - less than half the current. So that would explain why the PE fan is moving more air, and is noisier, at full throttle - it's simply a more powerful fan. So maybe, if you are planning to reduce the voltage significantly, it's better to start with a higher power fan - better low voltage performance?

But looking at this another way, if you half the voltage to the PE, it should be somewhat equivalent to the performance of the panaflo, since the PE is double the power?

[MikeC]

based on the voltage drop across it, which would be (12-4) = 8V? 8Vx0.24A=1.9 Watts

The PSU fan controller feeds ~5.5V to the fan, so the drop is only 1.5V. Just to clarity -- you will not be using the fanmate but the original PSU fan voltage feed. The resistor should be on the hot lead of the fan.

PE at -1.5V probably blows more air right from the start (Panaflo at 5.5V vs, PE at 4V) but by the time the fan controller circuit is putting out, say, 7~8V, there will be no contest. The PE is rated for something like 38 CFM compared to 24 for the Panaflo. If your PSU ever gets hot enough that the voltage output goes to max (12V, presumably) at 10.5V, the PE will blow a LOT more air than the Panaflo at 12V. The way you have it now, the Panaflo never speeds up unless you turn the fanmate knob anyway. (To me, it seems like a bit of a waste to spend the $$ on a specially noise reduced PSU and then bypass a key feature of the design helps make it quieter -- and more expensive -- in the first place.)

Anyway, I think this is probably worth a try. If your system draws the kind of power that mine do (not much, under 100W, always), I think you will find this setup quieter than the Panaflo at 5V -- almost all the time. NOt a big difference, mind you, but something. If ends up being louder, you can always go back.

[Steerpike]

I've now wired my PSU for 'flexible fan control' I've got the original PSU fan-power wires running out of the case, the original Zalman (PE) fan installed, with leads running out of the case; and a second thermistor glued to the heatsink next to the built in one. So I can power the fan from either the PSU's own temp-controlled supply, or my Zalman fan-mate supply, and monitor the temps. What else would one do on a Thansgiving weekend?

I was surprised by what you said at first ...

The PSU fan controller feeds ~5.5V to the fan, so the drop is only 1.5V.

I measured 9V across the built-in PSU power lines (un-loaded - ie no fan connected). But you are right - as soon as you hook up the fan, the voltage drops to about 5.5V. Strange; the zalman fanmate supplies a given voltage when it's unloaded, and maintains pretty much that same voltage when a fan is connected. Makes comparisons harder ...

Anyway I did a bunch of tests with both the built-in power supply lines and the zalman fanmate. With the PSU open, I inserted a card between the fan and the heatsink (to reduce the cooling effect of the fan), and measured the voltage on the actual fan at different heatsink temps. For the record, I'll record them here!

Heatsink temp and voltage using built in fan power supply:
Temp Volts
20 ---- 5.9
35 ---- 5.9
36 ---- 6.0
40 ---- 6.1
42 ---- 6.2
43 ---- 6.3
44 ---- 6.4
45 ---- 6.5
46 ---- 6.6
46.5 -- 6.7
46.7 -- 6.8
47 ---- 6.9
47.6 -- 7.0

After inserting a 22 ohm resistor in series with the fan supply (half-watt used - hope that's ok!)
Temp Volts
39 ---- 4.63
40 ---- 4.69
41 ---- 4.7
42 ---- 4.77
43 ---- 4.81
44 ---- 4.9
44.9 -- 4.98
45.4 -- 5.04
46 ---- 5.12
46.5 -- 5.2
47 ---- 5.38
47.5 -- 5.5
48 ---- 5.6

I also did the above tests using a 3 watt, 25 ohm variable resistor, and measured the voltage drop across it - and it was pretty consistently 1.35 to 1.45 V (as Mike suggested).

I also tested with the fanmate, at min and max settings. At max setting, unloaded voltage is 11.03 V, loaded (fan connected) is 10.42. At min setting, voltage is 5.02 (loaded or unloaded - makes no difference).

So - bottom line is, with a fanmate at the minimum setting, the fan is getting 5.02 V, while with a 22 ohm resistor (and a cold PSU heatsink!), the fan is getting 4.63V - half a volt lower, and slightly quieter. And as Mike said, I have the knowledge that if the heat rises, the voltage supplied will rise also.

I pointed a can of compressed air at the heatsink to cool it down even more, briefly, and with the heatsink temp at 32C, got a voltage reading of 4.58.

It seems that the built-in circuit really starts to increase the voltage to the fan at about 40C - increasing about 0.1V per degree, and ramps this up more at about 46C.

Another observation - with the PSU case off, and no fan, the heatsink temp rises quickly; I gave up measuring when it hit 50C. Clearly this PSU can't be run fanless without some serious modifications. (well, I was too chicken to see if it stopped rising once it hit 50C!).

Also noticed that the PSU case greatly helps in cooling (with the PSU out in the open, not enclosed in the PC case). With the PSU case off, for a given fan speed (when I was using the zalman fanmate, which is not temp controlled), the PSU heatsink temp was 35C, and with case on, 27C. So the PSU case really helps 'funnel' cold air over the heatsink.

Whew ... what a waste of a day that was ... Now I need to get back to my real purpose in life and scan some pictures!

[MikeC]

3 watt, 25 ohm variable resistor

That's actually a good idea! If you put in one ~35 ohms (will give you a bit more that 1.5V drop), you could vary the fan voltage/temperature curve up/down in accordance to your ambient room temps. In other words, use a higher resistance setting when it is cool and turn the resistance to zero if you are going to play 3D shooter games for 6 hours. Or hook a single resistor to a switch to take it in/out of the circuit, kind of like the 12/5V switch.

Oh, BTW, after all that experimentation (now you know how much work all this stuff is!!), is my prediction/hunch confirmed? Is the PSU quieter with the Panaflo or with the resistor-added original fan? It may be too close to tell without an immediate A/B comparison...

PS - 1/2 watt is fine, it sees so little current anyway. Power equals Voltage squared over Resistance, if I recall. I think that applies here.

[Steerpike]

Your hunch was exactly correct - it's quieter. The Zalman (PE) fan is hardly audible. The panaflo had a distinct 'whine' - higher pitched, more annoying.

In fact, I bought 4 panaflo's and tested them all; quite a range of sounds. I think their 'qc' may not be that great. I think there's a little eccentricity on the fan blade assy, giving rise to bearing noise - or something. I think we need to find a better 'standard' fan! Also, I'm wondering if anyone has ever done a decent review of fans taking into account their performance at (eg) 5V, and taking into account the amount of air moved. I've seen theoretical analysis, but no actual comparisons. When I first got started in my 'silence' quest, I bought one of each type of fan I could find - but I haven't had the time to compare them!

And yes, wow, this stuff takes time! To make it worthwhile, you have to measure measure measure. And, to be repeatable, you can't really use the PC (use causes CPU to heat up/draw current from PSU) so you can't do much else either! And, I learned (again!) that crimping is not a good idea with thin wires. To try to speed up my experiments, I crimped all kinds of male/female connectors to various wires ... but discovered that the wires were ever-so-slightly loose in the connectors (barely noticeable). I ended up putting a solder 'coating' on the bare wires to bulk them up so the crimp would have more to squeeze down on, and also have a more 'pliable' material (solder vs. wire) to squish. But in the end, soldering the wires is the way to go!

[djhardkhor]

guys this is a great thread. Just bought a zalman 300a not so long ago and was quite dissatisfied by the noise levels over my mod'd generic with a L1A@7v. (which died, possibly due to either the mod or 'old age')

Will consider wiring in the var resistor like you suggested, and putting a control on the outside of the psu.

cheers

djhk