• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Need help with TAM cooling fan!

pb3623

Well-known member
Hello all,

I got a second TAM last week off of one of the Facebook groups and, like the one I bought last fall, it also came well-appointed w/ a Sonnet G3/400. After some machine freezes, I took off the Fat Back and saw the frameless in the HD bracket was not on. I've since been working on certain tests, to the extent of my "trained monkey" abilities.

  • Fan header J17 next to the ROM/cache slot measures 12.00 volts across both pins.
  • The installed fan, using bridge wires, runs off of a 9V battery.
  • I don't have the specs on this stock fan since the label isn't visible but I assume it's 12V, 600-900mA based on its size. I know the amperage needs to be known if a replacement was deemed necessary but the 9V battery should rule that out.
  • I've tried a variety of other 12V fans of similar size, and they all work off a 9V battery and their original machines (one of which was a 603e CPU fan on a 6500) but none work on the TAM's fan header.
  • The TAM doesn't appear to have another fan header for me to try. There is a 3-pin J7 behind the ADB ports (above the Cuda) that I'm not sure what it does but it measures 4-5V across two pins. 
  • I've reset the PRAM and CUDA just to say I did but again, J17 measures 12V.
  • Seems the secondary fan (up in the upper right of the case, if you are looking at the rear) is connected to the backplane somehow. I can't get to it and it doesn't appear to run either.
  • I've swapped out the Bose power supply/sub just in case it was a +12V supply issue.
  • No other issues with the hardware that I'm aware of but I haven't gotten very far with it.
  • I do know that Accelerator, Ethernet and HD both work without issue.


Here is where it gets interesting. When I measure the voltage in series with one of the fans, the 12V drops to around 1 volt. When I take off the fan the supply goes back up to 12V. So, am  be correct in stating that, under load, the voltage is dropping to where it isn't enough to move the fan? 

I wouldn't think there is a fan controller or any sort of fan logic since it's a 2-pin header (ground and power). But could something have shorted? If so, why do I still see 12V without load?

Is there another way I can try to tap 12V from another location on the board?

Am I even following the right methodology here?

Thanks very much in advance!

 

Byrd

Well-known member
Hi pb3623,

Stepping back - I'm assuming it's an incomplete TAM that does not boot - if so why are you focusing on the fan issues initially?  Why is there a fan running off a 9V battery - can you include some pics?

I contend that the TAM does have some sort of voltage control on the fans; I'm currently using a 3DFX Voodoo 1 card in mine (with a low RPM and amperage 12V 120mm fan tucked inside the fat back mounted on plastic risers), and all fans do get louder and faster as the internals heat up - I've never checked voltages though.

What's the condition of the umbilical cord, condition of motherboard?

JB

 

pb3623

Well-known member
Hi pb3623,

Stepping back - I'm assuming it's an incomplete TAM that does not boot - if so why are you focusing on the fan issues initially?  Why is there a fan running off a 9V battery - can you include some pics?

I contend that the TAM does have some sort of voltage control on the fans; I'm currently using a 3DFX Voodoo 1 card in mine (with a low RPM and amperage 12V 120mm fan tucked inside the fat back mounted on plastic risers), and all fans do get louder and faster as the internals heat up - I've never checked voltages though.

What's the condition of the umbilical cord, condition of motherboard?

JB


Hey Byrd -

Thanks for your note!

The TAM absolutely boots - in fact, it runs fine (until it freezes), which led me to take off the fat back and see the bracket fan wasn't running. I brought up the 9V battery since I used it to test the fans themselves to make sure they weren't seized up w/ crud.

I read somewhere once that the secondary fan (that's mounted behind the outlet on the back case) only comes on when needed but I can't find that reference anywhere. On my other TAM, the fan comes on immediately, which is what I assumed is normal behavior for a 2-pin fan... a 3- or 4-pin header would be needed to control the tach and report back to the BIOS. Maybe you hearing the second fan is why it sounds louder?

The machine itself is in excellent shape - I really lucked out - cosmetically and electrically. It's been well taken care of.

I won't profess to being an EE but I assume a properly sized fan will pull the difference between the ground (+0 V) and power (+12 V). So, if a fan load of 600mA reduces the voltage to +1 V, does that mean ground is bad (and is hot)?

So @Byrd, can you measure across J17 without the fan and in series with the fan?
 

@JDW You're a PSU and fan guru. Do you have any thoughts on this conundrum? 

 

JDW

Well-known member
I won't profess to being an EE but I assume a properly sized fan will pull the difference between the ground (+0 V) and power (+12 V). So, if a fan load of 600mA reduces the voltage to +1 V, does that mean ground is bad (and is hot)?

So @Byrd, can you measure across J17 without the fan and in series with the fan?
 

@JDW You're a PSU and fan guru. Do you have any thoughts on this conundrum? 
I will profess to being an EE, but I also will admit we are certainly not know-it-alls (although many EE's try to act like it).  Electrical Engineering is a broad field, so many of us are highly focused in certain areas.  I focused on analog communications in college but ended up working more in digital (embedded MCUs, programmed in assembly language).  Even so, I am not an expert on repairing all problems in vintage Macs, which is often why I have posted here through the years to obtain sound advice from those more savvy than myself in those areas (e.g., TechKnight, etc.).

I have zero experience with a TAM.  As such, I can only comment on what I read in this thread. 

First, you said you are guessing the stock fan draws 600 to 900mA.  That is crazy high, I think, in light of the fact many stock SE/30 fans are 55mA or so (and even say that on the fan label).  My SE/30's replacement IXP-34-16 SILENX fan's specifications say it draws only 28mA.  Even in a worst case scenario for a 60mm boxer fan, I wouldn't see one drawing more than 150mA tops.

With that in mind, I am pondering what you said about the fan pulling down the 12V rail to 1V.  You said removing the fan allows the 12V rail to rise to 12V.  I would say that indicates a problem with the fan.  Cutting both wires of the fan and powering it from a 12V source (battery or power supply) and then doing voltage and current measurements would be very enlightening.  If the fan itself is bad (and it seems like it is), swapping out for a new fan should fix the problem.  If you know the CFM rating of the stock fan, that can guide you when choosing a suitable replacement.

 
Last edited by a moderator:

Franklinstein

Well-known member
 If the fan itself is bad (and it seems like it is), swapping out for a new fan should fix the problem.  If you know the CFM rating of the stock fan, that can guide you when choosing a suitable replacement.
The CFM can't be very high; they don't move much air even when they do work. The TAM was built more for form over function, so I'm sure they chose the quietest fan they could with little regard to performance. If you don't care about a little fan noise, anything in the same physical form and in the ballpark of the original fan's power rating will undoubtedly be an improvement and not overly tax the power supply.

The aux fan control on these things is totally dumb: it's something akin to a thermal switch that when tripped allows a power transistor to flow current to a 2-wire fan. There's no software control or system monitoring of fan status. The aux fan in the TAM appears to be an on-or-off sort of thing in my experience, not a variable temperature-regulated flow. The power supply fan should be a variable thermally-controlled fan, though, but the CPU fan is always on at full speed. Anyway if the aux fan doesn't turn, the power transistor doesn't know or care; it just gives power as required. If the fan overloads it, though, you'll have problems, as you seem to be experiencing. 

 

JDW

Well-known member
The CFM can't be very high; they don't move much air even when they do work.
The 60mm Silenx in my SE/30 is rating at 18CFM.  I have been told that's the same CFM as the stock fan, but when I hold my hand up to the stock fan and compare with the Silenx, it seems the Silenx moves a bit less air.  So either my Silenx fan isn't putting out 18CFM, or the stock SE/30 fan is actually rated higher than 18CFM.  18CFM isn't really pumping out a lot of air, so I wouldn't advise any lower CFM rating for a fan replacement.

 

pb3623

Well-known member
I will profess to being an EE, but I also will admit we are certainly not know-it-alls (although many EE's try to act like it).  Electrical Engineering is a broad field, so many of us are highly focused in certain areas.  I focused on analog communications in college but ended up working more in digital (embedded MCUs, programmed in assembly language).  Even so, I am not an expert on repairing all problems in vintage Macs, which is often why I have posted here through the years to obtain sound advice from those more savvy than myself in those areas (e.g., TechKnight, etc.).

I have zero experience with a TAM.  As such, I can only comment on what I read in this thread. 

First, you said you are guessing the stock fan draws 600 to 900mA.  That is crazy high, I think, in light of the fact many stock SE/30 fans are 55mA or so (and even say that on the fan label).  My SE/30's replacement IXP-34-16 SILENX fan's specifications say it draws only 28mA.  Even in a worst case scenario for a 60mm boxer fan, I wouldn't see one drawing more than 150mA tops.

With that in mind, I am pondering what you said about the fan pulling down the 12V rail to 1V.  You said removing the fan allows the 12V rail to rise to 12V.  I would say that indicates a problem with the fan.  Cutting both wires of the fan and powering it from a 12V source (battery or power supply) and then doing voltage and current measurements would be very enlightening.  If the fan itself is bad (and it seems like it is), swapping out for a new fan should fix the problem.  If you know the CFM rating of the stock fan, that can guide you when choosing a suitable replacement.


So, I had so many fans out, just for size comparisons, yeah - my scale is off. I had a couple of 1.0 and 1.2-amp fans that were obviously on the big side (~120 mm) and I didn't use them for this test. But I did have a pair of 40 mm units - 12 V @ 60 and 90 mA (not 600/900) and I know they electrically work because I used a 9V battery to power them. I agree, using something like a 12V sealed lead acid battery would be apples to apples but didn't see 9V as being a bad short-term power source (at least to rule out fan failure - debris, etc). They're obviously low torque so I'd assume you could continue to drive one all the way down to maybe 4V or so.

When I see "12V rail" I automatically assume the +12V supply coming from the PSU. In this case, it's a 2-pin fan header on the board that measures +12V, pin to probe. If I splice the wires and probe in series with the fan (thus sharing the 12V), the voltage drops to around +1V.

Am I doing this wrong? Should I be measuring current in series instead of voltage? Or measure in parallel?

Either way, what I intended to say (and this seems normal/WAI) is that the fan header appears to be capable of providing 12V, and I'm able to power the fans externally, so could it be something else?

What you seen in the picture (since you're not familiar with the TAM) - it's basically a PM 6500 logic board turned 90 degrees counter-clockwise (ports on the right). the +12V fan header J17 is half-way down on the left, above the RAM. There is a 40 mm fan connected but not running.

The stock fan has the red/blue leads. I ran two trace wires from the fan connector to test with the 9V battery (which worked).

Screen_Shot_2019-03-14_at_2_14.26_PM.png

 

trag

Well-known member
Yes, measure voltage in parallel.  Current in series.

So, if your 12V header has two pins labeled 3P and 3N (making labels up for illustrative purposes).

And your fan has terminals 4P and 4N.   You would connect 3P to 4P and 3N to 4N.   Then take your voltage measurement with the positive DMM probe on the junction of 3P and 4P and the other probe on the junction of 3N and 4N.

If you wanted to measure current you could connect things (calling the probes 5P and 5N):

3P to 5P.   5N to 4P.   4N to 3N.

Remember, the positive probe plugs into a different jack on the DMM when taking current measurements from the one you use when taking voltage measurements.

Now, all that said, I don't understand why you would get a 1V reading, if you actually inserted your DMM in a 12V circuit in series.   A DMM should have an extremely high impedance (resistance) in voltage mode, and so almost no current should flow.  With no current flowing in the circuit, the voltage reading should stay near 12V.

Either you're not actually putting the DMM in the circuit in series, or the DMM is not configured properly, would be my guess.  Possibly you have the dial set to measure voltage but the positive lead plugged into the DMM's current port rather than voltage (everything else) port.

In any case, something is amiss.

I suggest measuring the current in the circuit when you connect the fan to your 9V battery.  Also go ahead and measure the voltage while the fan is running.

 

JDW

Well-known member
@pb3623, please consider making us a short video (shot in daylight so we can clearly see everything well), showing us precisely what you are doing when you take that voltage measurement. You could upload that video to YouTube as an UNLISTED video and then post the link here.

 

pb3623

Well-known member
@pb3623, please consider making us a short video (shot in daylight so we can clearly see everything well), showing us precisely what you are doing when you take that voltage measurement. You could upload that video to YouTube as an UNLISTED video and then post the link here.
Haven't forgotten about this request. I'm happy to do that as soon as I get back in town from Spring Break holiday... probably Wednesday. 

Maybe it's my DMM, but putting it in series (3P to 5P, 5N to 4P, 4N to 3N) with the fan measured 0 amps, which I know is invalid and suggests the circuit is broken. Now, when I measure voltage IN SERIES (which I believe represents the sum of the voltage drops across each component), it's also 0. So, if I did everything right, this means the fan isn't working? But connect it to a 9V battery again and it spins up.

 

JDW

Well-known member
Be careful when using a DMM as an ammeter to measure current as some meters can't measure very much current at all before blowing their internal fuse.  Even my benchtop Fluke 8845A has two different current levels -- up to 400mA and up to 10A.  I even have to put the red probe in a different socket for each of those measurements.  If you put your meter in series with a given circuit and if you blow a fuse, your meter will read 0A (or close to it).  But in such a case, the circuit should shutdown because the only way it should get current is through your meter, and if the fuse blows, the current flow should stop.

Again, if you have a known-good 12V fan that is already disconnected and ready to be tested merely by connecting it to a +12V or +9V source, do that test to confirm operation, then put your DMM in Current Measuring mode in series with the (+) line of that fan and try to measure current.  If you still get 0A, then I would suspect a fuse inside the DMM is blown and needs to be replaced before you can measure current again.

 

pb3623

Well-known member
Be careful when using a DMM as an ammeter to measure current as some meters can't measure very much current at all before blowing their internal fuse.  Even my benchtop Fluke 8845A has two different current levels -- up to 400mA and up to 10A.  I even have to put the red probe in a different socket for each of those measurements.  If you put your meter in series with a given circuit and if you blow a fuse, your meter will read 0A (or close to it).  But in such a case, the circuit should shutdown because the only way it should get current is through your meter, and if the fuse blows, the current flow should stop.

Again, if you have a known-good 12V fan that is already disconnected and ready to be tested merely by connecting it to a +12V or +9V source, do that test to confirm operation, then put your DMM in Current Measuring mode in series with the (+) line of that fan and try to measure current.  If you still get 0A, then I would suspect a fuse inside the DMM is blown and needs to be replaced before you can measure current again.


Good morning!

I will verify the condition of the DMM fuse when I get back to town. My DMM is nowhere near the quality of a Fluke. It does have 300mA/voltage and 10A sockets. I assume the 10A dial setting is for the range and not required for operating w/ the probe set to 10A. But I've been leaving it in the default/300mA socket since we're talking about 60-90 mA fans.

 

JDW

Well-known member
Good morning!

I will verify the condition of the DMM fuse when I get back to town. My DMM is nowhere near the quality of a Fluke. It does have 300mA/voltage and 10A sockets. I assume the 10A dial setting is for the range and not required for operating w/ the probe set to 10A. But I've been leaving it in the default/300mA socket since we're talking about 60-90 mA fans.


Right, but all it takes is a few millisecond of a short to blow that fuse.  You'll get better accuracy when using the 300mA setting, but if you're not extremely careful, it may be best to use the 10A setting most of the time.  I've never blown my 10A fuse on the Fluke 8845A, but I've blown the 400mA fuse 5 times, and those things are WICKED expensive too, let me tell you!

I look forward to hearing your findings when you get back home.

 

Byrd

Well-known member
Taking the EE stuff out, are you assuming the TAM is overheating - have you pointed a desk fan at the rear innards of the machine to see if it doesn't crash? 

Also - I'd tap +12V from the optical drive (which is a modified Apple CD600 unit), if you are still looking for that.

My TAM with a very low CFM 120mm fan spliced into the fan header is struggling with L2 G3 400 @ 500 and 3DFX card, in hot Australian weather mind you. I'm going to trial a Radeon 7000 PCI in there instead, but might also have to source 12V from elsewhere with a faster 120mm fan as am concerned about the amp draw with two fans at present.  It is great though having a second video card in a TAM, noting the shortcomings with the stock video chipset and low resolution LCD.

 

pb3623

Well-known member
Well, I did find out that, somehow, my DMM's 250 mA fuse had blown (it's opaque so I couldn't visually tell - I assume it's a ceramic - but there's no continuity). So I need to replace that.

Should I be able to measure low current like this (less than 1A) using the 10A socket? As long as the range is properly set?

Talked to the seller, he states it definitely was working without issue before he shipped so it must be something that failed spontaneously (and coincidentally) or failed due to internal damage in transit.

I know voltage (potential) can exist without current (flow), but is there anything that would measure a 12V difference in potential without current? A bad ground? Does LaForge need to reverse polarity on the plasma coils? 

 

JDW

Well-known member
First things first. You need to get Commander LaForge to fix that blown fuse!   :)   When you know you are properly measuring current, you will then better understand what's going on in that circuit.

Yes, the 10A socket will allow you to measure current, albeit slightly less accurately than your 250mA socket.

 

Alexlfm

Member
How and when exactly does the machine crash? I would take things step by step and first check that it is indeed a thermal issue. I'm not convinced that it necessarily is. First of all have you checked and changed the thermal paste on the CPU? I would record the time that it takes to lock up and then immediately restart and see if it locks up faster. In theory if it is a thermal issue when its warm it should crash sooner. Also make sure you are running these tests with no accessories installed, especially the L2 cards.

The one issue I see is if you have a cheap DMM your 10A range likely is not going to be very accurate or fused so be careful not to short it out as it will not stop the current. To me it sounds like an issue with the fan control circuitry or temperature probe on the machine. As trag said, I would check the current on the fan when connecting it to the battery. It could be an issue of the fan drawing too much current on startup and then it stalling out or the machine cutting power.

 

KnobsNSwitches

Well-known member
Resurrecting an old (but not THAT old) thread, has anyone ever replaced the fan in a TAM? Mine works but is a little noisy, I'm assuming something modern would be quieter. It's an Elina Fan OPF6015L-12MB-2, but in searching for other 12v 60mm fans, none that I've seen have the odd triangle-shaped mount that the TAM fan does.
 

Byrd

Well-known member
A slimline 60mm fan with the outer shroud removed might do, but I’ve never seen the type of fan used in a TAM elsewhere. Expect some experiments and donor fans.

The TAM cooling is quite atrocious especially when you add a largish PCI card which blocks any air convection from the top vents! Im tempted to one day make a custom plexiglass or wooden “fat back” with a couple of large slow fans positioned at the back.
 

KnobsNSwitches

Well-known member
I've been incredibly pleased with the noctua fan I put in my color classic, I guess I will just try and get a 60mm noctua and see how it works.
Worst case scenario I can leave the back off, the way my TAM sits the rear isn't visible.
 
Top