OK, first a little background: I know nothing about electricity...I vaguely remember touching an electric fence once when I was little and not liking it. So on to my question. I bought this adapter to switch fans to 7V from directron.com.

http://www.it9.net/quiet/7v.jpg

I have tried it and it works. I understand it combines the 5 and 12 volt leads to somehow get 7. Im wondering what the use of this particular adapter is besides just converting to 7 volts, as it appears to be some kind of through connector and I should be able to plug other stuff in, yet the other wires dont follow through. Is this bad design, or should it work anyway? (Im sacred to try for fear of frying something). Can someone explain what this adapter does?

nate, there's 3 plugs on that adapter, one's a female molex plug (the one with the wires on the plug spaced far apart) this is the one that is tapping 12V and 5V from the normal molex connections. Then there's a male molex plug ( the big one with the four wires going into it) .... with the plug 7V comes out of the plug that has four wires going into it and the smaller plug that is there is a 3 wire plug (looks like the one's on your motherboard for plugging in CPU fans and the like) that one is also setup for 7V as well, I think they've done this so that u can run the fans that have 3 pin connectors on it (the smaller ones like one's u find on CPU heatsinks) at 7V.....
Hope this makes sense to u
PS. Nice webpage u've got there as well

Thanks
I think I get it quokked: Youre saying that with this adapter I can get 7 Volts from either the 3 pin connnector (which I have been using) or the 4 pin male molex plug. That makes a little more sense. It would be cool to make one that would be a through connector with just a 3 pin 7V connector coming off to the side. Is that possible? Also, I was curious about something I read (I think in these forums) about 7 volt mods being hard on the PSU...is this true? whats that all about?

The PSU doesn't have a 7V line; this is derived from the positive leads of the 5 V & 12V lines -- it's the difference, so to speak. Without getting technical, as long as you have lots of other stuff drawing power from both 12V and 5V lines, and you don't have a lot of fans on 7V, it's safe. But it *can* be more stressful for the PSU. If someone (like TerryW or a quietly lurking EE) wants to spell it out in glorious detail, by all means.

yeah u can do that as well, get a molex double adapter plug and cut off one of the two male molex plugs on it.
- Cut the ground wires off that plug (black) and wire the yellow wire to the +12V of the 3 wire plug
- and the red wire (+5V) to ground of 3 wire plug.... http://www.overclockershideout.com/7voltmod.shtml this article should help u with the info as well.....
good luck

PS.
- get a pin out of the 3 pin connector off the web somewhere, I've confused myself a few times and it's only 3 wires

- Solder and heatshrink the wires 2gether on the 3 pin connector, easy and quick to do and gives u extra security, I shorted my 12V wire to ground on some of my earlier fan connectors killed a couple of mobos that way

Here's the easiest to make 12V to 7V or 5V molex: http://www.dslwebserver.com/main/fr_index.html?/main/5-7-adapter.html. I use these in several of my systems to test 5V vs. 7V and end up leaving them in most of the time.

Here's another way to change a 4 pin to 3 pin molex cable to use the 3pin fan connector at 7V (or 5V): http://www.overclockers.com/tips766/.

Note that there is no need to cut the wires or solder anything in either of these HOWTOs.

And here's a guide to the wiring on the 4 pin and 3pin molex: http://www.heatsink-guide.com/psplug.htm

Hi there.

My 2-bits worth: I like the links that Renod provided. Easy and straightforward. No "rocket-science" here!

If anyone can explain this I would be interested. I have 5 fans running at 7V at the moment from a 300w PSU.

Hi there Shunx.

I am no *expert* in switching PSU, but can only relate my experiences as a technician for over 30-years. I have seen most every type of PSU designs, & even designed & built some.

The switching PSU is designed with multiple windings and/or multiple transformers to provide the various DC outputs @ rated currents. To achieve 7VDC, one connects the red fan wire to the +12VDC and the black fan lead to the +5VDC, hence 7VDC. Easy.

One possible concern: some switching PSU designs may not like a 7VDC hookup @ high currents, as it may unbalance the PSU. I seriously doubt whether PSU designers took into account that users might use the +12/+5 in this manner. However, I would think that small loads @ 1 or 2 fans might be okay, but 5 or more fans could cause a problem.

Of the switching power supplies I have used @ work, the manufacturers recommend against using such a configuration. What is recommended is starting @ 12VDC and regulating down to +7VDC. To lower the 12VDC down to 7VDC can be done various ways: resistors & zener diodes, 3-terminal regulators, etc.

Just some of my thoughts.

TerryW

Here's another explanation of why the 7V trick is not recommended by PSU makers: http://people.freenet.de/s.urfer/7vtrick.htm

Generally, it seems not to be a problem unless you're trying to draw a lot of power from this 12V / 5V "difference" line. How much is too much? I don't know, but I have had 3-4 Panaflo at 7V w/o apparent problems on a few PSU. I say apparent, because that might have been one of the things which hastened the demise of some of these PSUs.

Thanks MikeC

Good comments & link.

Many years ago, I used a variable DC circuit for an alarm for my classic Mustang. I started with connecting a Motorola MC7805 (3-terminal 5VDC regulator) onto the 12VDC souce and putting a potentiometer in the "common" lead to vary the output voltage from 5 to 12.

One 7812 can handle up to 1-Amp and could easily run up to ten 12VDC fans (assuming fans are 100mA each).

Mike, maybe you & I should get together & make such a device. The downside is we need to rival the $10.00 fan speed (5-7-12V) control that is locally available.

Your comments?

TerryW

Hope you don't mind my butting in... 25¢ worth...

A PICµ could do a good job of controlling several fans, and reporting back via the SMB as per fan speeds (make sure it senses well at low voltages). Maybe even SMB based speed control (is this the way it's done on intelligent mobo's?) to give less advanced mobo users access to new tech (assuming driver interoperability). The best voltage regulation scheme (as demonstrated by PSU makers) is switching. That's easy fair for PWM outputs on a PIC driving some cheap FET's. Fancy front panel controls would be nifty, but a well done 'black box' scheme is perhaps just as viable; maybe two versions. Also easy enough to do would be thermistor based auto temp control, although it certainly would up the ante a bit.

Well, one of my next projects is probably building a PIC based survey-instrument-driver / to-GPS-format converter, using my PICstart rig. Maybe it will put me in the mood, given apparent market at that time

That all sounds good Cris, but PWM is OUT for quiet fan speed control. I have been working with a PWM-based thermal fan controller & it is not viable, because at slower speeds the pulsing causes terrible variable noise from most fans. Something about timing issues. BRRR-mmmm-BRRR-mmmm-BRRR-mmmm -- faster and faster till it's almost at full speed. It may have to do with the position of the armature when the pulses hit.

I think PWM is probably fine for fans at higher speed in louder apps, but for quiet computing, the only way it could be made to work well is to build the fan and controller as an integrated pair -- and the inflexibility would be very limiting. I recommend you stick with straight voltage control with ensured starts for low speed RPM using z-diodes or similar.

Mike, I think the key to your statement is that the one you've been trying isn't viable. PWM must be viable since that's how the whole PSU works. My guess is that the unit you have uses too low a frequency ("pulsing causes terrible variable noise"), and doesn't filter the output because inductor+capacitor filters are more than the designer felt they needed to invest in the problem.

Typical 'better grade' switcher/pwm designs use 100+khz, and L-C output filters to pass only clean DC to the loads. Like the PSU and the mobo's Vcore regulator. I know those two examples are high end compared to the target cost fit of a fan controller, but I expect that a suitable lower grade comprimise is possible. One potential snag off the top of my head though is the available source voltage. Most switchers use either high voltage, or invert their own. PSU's use rectified and/or doubled 120v AC, around 170-340v DC (those caps can pack a punch!). Starting at 12v and not inverting might seriously limit the maximum output, like maybe you would loose the top 25% of the speed range. Even a linear regulator will have that problem, but to a lesser extent (approx -.5v).

An intelligent control could also be done by using D to A output to control linear regulators. PIC's use PWM to make up their D to A outputs, but I think the frequency is quite high, and can be very economically filtered at such low power levels. To appeal to the flashy modder crowd, you could even use tubes for the output stages Maybe don't laugh, there are $30000 audio CD players out there, and I don't think that price even includes the D to A converter I'm glad the computer crowd doesn't seem to have gone that way yet, and I'm not likely to encourage it, perhaps unlike Aopen.

Yes, but the app is so different. I agree it is possible the particular PWM fan controller I have here may not be a particularly good one, but the experience I have is not unique -- others report similar with different PWM fan controllers. I've been told that a single cap would solve the problem -- where/how do you think?

I will call the controlled positive output Vfan. Put a big-ish cap on the output between Vfan and 0v. A small resistance in series after the cap (between the cap and the fan) will also help, but too much will cut the fan's top end too much. Or better yet an inductor. If you want to get fancy, use a multi-meter with a frequency measurement setting and find out the modulation frequency of the controller, then use the L-C equations to calculate the inductance to match the capacitance at that frequency. AC on the output is the problem. So putting in a cap (Vfan -|(- 0v) will buffer the AC, and an inductor in series (controller -UUU- fan) will resist the AC. Sorry for the cheesy ASCII graphics, I couldn't help myself

The only other caveat is that using too large a cap can cause a large inrush surge when the power first goes on. The cap has to get charged up to Vfan from 0v, and behaves almost like a dead short for a breif time. If you go too big, that surge could smoke your controller. This is not a big worry, just don't hook on a surplus pop-can sized monster cap thinking that 'bigger is always better'.

I would start with something like 500 to 1000µf, and work up from there. look at the caps on the MOBO. They are trying to do pretty much the same job.

Crisspy,

My turn to butt-in! My 2-bits....

Ooooo - A very detailed description for a fan controller, but not likely to be inexpensive.

I am in awe - a fellow tehno-junkie after my own heart! Arise!

Where do you get all these neat ideas? Do you spend your waking hours delving into techy manuals and reading up on this material?

As for the caps: I would probably start with something smaller than 500 to 1000uF. The inrush current @ 12VDC @ 1000uF is quite high and could damage your controller.

Cheerio!

TerryW

Unfortunately Terry, as yet with analog electronics my knowledge is mostly cut&paste. I understand enough to be dangerous, and I understand quite well in principal, but I still haven't done my homework with all the math so I can't quite design all the stuff that I generally know about. Kids and women came first I guess aaah... maybe not quite in that order I have done a fair bit of digital though, and the example circuits for most IC's are pretty straight forward and workable as long as they don't get to complicated.

I get the impression that when any electronics start to get more sophisticated, you need to be a big manufactury to get the costs down and crank them out at profit. The smaller volume stuff we see in computer niche markets seems to come mostly from Asian tie-ins or else is fairly basic stuff. It should be possible to do a good multi channel smart PWM fan controller, but keeping it cheap would be big-boy high volume stuff. Doesn't seem very viable at the homebrew/cottage level, too bad. I won't forget about it though, and I'll keep my eyes open when I dig into my next PICµ project. There are some pretty cool shortcuts and who knows what might turn up.

An LC circuit is an capacitor (C) and an inductor (L) connected together. This is a circuit that acts as a band pass filter that resonates where the frequency of the effective resistance of L = the effective resistance of C.

This is called a tank circuit and it is used extensively in RF (radio) circuits where these are usualy called RLC circuits. The R is for resistance because the Q (bandwidth) of the circuit can be changed by adding or removing resistance. More resistance will decrease Q which means the bandwidth will increase.

The output of a PWM is a square wave which is basicly an AC current. Square waves because of thier shape (the sharp corners) have very high level of hamonics (noise). A tank circuit tuned to the same frequancy as the PWM output will remove the harmonics and give you something that approachs a sin wave AC current. So saying that this type of circuit is removing the AC is not correct as it is actualy removing the noise components from the AC and giving you a cleaner AC current.

Will adding a tank circuit to the output of a PWM make the fan run with less noise? I don't know and I would think that it would depend on what is causing the noise. If the noise was caused by the shape of the pulses it could.

Hal (AC6VZ)

A band pass filter is sort of what we need. One not tuned to the PWM frequency. Simple L-C tanks act more as a cut-off than a band pass. That is, they reduce anything above or below the LC resonant frequency, depending on the configuration. The configuration I described is a simple low pass. IE. reduce/block energy at frequencies above resonance. That's exactly how almost every I/O on the mobo gets treated before it hits the back panel (very small cap to ground, ferrite bead around wire), and how a noise trap for car radios works (in power leads), and why computer cables sometimes go through ferrite rings. All of these use a series L and a C to ground. Passes lower frequencies (including of course DC), reduces higher frequencies (RFI), helps meet FCC specs (is why they spend money doing it at all).