Will resistors really do the job?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
Will resistors really do the job?
I am thinking about using resistors to mod a few fans down to 7 volts. I would rather not have to deal with the extra cabling a passthrough cable or a fan controller would bring and I want to still get RPM readings from my motherboard. I don't have the need to adjust the speed around, I'd rather have it hardwired to a particular voltage.
Using the FanCalc .4 tool found here to get 7v on the two different fans I would need around a 85 Ohm resistor on one fan and around 140 Ohm resistor for the other fan.
Would these really get the job done to the figures of the FanCalc tool or should I explore a different route?
Using the FanCalc .4 tool found here to get 7v on the two different fans I would need around a 85 Ohm resistor on one fan and around 140 Ohm resistor for the other fan.
Would these really get the job done to the figures of the FanCalc tool or should I explore a different route?
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
That sounds about right.tay wrote:I have never used the fancalc tool but I've done this several times. 66 ohms gave me 8.5 v on a L1A panaflo 80mm for reference. Can you link to the tool?
http://www.silentpcreview.com/files/misc/fancalc.zip
there is a much easier and neater solution: you can get 7volts directly from your power supply! that's what i've been doing for my psu fan, case fan and hdd fan for the past year or so.
all you have to do is modify a molex plug (male or female, although male is easier) to take the (+) power from the 12v cable (yellow) and the (-) power from the 5v cable (red). voila, 7 volts
all you have to do is modify a molex plug (male or female, although male is easier) to take the (+) power from the 12v cable (yellow) and the (-) power from the 5v cable (red). voila, 7 volts
-
- Posts: 30
- Joined: Wed Aug 04, 2004 4:55 pm
- Location: Pelham, AL
Zener Diode
There is yet another way to 7 volt a fan. Go to Radio Shack and buy a 5 volt Zener diode. Insert the diode into the 12V fan lead. The Zener will drop 5 volts, leaving you with 7V to the fan. You can buy a 2 pack of 1 watt Zeners at Radio Shack for about a buck. If you are not sure that a 1 watt rating is sufficient, simply connect both of them in parallel.
In the future I will be using the diode method if necessary mentioned by Bill (I am all out of resistors). I am currently using the 7-volt method mentioned by chylld on systems I use but never for ones I build for relatives or friends (too risky they'll switch plugs). I have used resistors in the past but theyre too much of a pain to match with the fans etc. The diodes have a constant voltage drop which is a huge advantage.
You can use a zener or even an avalanche diode. The avalanche diode is meant to be used reverse biased, but is difficult to find. Radioshack doesnt carry them unfortunately.
You can use a zener or even an avalanche diode. The avalanche diode is meant to be used reverse biased, but is difficult to find. Radioshack doesnt carry them unfortunately.
that's a very good pointtay wrote:I am currently using the 7-volt method mentioned by chylld on systems I use but never for ones I build for relatives or friends (too risky they'll switch plugs).
i don't understand why nobody makes a simple plug that drops the fan voltage using zener diodes, they seem to be very effective for this purpose (or maybe someone does?)
Putting zener diodes in parallel to share the power dissipation is a bad idea. Because the breakdown voltages won't be exactly the same one of the diodes will hog all the current and get hot and the other will not. Just get higher wattage zeners, or if you must, put two in series that are half the voltage. Personally I find the $4 Zalman Fanmates irresistable. They are very compact, adjustable, cheap, and trivial to install.
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
I am having a hard time finding anything over 1w, the next level seem to dissapate 5w, but that would be OK to use right?ferdb wrote:Putting zener diodes in parallel to share the power dissipation is a bad idea. Because the breakdown voltages won't be exactly the same one of the diodes will hog all the current and get hot and the other will not. Just get higher wattage zeners, or if you must, put two in series that are half the voltage. Personally I find the $4 Zalman Fanmates irresistable. They are very compact, adjustable, cheap, and trivial to install.
i'd agree as well. i use one of them to slow down both my 4412fgl's (wired in parallel) and the ability to vary the voltage at the turn of a dial is a godsend. (even though i only adjust it between minimum and ~7v )ferdb wrote:Personally I find the $4 Zalman Fanmates irresistable. They are very compact, adjustable, cheap, and trivial to install.
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
Don't care for the extra wires. I have a Antec 3000B mid-tower with acoustipack in it, space is a premium.chylld wrote:i'd agree as well. i use one of them to slow down both my 4412fgl's (wired in parallel) and the ability to vary the voltage at the turn of a dial is a godsend. (even though i only adjust it between minimum and ~7v )ferdb wrote:Personally I find the $4 Zalman Fanmates irresistable. They are very compact, adjustable, cheap, and trivial to install.
I have had a 12-7v Speed Reduction cable for ages & never used it. Basically it's 3 components sealed in heatshrink with m&f 3pin plugs either end. I am sure everyone will be familiar with these.
I always thought it would contain some oversized resistors. Seeing this post I decided to dissect it! Let's say in the pursuit of silence. I was surprised to find 3 Rectifier Diodes in series. They are marked 1N5401 gave a voltage drop of 4.6v.
Maplin (UK equiv of Radio Shack) charges 10p/20c each for these Diodes, retail. In the Far East it's probably a fraction of that. As Mr Akasa charged mugface here about £4/8USD for this cable I feel a bit swindled now I know what's inside it. Having said that I was planning to fit it in someone elses PC, so safety was a consideration.
A similar cable could quite easily be made at home. Whether or not 3 Rectifier Diodes in series are suitable for this application I really don't know, but they did seem to work.
I am no electronics whiz & have only reported my findings for the sake of this thread. As these cables are being retailed it would be interesting to hear any thoughts?
I always thought it would contain some oversized resistors. Seeing this post I decided to dissect it! Let's say in the pursuit of silence. I was surprised to find 3 Rectifier Diodes in series. They are marked 1N5401 gave a voltage drop of 4.6v.
Maplin (UK equiv of Radio Shack) charges 10p/20c each for these Diodes, retail. In the Far East it's probably a fraction of that. As Mr Akasa charged mugface here about £4/8USD for this cable I feel a bit swindled now I know what's inside it. Having said that I was planning to fit it in someone elses PC, so safety was a consideration.
A similar cable could quite easily be made at home. Whether or not 3 Rectifier Diodes in series are suitable for this application I really don't know, but they did seem to work.
I am no electronics whiz & have only reported my findings for the sake of this thread. As these cables are being retailed it would be interesting to hear any thoughts?
-
- Posts: 30
- Joined: Wed Aug 04, 2004 4:55 pm
- Location: Pelham, AL
Using a string of diodes is an old tried and true method of dropping voltage. I'm a bit puzzled by your results though. The 1N5400 series are 3 amp rectifiers and usually have a drop of around 0.6-0.9v per diode, how you got 4.6V drop out of three seems strange unless you were running a great deal of current through them.Krispy wrote: I was surprised to find 3 Rectifier Diodes in series. They are marked 1N5401 gave a voltage drop of 4.6v.
1N5401's will work fine for the application but they must have had a cheap surplus supply of them to use them in that application since the 1N5400 series is usually more expensive and the voltage drop per diode is a little lower than the 1N4000 series. Either that or they expected them to be used in an application where the current would exceed 1 amp. If you want to construct your own diode chain voltage dropper I would suggest using any of the 1N4001 - 1N4007 type diodes. They are the most broadly available and usually very cheap. They are rated for 1 amp btw.
Hi ferdb, I agree with your point about the 0.6-0.9v drop per diode in this instance & that a 4.6v drop is strange. When I was taking these readings I thought it would be straight forward, but actually it was all a bit wierd.
To measure this drop I used a small brick transformer, the diode cable & a 12cm Panaflo fan. I have since repeated this test & got broadly the same result. I have also taken measurements from a spare 3pin plug on my PC & got quite a different result
As I said I am no electronics whizz & this was how I obtained these readings: I first multimetered the source, no problem.
I then connected up the diode lead & recorded only an approx 1v drop @ the end of the lead. I was confused.
I then connected a 120mm Panaflo & took a reading from the slightly exposed pins where the wires connect to the fan. This was the only way I could get a higher voltage drop & leads me to think this must be load related? Or maybe I am doing something wrong?
Anyway, here are the figs. If you could make some sense of them, that would be great
Brick ......13.16v
@ Lead ..12.21v...(-0.95v)
@ Fan ......8.70v...(-4.46v)
From PC ...12.20v
@ Lead ....11.09v...(-1.11v)
@ Fan ........9.95v...(-2.25v)
It seems that when connected to a PC this cable doesn't seem to do what it's supposed to? This seems to be another fine mess I've got myself into!
Maybe Sizzle should stick with resistors?
To measure this drop I used a small brick transformer, the diode cable & a 12cm Panaflo fan. I have since repeated this test & got broadly the same result. I have also taken measurements from a spare 3pin plug on my PC & got quite a different result
As I said I am no electronics whizz & this was how I obtained these readings: I first multimetered the source, no problem.
I then connected up the diode lead & recorded only an approx 1v drop @ the end of the lead. I was confused.
I then connected a 120mm Panaflo & took a reading from the slightly exposed pins where the wires connect to the fan. This was the only way I could get a higher voltage drop & leads me to think this must be load related? Or maybe I am doing something wrong?
Anyway, here are the figs. If you could make some sense of them, that would be great
Brick ......13.16v
@ Lead ..12.21v...(-0.95v)
@ Fan ......8.70v...(-4.46v)
From PC ...12.20v
@ Lead ....11.09v...(-1.11v)
@ Fan ........9.95v...(-2.25v)
It seems that when connected to a PC this cable doesn't seem to do what it's supposed to? This seems to be another fine mess I've got myself into!
Maybe Sizzle should stick with resistors?
If one has a fan with a pass-through cable (both male and female Molexes, so that one can chain devices), a viable solution is to modify both Molexes the same way. This way the other connector can still be safely used for other devices. It'll look wicked though, since the wire colors don't match between connectors...chylld wrote:there is a much easier and neater solution: you can get 7volts directly from your power supply! that's what i've been doing for my psu fan, case fan and hdd fan for the past year or so.
all you have to do is modify a molex plug (male or female, although male is easier) to take the (+) power from the 12v cable (yellow) and the (-) power from the 5v cable (red). voila, 7 volts
Another thing is whether the 7-volt mod is safe or not. It can easily kill a PSU.
Cheers,
Jan
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
Plus it means more messy cables. Only have room to hide so many.Jan Kivar wrote:If one has a fan with a pass-through cable (both male and female Molexes, so that one can chain devices), a viable solution is to modify both Molexes the same way. This way the other connector can still be safely used for other devices. It'll look wicked though, since the wire colors don't match between connectors...chylld wrote:there is a much easier and neater solution: you can get 7volts directly from your power supply! that's what i've been doing for my psu fan, case fan and hdd fan for the past year or so.
all you have to do is modify a molex plug (male or female, although male is easier) to take the (+) power from the 12v cable (yellow) and the (-) power from the 5v cable (red). voila, 7 volts
Another thing is whether the 7-volt mod is safe or not. It can easily kill a PSU.
Cheers,
Jan
i think with low speed fans it isn't a problem. i've been running 2x l1a and the bqe case fan this way for a year or so, and i can't remember the last time i had a system crash also i'm running a fairly power hungry cpu (axp@2v) and the psu is modified for low airflow, so i'd think it's safe to say that the average user with a 300w+ psu will be able to do this trck safely for at least 1 or 2 low-speed fans.Jan Kivar wrote:Another thing is whether the 7-volt mod is safe or not. It can easily kill a PSU.
ymmv of course
It's all about how the PSU's protection circuit is made, and if there are other devices connected normally in the same set of wires.chylld wrote:i think with low speed fans it isn't a problem. i've been running 2x l1a and the bqe case fan this way for a year or so, and i can't remember the last time i had a system crash also i'm running a fairly power hungry cpu (axp@2v) and the psu is modified for low airflow, so i'd think it's safe to say that the average user with a 300w+ psu will be able to do this trck safely for at least 1 or 2 low-speed fans.Jan Kivar wrote:Another thing is whether the 7-volt mod is safe or not. It can easily kill a PSU.
ymmv of course
Basically one is pushing current back to the PSU via the +5V wire. If there's a hard drive or a CD/DVD in between (or other device which uses +5V rail), the illicit current will go there. If there are no other devices connected to the cable from the PSU, all sorts of things could happen. Normal operation, PSU doesn't turn on at all or shuts down after 30 secs, smoke...
I wouldn't recommend the 7-volt mod unless there is a device "upstream" which will use the illicit current.
Cheers,
Jan
This makes sense really. It's only a measurement error. Your wall transformer is either AC or has a large amount of ripple on the output. Your meter is probably not a True-RMS type and so is getting fooled by the crest factor of the AC it's seeing. The results you get when you hook it up to your PC powersupply are the ones I believe to be fairly accurate, they are right around where I suspect they should be for 3 diodes.Krispy wrote:To measure this drop I used a small brick transformer, the diode cable & a 12cm Panaflo fan. I have since repeated this test & got broadly the same result. I have also taken measurements from a spare 3pin plug on my PC & got quite a different result
Anyway, here are the figs. If you could make some sense of them, that would be great
Brick ......13.16v
@ Lead ..12.21v...(-0.95v)
@ Fan ......8.70v...(-4.46v)
From PC ...12.20v
@ Lead ....11.09v...(-1.11v)
@ Fan ........9.95v...(-2.25v)
It seems that when connected to a PC this cable doesn't seem to do what it's supposed to? This seems to be another fine mess I've got myself into!
Maybe Sizzle should stick with resistors?
Although people always say diodes drop 0.6-0.9 volts it's not a fixed voltage but depends on the current through the diode. The 0.6-0.9 volts figure is for currents in the 1mA to 1A range. Depending on the meter you use it may read much less than that if the current from the meter through the diode is in the microamp range.
Diodes work fine for voltage dropping. Just put enough of them in series to get the results you want with the load you are working with. There are actually one or more diodes in the DC fans themselves.
-
- Patron of SPCR
- Posts: 634
- Joined: Wed Mar 10, 2004 5:01 pm
- Location: Saginaw, Michigan
- Contact:
Success! I got a pack of 5.1v 2w Zener diodes from Digi Key in the mail today. Tried one out on a Panaflo 92mm L1A-BX today. Just going by the rpm monitor, the Zener has dropped the fan to 6.8v. The diode is warm, but this fan has input power of 1.32 watts, the fans I am going to use these on have input power rated at 1.2 watts.
Thanks for the good advice.
Thanks for the good advice.
A 1W 5V zener (in theory) can provide up to 200mA of current. (Power = Voltage * Current, or Current = Power / Voltage)
So, if your fan is rated at pulling less than 200mA of current at 12V, then you will be fine, since your fan will be running at 7V, and therefore will be pulling less current anyway.
What if your fan is rated in power? Then the 1W 5V zener should be able to handle a fan that uses 2.4 Watts (0.2A * 12V). Again, in this case, I used 12V, when your fan would be running at 7V, so it would be consuming less than the 2.4W.
So, if your fan is rated at pulling less than 200mA of current at 12V, then you will be fine, since your fan will be running at 7V, and therefore will be pulling less current anyway.
What if your fan is rated in power? Then the 1W 5V zener should be able to handle a fan that uses 2.4 Watts (0.2A * 12V). Again, in this case, I used 12V, when your fan would be running at 7V, so it would be consuming less than the 2.4W.