SPCR rpm sensing device

Scooby · Post by **Scooby** » Fri Feb 20, 2004 8:59 am

During testing, the ZM-MFC1 did preserve the PRM monitoring. How low you can go depends mostly on the monitoring circuit. Many of those found on motherboards stop working properly below about 1000 RPM. I was able obtain RPM readings from a Panaflo 80L1A using the an external RPM sensing device made by fancontrol (of the SPCR forums). Temperature/rpm monitoring obsessives will love this product for use with the 2-lead Panaflos! (Look for a review on it soon.)

Sounds interesting. Any idea when the review will be out?

fancontrol · Post by **fancontrol** » Fri Feb 20, 2004 5:52 pm

I think I still have some lying around...

Scooby · Post by **Scooby** » Fri Feb 20, 2004 5:54 pm

could you post some details & pics please mate?

fancontrol · Post by **fancontrol** » Fri Feb 27, 2004 8:44 am

maybe. I'd have to dig them (and the documentation) out. Maybe someone that has one could do better?

The thread about making them should be helpful, but I bet most--if not all--of the pictures are dead.

patord · Post by **patord** » Fri Mar 05, 2004 2:03 am

For some dumb reason I only buy fans with built in rpm monitor, so it limits the type of fans can I buy. I saw that thread a while back but never bothered to ask if any of these devices are still in circulation...

you still ahve any for sale fancontrol?

fancontrol · Post by **fancontrol** » Fri Mar 05, 2004 10:26 am

patord wrote:you still ahve any for sale fancontrol?

I think I've got 6-10 lying around. I'd have to test them to know if they're any good. Send me a message if you want one. Price will vary based on shipping.

fancontrol · Post by **fancontrol** » Sat Mar 06, 2004 4:29 pm

I just wrote this to patord, thought I should post it here:

I'll send you as many as you want (I found seven) as-is for the price of a padded envelope and postage. Looks like that'll be about $2. Of course I'd appreciate a couple of bucks for the parts and whatever, but I'd pretty much written them off a while ago.

I can test them, make the cables you need, etc. but I'd have to insist on a few more dollars for my time and parts. Probably $5 ea.

Let me know how you'd like to proceed.

//edit: what's with the ad?

fancontrol · Post by **fancontrol** » Sun Mar 07, 2004 12:41 pm

here's the best i can do in the time i have.
schematic:

pile-o-parts:

the two thru holes between the connector act as the switch on the schematic.

not pretty, but they work (and, now, they're cheap).

Gotta run; I'm pouring new piers for the back deck. Concrete waits for no man.

Money talks, as they say. paypal [email protected].

patord · Post by **patord** » Sun Mar 07, 2004 5:03 pm

I have dibs on a few of these... PMs sent to fancontrol.
Thanks for the pix.

sthayashi · Post by **sthayashi** » Mon Mar 22, 2004 10:09 am

Fancontrol, I don't want to subvert your efforts on your sales, but maybe you can help me out.

I went through my pile o' components and found that I had everything in that schematic (well, almost, but I was ordering some switches at work from Digikey anyways). I put it together and got some interesting readings on my oscilliscope. Basically, it was an interesting signal output, but not very square wave (and it didn't vary when the fan speed changed). The only substitution that I made was an Electrolytic capacitor rather than a ceramic one.

fancontrol · Post by **fancontrol** » Mon Mar 22, 2004 11:58 am

did you stuff R1? I know it's listed as a DNS part, but that's because I wanted to try different resistances once the boards were made (hence the thru-hole requirement). ~10ohm is about right.

Once you get that, try messing with the pot until the output starts to show up in a even pattern of negative spikes. Watch the + side of the comparitor, you should see the same thing. The LM acts as a level shifter and buffer btw the current sense resistor (R1) and the tach input.

And, I have no efforts on sales. No worries. Best of luck.

sthayashi · Post by **sthayashi** » Tue Mar 23, 2004 9:32 am

I'm wondering if I'm missing a component somewhere. The oscilliscope signal bounces slightly when the fan is on, and the bouncing is proportional to the fan itself. Yeah, I have a 10 Ohm current resistor acting as current sense.

The oscilliscope output starts with a 20mV 2.5ms pulse, 7.5ms rest, and then a -20mV 2.5ms pulse. That basic shape stays, even when the fan isn't spinning

I don't suppose you'd have any idea what might be causing it. Maybe tomorrow I'll take a picture of the circuit in the breadboard to show how it looks.

Scooby · Post by **Scooby** » Tue Mar 23, 2004 11:09 am

fancontrol wrote:
patord wrote:you still ahve any for sale fancontrol?
I think I've got 6-10 lying around. I'd have to test them to know if they're any good. Send me a message if you want one. Price will vary based on shipping.

I'd be be interested in having one if you've got any left...depends on how much it will be to ship to the UK.

fancontrol · Post by **fancontrol** » Tue Mar 23, 2004 3:57 pm

sthayashi - no clue. assuming you're looking at the (+) input to the comparitor you should never get negative pulses; that implies the fan is generating power. 20mV is equivalent to only 2mA, I bet your fan is drawing more current than that.

Are you sure you have the scope set to a DC coupling and that it is well grounded to the circuit?

What do you get when you just hook the scope to +12? You could be seeing the switching frequency of your power supply, but that's a long shot.

Scooby - no more.

afiuw · Post by **afiuw** » Wed Mar 24, 2004 9:20 pm

Sthayashi, did you manage to figure out your problem?

Also, what's a "DNS part"?

sthayashi · Post by **sthayashi** » Thu Mar 25, 2004 6:29 am

Not yet, but I haven't been working on it much. I work on it at work, but after hours or during lunch. But today and yesterday, I've been working with contractors at their location and I can't bring myself to come in to the office JUST to work on it.

fancontrol · Post by **fancontrol** » Thu Mar 25, 2004 7:30 am

DNS stands for "Do Not Stuff" i.e. leave the part out when the board is populated.

One more thing occurred to me last night. The '393 is an open collector output, just like a fan; it won't drive the signal one way or the other. If you want to see if it works on a bench, you'll need to add a pull-up resistor (10k is good) between +12 and the output.

Tach sensing circuits, typically, have that pull-up built in.

sthayashi · Post by **sthayashi** » Mon Apr 05, 2004 2:24 pm

I just got a chance to play around with it again now. Here's the update:

The circuit doesn't appear to work for the Panaflo 120mm L1A, and it barely works for the Panaflo FBL 80mm.

When I probe the non-inverting input of the comparator, I get a great signal that could be a passable fan Tach. For the 80mm, it's a peak voltage of 1.2v. The 120mm has a peak voltage of 2-2.5V (I'm checking out with an Oscilloscope).

The inverting input isn't good at all. Is that supposed to be a reference point that sits at mean voltage of the non-inverting input?

Thanks again for all your help on this.

fancontrol · Post by **fancontrol** » Wed Apr 07, 2004 7:13 pm

yes, the inverting input is a reference voltage.
different voltages are required for different fans.
hence the potentiometer.
hence the thru-hole current shunt.

sthayashi · Post by **sthayashi** » Thu Apr 08, 2004 3:46 pm

WOOHOO!!! It worked!!!

I had to swap most of the resistors with something else BUT IT WORKED!!!

What I did differently was:
a) replaced the DNS resistor with a low resistance current sensing resistor I bought a while ago (0.15 Ohms).
b) replaced the 100 Ohm resistor with a wire.
c) replaced the 15k resistor with a 100k resistor.

One thing to note for anyone trying this at home: There are 4 periods per rotation with this circuit, not the typical 2. The upshot with that is that it can be theoretically read on most computers. You just have to remember to divide the RPM signal by two.

Now what I need to do is figure out the best way to display tach onto an LED screen without needing to use a microcontroller.

fancontrol · Post by **fancontrol** » Fri Apr 09, 2004 12:27 pm

without thinking too hard about it, i think you could pull it off with an oscillator, counter, and--probably--a binary to bcd converter.

the counter gets hooked to the fan output. the oscillator latches the count and resets it. the bcd converter takes the count from the output and puts it in a format that the lcd will enjoy, and--presto--you're done. it wouldn't be that hard to do in a PLD, either, and that's not technically a micro

frankgehry · Post by **frankgehry** » Sun Mar 06, 2005 9:15 pm

I'm interested in the 2 wire fan rpm sensing project that bxwrapper and fancontrol worked on. Unfortunately the links to the photos and diagrams have been broken.

Did anybody make copies? TIA - FG

jinu117 · Post by **jinu117** » Sun Mar 06, 2005 10:10 pm

Was looking at it myself yesterday...
http://forums.silentpcreview.com/viewto ... an+panaflo
Still reading through and trying to figure out from whatever information it had. Wonder if that is good enough information. (lot of good fan controllers out there now with good functions except lot of them will go nutz on 0 RPM)

frankgehry · Post by **frankgehry** » Sun Mar 06, 2005 10:26 pm

J,
I sent a pm to fancontrol, bxwrapper, and hopefullly they will still have some of the stuff. There are a lot of circuits on the web for contolling speed but I haven't seen one for getting rpm from 2 wires. I have a t-balancer and that's what made me interested in how all this stuff works. - FG

frankgehry · Post by **frankgehry** » Mon Mar 07, 2005 4:57 am

For the programmer types I found this developement system that includes a programming language, a usb interface, a project board, and some circuits. In particular I'm interested in the miniature stroboscope using leds to measure fan rpm. You create a target on the rear hub of the fan. Half is black and the other half white. Its really clever. Its recommened that you be 17 or older. - FG

http://www.parallax.com/dl/docs/books/edu/ic.pdf

JanW · Post by **JanW** » Mon Mar 07, 2005 5:41 am

frankgehry wrote:There are a lot of circuits on the web for contolling speed but I haven't seen one for getting rpm from 2 wires.

FG, there is one site linked from the end of the original 2-wire RPM sensing thread.

SPCR rpm sensing device

SPCR rpm sensing device

Anyone have fancontrol rpm prj. diags.? - links dropped

rpm sensing