Electronics SATA Question/Challenge

Hi there i love the ideas coming in on this forum when i get some time i might try a few of them especially the cable tidying ones since mine are a mess, but heres the question im using sata drives in my computers, how can i make an led activity monitor for it, is it possible?

 

Moving to electronics...

 

Your question doesnt seem to make sense. What do you want to make an LED activity light for?

edit: Ah now your question makes sense. I will look into how to do it.

 

for the hard drive, like zapwizards one

 

Ok looked into it abit. According to some very preliminary research there are 2 pairs of data wires going to/from an SATA drive Tx+ Tx- Rx+ and Rx-. There is no activity wire so you would have to make one. The easyest way I can see of doing this is to moniter Tx+ and if that is "high" (i.e. Serial data is going down it) then turn the light on and if it is low turn it off. You will probably need a high impeadance takeoff from the Tx+ in order to prevent signal degridation. It should be very doable though.

 

ummm ok if you havent guessed already im new tot his and that just gave my head a banging . maybe if you could explain with diagrams (i kno lots of hassle) dont worry if u dont wanna im sure i can jus buy another sata cable if it screws up. , but could you post just a little more simple explaination but thanx for you help so far.

 

OK your SATA cable has 4 pins. these are 2 data pairs. This means that the data from 2 pins goes into producing the output. The pairs are Tx (for data from the mobo to the HDD) and Rx (the other way round). Each pair has a + and a - pin. as that surgests one wil have a +ve voltage and the other a -ve voltage. What I am surgesting is that you tap off the Tx +ve pin and if the value of that pin is high then turn the LED on and if it is low turn it off.

However as the Tx pin is a data pin you have to prevent data being lost when you tap off it so you will have to use a buffer to prevent the value of the Tx pin being changed too much by you reading off it.

So basically what you will need to do is connect a buffer to the Tx+ pin and then use the output of that to power your LED.
I dont own an SATA drive so I cannot help that much however.

But the info I have posted here should get you most of the way.

 

ok thanx ill give it a go lol ill need to get the parts first and another sata cable maybe 3 cause ill screw it up a few times lol

 

You could use an op-amp for this, they usualy have a fairly high input inpedance. Just for kicks you could throw a 1Megohm resistor in series with it.

 

IMO this is entirely the wrong way to go about this.

What I would first do in your place is look into your SATA controller chip. All of the ones that I have looked at have had individual activity pins for the individual hard drives. This doesn't mean that every controller has them - but all those that I've checked do have them. After that - you'd just add an optoisolator to the activity pin and you'll be golden. Hope this helps.

 

IMHO voiding your mobo warenty and soldering onto an SMT device is the wrong way to go about it. Anyway all that is required to be connected to the SATA cable is an OPAMP which most likly can drive the LED directly.

Lynx

 

I don't agree. The data lines on SATA cables are not designed for that - and that could cause unwanted interference and signal degredation. And I forget what frequency SATA operates at - but I have a feeling that it will be rather fast - so you'll need additional circuitry to leave the LED on for a visible amount of time.

 

its actually because how fast the data is transmitted that you dont need any extra circuity. As long as any data is flowing it should work. Any signal degredation should be avoided by using a high impeadance device such as an Op AMP. SATA is very good at noise reduction as it is a paired design to reduce noise.

 

Actually as SATA is LVDS my way wont work. Sorry

 

Soldering to the controller chip sounds easy enough. Every controller chip I've seen is at worst in TSOP packaging, which a steady hand can handle. Just use 24 AWG wire or thinner. Finding the pin assignments should not be hard; I've found adequate documentation on all of the popular Serial ATA chips.

 

just because its LVDs dosen't mean it won't work, all you need is a very high Z sensor.

But tbh this is way too much hassel as its a high speed serial (VERY high speed) you have to worry about sheilding, then not to mention the change of impedance that even the highest Z sensor would bring. (A researched put a scope across his one, only to corrupt shedloads of data).

Myself i would look at the drive, a LOT of high end HDDs have access indicators on them that you could potentially tap off. However with SATA this might be one of the cost cutting things that only SCSI users have.

(a lot of SATA drives are taken from the SCSI 10krpm version, often the platter size increased to make it cheaper, at the expense of seek time).

Also guys, play nice, both points have merit, but also have drawbacks, keep this objective rather than school yard stuff.

 

My way wont work as the Tx+ will not nessasarily be at 0V when there is no data goign through it. As the data is the difference bettween Tx+ and Tx- it could be any value in a given range. Hence my way wont work. Tapping off the mobo# controller and voiding your warrenty is the best way.

 

you could mesure it of the T line (the pair constitute one signal).

For intelectual reasons i could scetch up a circuit, but see above why i wouldnt recomend it.

 

Odd, I don't remember intel southbridges having a TSOP package. Some sort of mBGA I think. Or is there a seperate chip for intel southbridges? Easy enough solution for any other controller chip though, if it's really worth that much to you.

That's why I like my highpoint controller. It has LED headers (and drive present headers) right on the card Unfortunately putting a PCIX card in a PCI slot is bottlenecking it so bad that RAID0 raptors are performing about the same as 1 7200rpm sata.

 

hmm, my searches are proving futile, byut a little while ago, Zapwizard did post a how-to on adding SATA activity LEDs from the controller chips.

I have some thoughts on how you might be able to do it, but I agree that they could potentally be very bad ideas. I'm not sure what degree of slop SATA will tolerate, but I'll mention the ideas anyway.
1)opto-isolators: You'd need a pair each line wired in an 'opposing pair' (think bi-color LED) so current can continue to pass normally, but I don't know what sort of currents SATA provides. Plus there is the added delay from the diodes recieving enough voltage to turn on.
2)op-amps: If you had op-amps that behaved as close to ideal as possible (infinite input impedance), and made only small incisions in the SATA lines to preserve the shielding, you might be able to do it. An ideal op-amp won't draw any current so any parallel device (like a controller and HD) won't even see it.
3)inductive pickups: These would have to be immensly sensitive so you ony needed a loop or two. While the wires wouldn'r need to be cut, their shielding would need to be removed. Making sure you had pure signal from the cable could also be a problem. My limited experience with any system such as this has been with oscilloscope current probes, so it's really limited.