Electronics ATi AIW 9800 PRO problems - Power Delay Relay Circuit needed

Hi

I have recently purchased a ATi AIW 9800 PRO for my Jetway MiniQ 760.

Well, when I switch the PC on, the grfx card spins for half a second and then nothing... =(

After searching a few forums, it seemed to be the same power overcurrent issue many AIW users suffer from. So, I have tried the manual 2s 5v power cable plug in workaround, and it works perfectly. I have decided not to RMA the card or PC just yet, instead I want to build that simple Time Delay Relay circuit.

If anybody can point me to good threads, diagrams, kewl images of their Radeon power delay circuit design, a step-by-step article, etc, it will be much appreciated. =)

I am a n00b when it comes to electronics, so I would need a little more help than a circuit diagram to get started.

Thanking you in advance!
fuekewl

 

This site holds many circuits to what you need: Discover Circuits

For an easy curcuit, use this except invert the output via a 2 state-2 output relay or transistor inversion. Time Delay Relay

Hope this helps; if you have any questions feel free to ask.

 

Thanks Xiachunyi, that link gave me some ideas to simplify the circuit I have designed to solve this problem.

The proposed solution circuit and a longer thread on this issue is in the graphics cards section here: (The same question post was put in both forums)

http://forums.bit-tech.net/showthread.php?t=50572

UPDATE: (Sept 04) The original proposed circuit diagram is re-posted below, since I cannot alter fuekewl2's posts:

​

This is the schematic referred to in the post below. (the red 'x')

 

My apologies for the dual thread in the Graphics card and Electronics forums, This forum is probably the best for this thread, and should be used henceforth.

Below is my response to "Thr Bodger"'s post, from the "Graphics Card Forum".

=====

Bodger, You've been busy! =)

That's awesome! =)

As long as I dont have to do the 2second workaround everytime I switch on the PC, its perfect! =D

Thanks for taking your time out to simulate your circuit you've designed!

WHen you have some time in the week, could you also draw up a final components/parts list please? I'm such a n00b at all of this, its embarassing =(

How do I do that, is there a setting? Do I have to ask a moderator? Or can we just continue this thread in the other Electronics forum?

I'll reply to the other forum's thread now!

Thanks again
fuekewl

 

Thanks for moving the info in the other thread across to here; I think this project is better suited to the 'electronics' forum. As I mentioned in the other thread, if anyone has a simpler circuit, or suggestions for improving mine, feel free to comment.

As it happens, I actually have a use for a fairly similar circuit myself, so I intend to construct a completely working unit and test it. Therefore there should be no problem in sending you a parts list and a 'stripboard' layout for the circuit.

As a rough price guide, I am pretty confident that the circuit will cost under £5, and if you live in the UK, you will be pleased to know that all the parts I am using are available from Maplins and from Farnell.

I've have started making a prototype on 'breadboard' now, all I need are the TIP120 darlington pair transistor and the zener diode. (Will raid the Uni components bins on Monday). Therefore I should hopefully have some news by Tuesday at the latest.

 

Just a quick update, to show that I haven't forgotten! I have now got the zener diode, but still don't have the TIP120 transistor, as Uni were out of stock. Will be going to Maplins tomorrow to buy one.

 

Update!

I've got the transistor today - couldn't get a TIP120, but got TIP122 which does exactly the same job. Circuit appears to work fine - just need to change the value of VR1 (the variable resistor) to make the time delay easier to fine tune. I'll have photos of the prototype up by this time tomorrow.

 

Hi again, yet another update!

I've tested the circuit, and got it working 'about right'. Note that I've changed afew component values since the first design:

​

Note that TIP120 also works in the circuit.
By adjusting the setting of VR1, delays between 0 and about 8 seconds can be achieved. Time delay starts when circuit first receives power. The maximum current that can be drawn through the circuit is 5Amps at 12V and 5Amps at 5V. (60Watts @ 12V and 25Watts @ 5V, simultaneously). Note this will only be the case for the final soldered version at the bottom of the page; prototyping boards cannot cope with this much current.

Below is a photo of the working prototype. (On prototyping board):

​

Power comes in from the left hand side of the photo, (Yellow = 12V, Red = 5V, Black = 0V) and exits at the bottom right.

The relay used is a Double Pole Double Throw (DPDT) type, with a maximum rating of 5Amps. It uses a 12V coil, with 300 ohms resistance. Maplin order code for the relay used is QC95D. (Will put full order code list up when unit soldered together) Note that all the components for this circuit can be bought at Maplins.

Finally, if someone wants to push ahead and build a complete circuit on veroboard before I have, below is a stripboard version, as viewed from the component side:

​

Note that this final layout HAS NOT BEEN TESTED YET - THERE MAY BE ERRORS IN THE DESIGN!! Red crosses mark the points where tracks are broken. Note also the R2 is vertically mounted, to save space. The final board would measure about 5cm by 6cm.

List of component Values:
R1 = 1k ohm
R2 = 5.1k ohm (anywhere between 5k and 10k will do)
VR1 = 22k ohm
TR1 = TIP 122 (TIP120 will do)
D1 = 1N4001 (1N4003 will do)
Z1 = 6V8 (6.8V) zener diode.
C1 = 220 microfarads, (16Volts minimum working voltage)

All resistors and diodes are .25Watt minimum rating. (This is standard)

Relay: See above for info.

Thats it for now. If anyone spots an error in the stripboard circuit, please let me know - I'm building it on Sunday!

 

Good news for some (not for me though) overclockers.co.uk have in the shuttle silentx 250w psu that is not affected by this issue. This is a fairly common problem for shuttle users so hopefully this should help a few of those ppl if they arent up for a bit of soldering, i myself was going to go for this option but unfortunately this psu doesnt fit my sn85g4 so check the shuttle site before you buy. how daft was i, thinkin a shuttle psu would fit in a shuttle system. i now realise my mistake

SIDENOTE: didnt fit due to the silentx being larger height wise than the sn85g4 standard psu putting the position of the mains connector too low for the hole in the back of the case it also made fitting memory a little tight, however this could probably be made to work with the assistance of a dremmel ( or other such rotary tool ) but i've only had this a week and im far too new to modding to risk this quite yet

 

Not to worry, I haven't gone away. Better still, I'm going to build a soldered unit tomorrow, and will photo - document the build for you so that it can be repeated.

 

I posted a fairly similar delay circuit at Phaeton forum a few weeks ago, application was to run a fan at full speed for a few seconds if the controller was turned low, but the relay switch can drive anything. Uses a MOSFET to give a higher trigger voltage than a BJT but the zener idea is neat . Making R1 a pot will vary the delay.

 

Hi all,

I've completed the soldered version (as shown in plan above) and it definitley works Have tested it successfully using a PIII computer and an old hard drive as the load. Haven't got time to photograph it and post it up now, will do so later.

 

Works great for me too, built it to match the circuit in post 8 but on veroboard. i also used a 47k vr as i didnt have anything else handy

 

Hi all,
This will be my final post in this thread (unless anyone has any questions about the circuit for me) Below are some photos of the final circuit on veroboard (stripboard). Note that the hand drawn circuit shown in post #8 works fine, and was the blueprint for the unit below. The only reason that the two are slightly different is that I cannot count holes properly!

​

As previously stated, all the parts used to create the circuit can be obtained from either Maplins of Farnells. Component list is given in Post #8.

To n3vrmind, glad that I could help. If any others use this circuit, please let me know - I'm interested as to what other applications it may have! (I'm putting one in my car, to allow me to activate two circuits from one switch, with a 3 second delay between them powering up)

 

That really is AWESOME Bodger!

I will attempt to get all the parts this weekend and start building ASAP!

If I get stuck ... (probably will) ... Ill give a shout here, hope you won't mind ...

THANKS ALOT for your help, you truly truly rock!

 

Even simpler:



D1 = 1N4001
C1 = 3300 uF
R1 = 390 Ohms
R2 = 2200 Ohms

It stays closed for about 10 seconds.
It automaticaly resets after about 30 seconds of power off.

 

Wouldn't you want a de-bouncer type circuit on the output of that relay to smooth out the power a bit? Have you checked the output on an oscilliscope?

I'm curious as to how long it takes the voltage to stabilize after the switch, and whether it should be regulated a little.

 

The relay has a shorting diode to stop any current feedback when turned off.
The relay would actualy not see any abrupt drop in power anyways, as the current across the relay will drop as the capacitor in charging, eventualy lowering to only a trickling current to keep the capacitor charged.
(But not enough current to keep the relay closed)

As for the switched contacts on the relay, they need no de-bouncing as they are the same as if you pressed and held the reset button yourself.
(Your motherboard will already have it's own de-bouncing circuitry)

 

Delay Circuit request

If anyone would be willing to make a circuit for a reasonable price, please email me. I have an SN85G4 and a Radeon 9800 Pro AIW.

Thanks,
-s-

 

I am prepared to help if you're still interested. (Could make a copy of the circuit that I described earlier in this thread).

However, before anything is done, you would need to confirm that the problem is the same as fuekewl2's. This can be done as follows: (this is how fukewl did it)

1) Power up the PC with the extra power cable to the Radeon unconnected.

2) Before the POST is completed, (before the PC 'beeps') connect the power cable to the gfx card. (while the PC is running)

3) If the PC continues to boot correctly, and runs fine with no stability issues, and can be restarted using the 'restart' button in windows without problems, the problem is the same as fuekewl2's, and the circuit described here will help. Otherwise, I'm afraid you'll have to look elsewhere.

Sorry this is a little late, I didn't notice your post. Note that if one is made, it will be on a swanky PCB, not on prototyping board, too.

Prices will be calculated and posted here if you (or anyone else) are interested. Should be fairly cheap, as in approx. £20. (probably less)