Jump to content
SAU Community

Recommended Posts

Hey Jeff,

Ur a legend mate, I followed your instructions and got my unit to work momentarily. Here is what happened:

1. Measured the voltage, one of the 3 points measured 1.7 volts.

2. Adjusted it as high as possible to 2.2 volts

3. Turn the car on, 4wd and abs light remained on.

4. Turned the car off and adjusted my after market torque splitter higher (more front wheel bias setting). I'm guessing this increases the voltage too??

5. 4wd and abs light went off and all was working well.

The unfortunate thing is that it stopped working this morning :( any idea what I néed to replace on the board to get it to 2.5 volts and remain there??

Ps good on u for taking the time to trial build a new unit mate. I think there r a lot of ppl who desperately want to fix their gtrs. After all driving an expensive 1500kg gts-t is not much fun is it :P

Hi Jeff, good to hear that you are looking at new ideas, look forward to seeing how things progress. Would love to get my unit working again as the car has been in 2wd for a while now.

BTW mine too does not go above 1.6v, I think one of the voltage pots is buggered up.

Jeff,

i reckon if you build a working new unit, you could retire off it lol, theres that many GTRs limping in 2wd!

update on mine - works if I start the front of the car facing up the hill. haha, better than nothing i guess!

Did a tutorial on actually getting to the g-sensor the other day. Here's the link if anyone is interested..

http://www.skylinesaustralia.com/forums/topic/376142-r32-gt-r-lateral-g-sensor-check-and-more/

Without seeing whats on the other side of the PCB, the underneath pictured is very simple - three quad op-amp chips taking input from what I'm assuming are the G sensors on the other side of the board.

Surely someone has a crook one here in Melbourne that wants it fixed....

Hey Jeff,

Ur a legend mate, I followed your instructions and got my unit to work momentarily. Here is what happened:

1. Measured the voltage, one of the 3 points measured 1.7 volts.

2. Adjusted it as high as possible to 2.2 volts

3. Turn the car on, 4wd and abs light remained on.

4. Turned the car off and adjusted my after market torque splitter higher (more front wheel bias setting). I'm guessing this increases the voltage too??

5. 4wd and abs light went off and all was working well.

The unfortunate thing is that it stopped working this morning :( any idea what I néed to replace on the board to get it to 2.5 volts and remain there??

Ps good on u for taking the time to trial build a new unit mate. I think there r a lot of ppl who desperately want to fix their gtrs. After all driving an expensive 1500kg gts-t is not much fun is it :P

Hey mate. Not sure why it would start then stop working unless the pot moved just a tiny bit and flipped the bias voltage low. Might have happened as you set it really close to max... So it would only take a tiny bump...

Awesome to have Bozz onboard. The accelerometer is actually a really cool old school design. Is uses a magnetic pendulum suspended in about 30000~40000 weight silicon lube (for dampening) and a hall effect sensor to measure its position. Not sure witch components drift over the years... I'm guessing the hall effects. But they are moulded into the housing. I think (if I remember correctly) they there were the only components on the back side of the PCB. But when I pull it out again. (Hopefully this weekend) ill check. I was more interested in going digital so I could fiddle with settings on the fly etc, I kinda didn't look into the root cause of why the sensors drift and die. Does anyone have a broken unit that maybe they could donate/lend to Bozz or myself?

If you need a better quality image Booz of the top drop me an pm with your email.... There were also some resistors on a little side PCB between the output of the top PCB and the header... Didn't look like it did much so I didn't take a pic. Originally the pic was just for myself so I knew what it looked like before I fiddled with components.

I think the op-amps have gone are EOL but there might be a drop in ??? (haven't looked...)

Awesome to have somebody else gett'n into this.

Just landed back in Adelaide yesterday.... so I should be able to take a look at stuff as needed...

Jeff

Sent from my HTC

I'm in two minds about a digital accelerometer - the R32 4WD system is well known to be shit on a track and all the aftermarket stuff does is feeds different signals into a slow and shitty controller to try make it less shit in some situations.

An awesome project would be to replace the 4WD control system (leave the factory ECU doing ABS) with a MEMS accelerometer such as the ADXL335 coupled to an Arduino Duemilanove. $40 gets you the new sensor, new controller running at 16MHz and you can make it do whatever you want, track, drift, playstation :)

But back on topic, it would be good to play with a broken sensor as it's unlikely all three axis are broken so I could spin it up and plot the force/voltage and pretty easily make a new unit with a MEMS accelerometer. After the design work, I reckon a replacement adjustable accelerometer could be manufactured for under $100.....

Sent from my laptop using Firefox

I'm in two minds about a digital accelerometer - the R32 4WD system is well known to be shit on a track and all the aftermarket stuff does is feeds different signals into a slow and shitty controller to try make it less shit in some situations.

An awesome project would be to replace the 4WD control system (leave the factory ECU doing ABS) with a MEMS accelerometer such as the ADXL335 coupled to an Arduino Duemilanove. $40 gets you the new sensor, new controller running at 16MHz and you can make it do whatever you want, track, drift, playstation :)

But back on topic, it would be good to play with a broken sensor as it's unlikely all three axis are broken so I could spin it up and plot the force/voltage and pretty easily make a new unit with a MEMS accelerometer. After the design work, I reckon a replacement adjustable accelerometer could be manufactured for under $100.....

Sent from my laptop using Firefox

Yeah I totally get what you mean. I'd love to have the time to totally replace the controller with something better, but as i'm always so busy with work I don't want to start something I'll never get around to finishing. I agree, def would be the better choice... I'd love to have the source code of the current controller, just looking at the maths would be sweet. I think to start with i'm gonna stick with just a digital replacement for the G-sensor. But, if you wanna have a crack at replacing the controller I'd sure be interesting in lending a hand (if you want it).

My bro is also in Melbourne (sort of) and has a R32-GTR i'm sure avail for testing etc.

Just started the schematic etc for the digital G-sensor. I use the latest Altium Designer but can release it in whatever file type ppl want when it's done. (Maybe just the Gerbers so ppl can order the PCB's etc ??) The unit i'm building is a little overkill as I'm also going to use it for another project, but yeah I reckon around too that about $100 would be about right for a cut down unit just to replace the G-sensors.

Jeff

  • 5 weeks later...

Almost ready to send the PCB away... Just wondering does anybody have an R32 G-Sensor (a really good one) they I could borrow for some testing ? I think mine is ok, but i've done some tweaking so i'm a little unsure how "original" the outputs now are. I'll be at SAU Nationals (Just the Race Day - spectating) I can post it back after i'm finished with it... I don't plan on opening it up or anything so nothing *should* get damaged - I just want to monitor the output against my new replacement when it's up and running.

Cheers,

Jeff

My mate has a 32 with a cracked g sensor, its probably had a large shunt. Can this be repaired? Otherwise he will be interested in your board design if it works as planned.

If the G-Shock indicator has gone red then it's probably still fine (or might just need the bias voltages adjusted as per my pic at the start of this thread) If the whole unit has cracked and the silicon dampening fluid has leaked out of the hall effect / magnet pendulum cavity then it's prolly NFG. I'd take the top off and adjust the bias voltages first. If it's been hit the worst thing that can happen is that the magnets mounted on the small pendulum spring will be bent a little (ok) or bent a lot (nfg unit) inside it looks like this

Just pretend the magnet is a rock mounted on a piece of string. (Really it's a magnet moulded in a lead weight mounted on a spring all sitting in some thick fluid.)

"x"'s = Fluid

----------------------------------------------- (PCB)

xxxxxxxx|xxxxxxxx

xxxxxxxx|xxxxxxxx

xxxxxxxx|xxxxxxxx

xxxxxxxx|xxxxxxxx

xxxxxxxx|xxxxxxxx

xxx(Magnet)xxxx |Hall effect sensor|

As the car undergoes g-force the magnet moves closer or further away from the sensor.

After years of movement or an impact they look like this (Dramatised) (Edit - this should look bent to the right - added x's so the alignment works)

----------------------------------------------- (PCB)

xxxxxxxx|xxxxxxxx

xxxxxxxxx|xxxxxxx

xxxxxxxxxx|xxxxxx

xxxxxxxxxxx|xxxxx

xxxxxxxxxxxx|xxxx

xxxxxxxx(Magnet) |Hall effect sensor|

And therefore the output of the op-amps feed from the sensor needs to be re-calibrated, the bend in the spring taken out, or the unit replaced.

Hope this helps a bit and that the text pics look ok

Jeff

Edited by datsun_1600
  • 1 month later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



  • Similar Content

  • Latest Posts

    • The fab work can be as simple as a couple of silicon hoses and clamps to the factory piping. 
    • Just sounds like either way you need to do some fab work to get everything to fit, so why limit yourself at that point? If the GCG high flow option is zero effort in and out swap though I'd probably do that. It's almost certainly lowest risk, lowest cost, etc. The HKS GTIII-RS option that Kapr mentioned is laughably expensive for what it is, they charge the exact same for two turbos on the RB26 so their margins are off the charts on that thing.
    • Intake manifold is not a part of the issue. The turbo bolts to the exhaust manifold. That is easy. But close your eyes and picture the physical situation. That is a T3 flange on the manifold and a T3 flange on the turbo. As long as any new turbo has a T3 flange on the exhaust housing, that exhaust housing will bolt to the exhaust manifold. This places the exhaust housing in the same place as the stock one. But now move your mental attention a little further forward. The location of the compressor housing is set by the length of the turbo's core. The stock turbo had a long core. Let's say that it is..... 100mm long. So that would place the compressor housing 100mm forward of the exhaust housing. But a highflow, might well have a centre core that is shorter. Let's say that it is only 70mm long. Now the compressor housing will be 30mm further back in the engine bay than the stock one. This DOES move the turbo's compressor outlet backwards. It also moves the compressor's inlet backwards. You will very likely have to do some work to both the inlet and outlet piping to make everything connect again. I am not say this to make it out to be a bigger deal than it is. I am just pointing out that "bolt on" is sometimes not quite bolt on. The highflow from GCG that Murray linked above (https://gcg.com.au/turbo-charger-upgrade-skyline-gtst-2iu-xtrgts-s1.html ) uses a core that is the same length as the stock core, and so does not require this extra work. It will look very much like the stock turbo. No-one uses GTR turbos of any flavour (stock, or aftermarket) in a single turbo application on RB20/25. It's not a thing. Yes, people have been putting on GT3076, GTX3076 (and bigger and smaller versions of those) and G30s (of various sizes) onto RB20/25 since forever. But these are not "bolt on". Everything except the 4 bolts to the exhaust manifold change with these. And genuine Garretts are expensive. Non-gen, like Pulsar, etc, are cheaper, variously as good or nearly as good. But still not bolt on. No-one in the right mind would pay for an HKS turbo. Not in this modern day and age. Well, yes, the GCG highflow. You could ask HG what he can do to make the compressor housing sit in the original location. There are surely others doing highflows around the world. And some of the eBay/Temu ones (as reported by Dose) work and don't die. Bit of a lottery though. I would send your turbo to GCG (here in Oz) to be highflowed if you want a trivial no-extra-work option. But seriously, the work required to change inlet and outlet piping is not that hard. That's a boost control problem, not a turbo problem.
    • Thank you all for the replys 🙂 I know that intake would be different but that is one pipe at it is not that hard to get(custom one). I meant mainly bolt to the stock manifold and the turbo elbow. I looked and many sites/forums but they are just "old" with some old HKS turbos from GT-R i guess? What about some Garrets?  Or any other turbo? Is there even a turbo which i can just bolt on? 😄 And yeah i know about that new HKS but that is like 2000k USD without taxes/shipping in here   Iam getting a touch up tune but my "problem" is that on the "not" hot day iam getting peaks around 0,9 bar...and when it was around 15 Celsious i saw peak around 1 bar which is just too much for stock turbo. And of course turbo is old and i like to get some new one for a piece of mind 🙂 
    • I'm working on the assumption that our friend Jasmine here is a Russian (or, possibly Ukrainian) spammer/spambot, based purely on the number of such that I have been having to neuter in the last few weeks. IP address for the OP above was in WA. But that could have been via VPN. Posting at quarter to 4 in the morning is a good sign of being from somewhere in Europe. The last Jasmine that I kicked in the cooch was IP addressed in Ukraine. Even that could have been via VPN, and the bitchbot could have been from Russia, Serbia, China or anywhere. Regardless, was a spambot, so I killed it with fire. The fact that our new friend Jasmine here did not respond in any way to my tart query strongly suggests to me that this OP was just the establishment phase of a user able to be activated for spamming in a week, or 3 or 10.
×
×
  • Create New...