Alrighty.

So I'm using the Autoleads PC99-X46 Lead in my Mazda with the PC99-SON Sony adapter. Thats hooked up to my PCATS tuner and Empeg running firmware v2.0 and Hijack v479. The steering wheel on my car has buttons Up, Down, Menu, Mute, Vol+ and Vol-

As standard, the buttons do absolutely nothing.

I added stalk_debug=1 to my hijack config and looked for serial output while pressing buttons. BOTH Up and Down buttons giveand all the other buttons do nothing. Not ideal.

So I broke out a multimeter and tried measuring the resistance on the plug of the Autoleads part when pressing buttons. I've established that all the buttons do 'work' in that they do change restistance values. Its hard to measure the Up, Down, Menu, Mute button values because I have a digital multimeter which doesn't react fast enough when you press the button. The Autoleads adapter only seems to send a pulse on those. Holding down Vol+ and Vol- works great and I can get an accurate reading on them. I end up with something like this:

• Up: ~12.00 (maybe?) kΩ
• Down: ~14.30 kΩ
• Menu: ~11.00 kΩ
• Mute: ~34.00 kΩ
• Vol+: 17.66 kΩ
• Vol-: 21.98 kΩ

Okay, I'm not sure what to do now. By the looks of things, the stalk interface on the tuner should be able to look out for those values, they seem to be in range of what the Sony stalk uses (2.2-48.4kΩ), just off here and there.

Anyone got any ideas?

You can use stalk_lhs and stalk_rhs to modify the values for the stalk.

Does your multimeter have an option where it will hold at the minimum resistance it sees?

Thanks wfaulk, I had a quick look at those options but I don't know where to start since I'm not getting any debug messages in the console from most of the buttons.

My Multimeter has a 'MIN' mode that I was using which holds the minimum value it sees. I was using that, but I still think it doesn't have fast enough reactions to give a reading that I have confidence in. Every time I press the button I would see its little fast reacting LCD bar move but it would only give me a readout around one in ten presses. The Vol+ and Vol- readings are dead on though, every time they give exactly that reading (no need for the minimum function, just hold the button).

Oh yeah I tried my sisters PCATS tuner (just looked for button codes on the screen, no console), but no joy with that one either. Both tuners work perfectly for Radio.

Here's the thread where I got my steering wheel remote to work with a PCATS tuner and Hijack. I had to get Hugo to replace the resistor in the tuner to bring it into the correct range, but there might be something useful in there...

European Ford Steering Wheel Remote works!!!!

Thanks Roger, that does look handy. Now, even though my stalk (at least the Vol+ and Vol- buttons that I'm sure about) appears to be in the correct range, Vol+ is 0.86 kΩ too high and Vol- is 1.62 kΩ too low, but both SHOULD give me something on the serial console when Hijack has stalk_debug=1.

The interesting thing here is that my Up and Down buttons which have AT MOST 12 kΩ give me FF on the serial console, so it does point to something being askew on the Tuner which is making it 'top out'.

I am going to try bridging the 22 kΩ resistor R2 in the tuner. It sounds like the right thing to do.

I'd appreciate it if anyone lets me know if they think I'm going about this the wrong way or I should try something else first - I'm really no expert on this kind of stuff.

Oh, I guess I can unscrew the connector on the tuner stalk plug and bridge the two contacts with some tweesers while watching the serial port. That should give me something between 00 and 07 in theory (and turn the player off).

If it works I will be confused. If it gives me a serial message of around 80 then I think my resistor mod should work.

Silly question, but is the stalk disabled in config.ini ?

I have stalk_side=rhs under the [controls] section of my config.ini - thats it, right?!

I tried 'lhs' and 'none' too just for good measure while I was in the car.

Yup, should be fine.

Oh wait, actually its not stalk_side=rhs, its stalk_side=right. It should work though, it was set using Emplode. I've attached my config.ini anyway just in case I have anything weird going on. Will suppress_notify=1 suppress the debug messages? Is stalk_debug and trace_tuner the same wrt to the stalk debug output? Mark, is it possible to have the stalk debug output to the screen? Either on the 'view button codes' or a new 'view stalk codes' section of Hijack. Taking my laptop and dodgy USB-Serial cable over to the car is a lot of faffing around.

Well, the saga continues. I was in the car late last night taking things apart (again) and found some important facts. The Up and Down arrows on the wheel gave me "Stalk: in=02 00 ff ff == no button" on the serial port and with my multimeter I found that these are actually shifted buttons (i.e. they show continuity between the middle ring (B) and the ring closest to the plug (A) as well as a resistance value between the tip (C) and A.

On discovering this I got all excited and thought somewhere along the line on the Tuner side C and B have been swapped over - this would explain everything!

I salvaged a 3.5mm stereo plug and lead and socket and lead from a box of stuff and set about stripping the leads back - I twisted the cables together according to my theory... and it still didn't work (no serial output).

Feels like I'm getting closer to getting it working though. Going to pop over to my local friendly car stereo shop in a minute to see whether they have an RM-X2S or RM-X4S lying around that they can lend me. That should help me figure out whats going on.

Nope, they have no Remote Commander to lend me. Going to fiddle around with this some more today hopefully, its very frustrating. I will make it work!!!

"View button codes" should already show the stalk controller resistances on the screen. If it doesn't, they're probably out of range of the comparison resistor.

Oh right... My Up and Down buttons give me that FF message on the console, but not on the 'View button codes' screen (IIRC, will double check in a minute, just going to the car). Something is not right.

Definitely don't get any output at all from View button codes in Hijack right now.

My own config.ini file has only this for the stalk:

[controls]
stalk_side=right

There are no trace or debug options enabled whatsoever.

Mmm.. something's wrong there. They must have a capacitor or something (internal to the stalk) that's in series with the output or something. That has to be fixed before it will work with the empeg tuner inputs.




Okay, so those two should be usable, if they're within range. My real stalk reads 23.6K and 17.5K for those two, give or take a few percent.

The connector has three conductors, which I will call tip, ring, and collar, in order from the pointy end of the plug back towards the plastic housing and wire leads.

The tip is for the buttons, ring is for the "shift" function, and the collar is common to all.

Cheers

Thanks Mark!

I'm going to take out my debug lines, give up on the laptop and rely on the Hijack display. That way I know it should work and there are less things to go wrong.

In theory I should be able to do a simple test with the tuner stalk socket. If I unscrew the cover on the connector and short out the cable connection points for the collar and tip I'll see the Hijack display show the button code Stalk=0x01 and/or turn off the player... right?

Right, except the numbers will look like this instead:

Cool!

I'm going to go to the car and try that now. It would be a good start to get the basics working.

Yeahh... that didn't work.

I've taken my tuner out of the car and having a close look at it now.

Perhaps you've simply been trying with the microphone input, rather than the stalk input? The connectors appear rather similar.

EDIT: I just now hunted down the dock here with a PCATS tuner inside, and tried the same tip-to-collar test there. FAILED. Then I tried what I thought was the MIC input, and.. behold.. it lives!

That thin grey coax of the stalk input *really* suggests a MIC input to me!

Cheers

Yes, the thin grey coax cable from the tuner... thats not the stalk input?

But but, I thought the Mic input was on the sled? Surely the stalk input is on the tuner...

Mark, the connector that works for you, can you please trace the cables and see where its connected up - does it go to the sled? This is weird!

So far I've only been interested in the thin grey cable from the tuner. Mark, I'd really appreciate it if you could please double check whats happening with your PCATS tuner and whats connected where.

I'm going to bed now, thanks for all you help so far, back in a few hours!

No no.. don't let me confuse you. The thin grey thing from the PCATS tuner *is* the stalk wire.

The sled harness connector is indeed the MIC.

Cheers

Okay, I'm not going crazy

Right, well I've been out in the car again just to try that test once more to be sure. Shorting out the collar and tip of the stalk remote connector does not give me any output at all on the Button Codes Display or serial console with the debug options enabled. I've checked the wiring of the stalk remote line, and its okay. The cable is not broken either, all the traces are fine and its all connected the right way with no crossed wires as per the tuner assembly guide PDF.

With trace_tuner and stalk_debug both enabled on my player, shorting the cables for the ring and the collar (Shift) gives me the following on the console:

It gives me nothing on the Hijack Button Codes display.

I've attached the dmesg output from my Empeg in the car, but don't see anything broken in there really.

Whats my next step? If I can't get this basic test working, is there definitely something up with my Tuner? Could I send it over to someone to have a look at? Patrick?

Hi.

With reference to the attached schematic, I would proceed by first connecting the stalk and measuring the voltage at pin 2 of the PIC, IC4, as the controls were manipulated. If you don't get any voltage there, or it doesn't change, check the output of IC3 pin 1, the LM258. Also check it actually IS a 258

If there still isn't anything, check the input on pin 3 of IC3. If there's STILL nothing, your stalk or wiring is bad.

If you get a voltage at IC4 pin 2 that looks reasonable, it's possible that the PIC is faulty or damaged. D2 should protect the input from spikes, but it's not infallible. If this turns out to be the case let me know and I can swap it for another one.

Pca

That one has a "bad checksum", so Hijack ignored it.


Yeah, 0xff is the *only* stalk report that Hijack does NOT show on the front panel, because it means "button released"

So what you got was a 0x00 "button pressed (but bad checksum)", followed by a "button released".

Mmmm.. I wonder if it's just the checksumming logic that's busted on that tuner?

Here's a copy of Hijack that disregards the checksums:

http://rtr.ca/hj479+no_csum.zImage

Try it and report back.

It could be quite informative to see a full serial port trace (with the stalk/tuner debug/trace options you were using above) showing what happens with the stalk connected, and the vol up/down buttons being used.

Cheers

Mmm.. a correction on that. I misread the packets. The checksum is likely passing just fine after all. The button value isn't 0x00, but rather 0xff on both packets. So it really is invalid, and quite likely a circuit issue in the tuner.

I wonder if the in-car stalk somehow damaged the tuner's stalk input? We already know there's more in there than just a resistor array (because most buttons generate a pulse, rather than just a continuous level).

??

Thanks Patrick!

I really appreciate your help, I know you're a busy guy. I will try out your steps in the car later hopefully.

While my tuner was right here I had a quick look at it and IC3 is not an LM258, its a KA258.

They are both dual op-amps, are they both compatible/interchangable?

When I tried this in the car with my Autoleads adapter connected all the buttons produced no output except Up and Down, which gave me the same output I get when bridging the 'Shift' pair (collar and ring) as these are shifted buttons.

Oh, I'd really hate that!

I'll check the output of the stalk adapter for voltages (there should be none, right?).

We have 5.03V here. Fiddling with the stalk (tip to collar trick) doesn't change anything.


Again, 5.0V on idle. Pressing a button shows around 0.5V briefly (< 0.5 sec) and then 0.89V while the button is held down. As soon as the button is let go, its back to 5.0V

You're lucky the police didn't come around and nick you for hot wiring a car

Yeah should be compatible enough. Nat Semi (the LM258) initiated a lot of designs that were later copied functionality wise by others using the same numbers but a different prefix to prevent any trademark issues I assume. Happens a lot with common parts.

Whilst a quick glance shows they are not exactly equal, I think you should be OK. Patrick might know more though and may specifically say the Nat Semi LM258 is the one to use.

Edit: your photo reminds me of when I was hacking up some code to drive my Alpine CD changer. I was editing on my Windows laptop to the local disk shared via wireless back to my desktop running a VM with Linux doing to the compiling putting the object back on that file share for download to the empeg...

I've been talking to Patrick by PM and thought I'd just copy into this thread since its still ongoing and as yet unsolved.

I basically asked about a possible PIC problem, and said I was going to check my sisters tuner and get back to him, then:

Its alive!

After the tip to collar trick failing I was pretty certain my tuner was dodgy. I confirmed this further when the Sony Stalk I picked up on eBay didn't work either. So, I borrowed an original Empeg tuner. Plugged it in, and it all works straight away. Just like that.

So, turns out that the 'stalk' part of my PCATS tuner and my sisters PCATS tuner is faulty somehow.

Btw: does my PCATS tuner really give me better reception than my original tuner, or is it the case that the signal meter on the original one actually works as opposed to the PCATS one being stuck on max?

Very interesting. I had pretty much the same problem as you and I put it all down to my eBay Autoleads adapter (I have a PCATS tuner too).

I had to use Hijack's IR_Translate to get things working. The Button Codes display was a great help in determining the values to use in IR_Translate. After a few days, some of the button's actions became erratic and I found that some of the codes returned by the buttons had drifted a little. That's how I left things 18 months ago and, inspired by the notion that my Autoleads adapter isn't necessarily faulty, I will have another attempt at getting it working soon.

This is how my IR_Translate ended up:





I find that anything less than a full strength signal display coincides with poor reception.

Going slightly OT, another problem I have had since building/installing my tuner is that the AF (Alternate Frequency) mode keeps re-tuning from a good signal to weak/marginal signal (3.00a11). This happens a few times/minute so I keep AF turned off which is a pain when driving long distances.

I need to dig my tuner out from the depths of my dash and re-check the coax lead's grounding - I might not have bound the earth connection as well as I should have.
I also have a spare (unbuilt) tuner kit and this thread has enthused me enough to get it assembled so I can swap it for the one I'm using now. Thanks for the 'prod'!

Maybe your problem is not exactly the same as my problem. The stalk controls connected to my PCATS tuner gave no output to the Hijack button codes display.

But yes, although I had suspected my Autoleads adapter to be broken for a while it turns out that it works rather well. Its fast and seems pretty reliable. I'd recommend it.

For the record, the stalk codes I get from it are:

They make more sense when stalk_side=left and then still need a little ir_translating to make it really useful.