S2Kman,

Let's see if our collective brains can't help you solve some of the problems...

1) There seems to be a real problem with getting a correct program in, possibly due to a noisy power supply line. What sort of voltage reference are you using to drop the voltage? Any filtering of the incoming power line?

2) Crappy sound in the audio range...what crystal frequency are you using? Are you providing any filtering between the DCU and the head (along the link)?

 

I'm not sure that this is DCU related, but my battery went flat last night.

I haven't run the car since early this week and I've been doing some testing with the stereo and DCU but I wouldn't have thought it was enough to drain the battery. It's now on charge and will be monitored for excessive drainage.

 

MacGyver:

Thanks for jumping in here and throwing out some suggestions. I really appreciate it.

I think it makes sense to start off with a re-cap of exactly who has what problems first.

Everyone: The instructions suck. Writing instructions is not my strong point, and it shows. I've updated them and then put them on a web site that no one seems to be able to get to. I am looking into why the web site isn't available (it's some sort of DNS problem), but so far no luck. I sent it out as an attachment to everyone on the group buy today.

Func 10/11 doing func 1/2: Sime, AusS2000, Kaptain Insano, Kacee,

Strange electrical noise: Sime

Erractic behavior at car-start time: Almilli

I am also have an e-mail dialogue going with Kaptain Insano, and he has sent his unit back to me for inspection, testing and repair. 1 person's unit was lost by UPS, and a second unit was lost by a college mail system and I am working on resolving those.

This should be a list of all the problems reported both here on the forums and via e-mail.

For the func 10/11 problem, it appears to me that there is some common problem related to the wiring of the 1/8" jack. I will be able to verify that this is the case when I receive Kaptain Insano's unit in the next couple of days. Once I have verified the problem, I'll post what the fix is, whether I have to send a new harness or what, but we'll know then and I'll post here. I know it isn't a global problem, because I have received e-mails from a couple of people who are using func 10/11 just fine.

To answer your questions MacGyver,

1. I have only had reports of erratic behavior, from Kacee and Kaptain insano, although Almillis problem sounds similar, but not nearly as severe. This really sounds to me like a problem with the unit itself, or something unusual about their cars electrical system. There is filtering of incoming power and the voltage reference is seperate from the digital power supply. I am not 100% certain of the design, because I aid an electronic design engineer to do it for me to make sure it was right. A loose connection in the wiring, or a bad solder joint on the DCU itself could cause this type of problem, although I'm not sure why I didn't see it when I tested the units, but if Kacee couldn't get it to program with the engine running, I recommend that he send the unit and his entire harness including the Sony connections back to me so that I can troubleshoot it. You can send it USPS priority if that is convenient, just make sure it is trackable and insure it for $150 in case they lose it. I'm sorry for the problem, but as soon as I get it I can inspect it and find out what is wrong. Kaptain Insano's is already on the way USPS and I'll be able to troubleshoot it early this week when I get it.

2. On the strange electrical noise, it appears that this problem is isolated since only Sime has reported it. To answer your questions on this issue MacGyver, I am using a 3.6864 Mhz crystal. Although some functions of the unit run at much lower freqs in the 250-500 khz range and are driven by the processor, such as the A-D converter. Since no one else has reported this problem, (And I've never seen it in any of the installations I have personally worked with) it seems to me that this is not DCU related. Physical isolation of the DCU away from the Headunit should be more than enough to prevent any RF noise from entering the radio. The DCU has filtering beads on EVERY line that goes in or out, including the power lines and the 1/8" connector into the Sony, and it makes no direct connections to any audio lines. I've heard scratchy sounds from speakers in the past and it has always turned out to be a bad speaker connection. I can't think of anything else it might be, but I'm open to suggestions.

 

Before I relate my problems, let me first ask a question: can the resistances in the DCU change enough to throw off the programming?

It is cold here (in the thirties now, 20s last night) and I have gotten some very strange results from programming.

I am quite sure I am following the correct procedure, as the lights do all the on/off and flashing, etc that is indicated in the instructions.

Programmed func/ Actual func

mode P: source(3) / source
mode H: OFF / OFF
CH P: preset up (10)/ seek - (2)
CH H: preset dwn(11)/ vol dn (7)
Vup P: trk up (1) /vol up (6)
Vup H: vol up (6) / display (5)
Vdn P: trk dn (2) / vol dn (7)
Vdn H: vol dn (7) / mode (?)

I'll try again and see what happens; as a last resort, I will warm the inside up and see if that makes a difference.

 

S2Kman,

I don't want to pull up old threads (I'm lazy today)...what method are you using to compensate for temp differences? It's too late now for version 1, but the easiest/simplest solution is to PERIODICALLY sample a resistor on the board (make sure it's far away from any power-supply components that would artificially heat it). Since the low-end PICS don't have very many bits for their counters, this can be a short time, like every few seconds, if necessary. This should compensate for the cockpit warming up by the heater on cold days, or cooling down by the AC on those hot days.

You say that you have beads on every incoming and outgoing line (good practice), but what frequencies are you filtering for? What frequencies are the signals going across the DCU to head unit line? If those signals are are in the audio range or above (most likely they are), you cannot adequately filter out low frequency noise by necessity (that would destroy the signals!). So, you have a long line that will be picking-up and radiating some potentially strong signals. I'm not saying this is where the audio crap is coming from, and my first test would be the speaker connections as well, but keep it in mind.

My first bet on the programming issue is noise. Even on battery power, you're above 10V and the PIC is working off of 5V coming from an LM5x-type component (right?), so you have plenty of power, and with the engine running you should be around 14V. Those are pretty good about keeping the voltage stable at 5V, but if you have a lot of noise on the line, it can hop right over that converter and end up in your circuit. I once had the readings from a PIC's ADC jumping wildly around what I thought was a stable voltage...turns out the monitor I was using (a really old and crappy one) was heavily leaking the vertical refresh signal into the serial connection I used for communication. If any noise is getting through your power supply line, it'll have the same effect, and your ADC measurements will be crap. As a test, make a short little program that does nothing but spit the read data out, or blink an LED everytime the data is out of range...plug it in...I'll bet you'll see that LED flash on and off dimly.

While I doubt it's a problem, you should also look at harmonics from the crystal. The 3rd and 5th harmonics can be exceptionally bad if the board layout isn't correct.

 

I have had several reports of problems accessing various functions, especially the #10 and #11 Preset Up/Track Dn functions, and a three people have identified erratic behavior of their unit. I have identified the source of this problem and can now offer a solution. To put it simply, the 1/8" connector that plugs into the Sony stereo is the problem. The connector has three condutors, the tip, the middle, and the base. At the base is a metal ring that is the seating surface. This ring is too large to fit into the flange that Sony has installed to prevent headphones form being plugged into this connecotor and is preventing the connector from going all the way into the Sony stereos. The solution is to grind some of the metal away to make the connector seat all the way into the Sony. I recommend that this be done on ALL of the units, even if yours appears to be working fine. This problem is 100% certain to cause the functions 10 and 11 to not work, possibly can cause other functions to not work properly, and MAY cause erratic behavior of the DCU. Please refrain from using your DCU until this is done.

There are three options to resolve this problem, listed here in order of quickness, if I don't hear from you as to which option you want, you will automatically receive a fix in the mail within 2-3 weeks.

1. The quickest solution is to use a dremel tool or a file to grind the metal away near the base of the connector. If you choose to do this, be VERY careful to not grind the conductors, or to grind too much away. If you are not savvy with electronics (and dremel tools/files, etc...), please don't do this yourself and use one of the two options below. E-mail me at swinn@san.rr.com and tell me that you have succesfully completed this option.

2. The second option is to send me your wiring harness via USPS priority mail and I will repair it and send it back to you. E-mail me at dashcontrols@san.rr.com with your full name, address and the date you shipped the harness to to let me know you have done this. I do not need the DCU, only the wiring harness. I will modify it as soon as I receive it and ship it the next day. This is the second fastest solution.

3. The third option is to do nothing and wait to receive the adapter-fix. I am creating an adapter that will plug into the 1/8" connector on your existing harness, that will then plug into your Sony stereo. This short extension cable will solve the problem, and you won't have to do anything except wait for me to create the adapter and send it to you via USPS. After I finish these adapters, I will be sending them to ANYONE whom I have not heard from that is doing either of the above solutions. So basically if you do nothing, you'll get an adapter fix sent to the same address your DCU was sent to. I'm afraid it's going to take me a week or so to create these things, so this is probably the slowest solution, however it is the least amount of work for you.

I apologize for any incovenience this may have caused. For those of you that would like to know, this 1/8" connector has been a source of MUCH trouble for me. I originally purchased connectors in bulk from an electronics supplier, and during building of the wiring harnesses it was discovered that 75% of the connectors were bad from the factory. I then purchased a second bulk quantity of connectors from a different vendor. We discovered a 35% failure rate of those connectors. I then went to a third supplier and tested the connectors before I bought them and discovered that this supplier had a 2% failure rate. This was the best I could find, so I purchased a bulk quantity of these connectors. These are the connectors that I shipped with, and I tested them with a Sony stereo here at my home. However, since this was not the original connector manufacturer which I designed the DCU to work with, I did not realize that this connector would not fit ALL sony stereos without being modified as outlined above. I hope you can find one of the above solutions to this problem acceptable.

 

To answer MacGyver's questions:

Temperature compensation is done periodically. There are several timers available in the micro I'm using and I went to great lengths to verify that the temperature compensation works fine. I even went so far as to put my S2000 into a commercial freezer warehouse for a couple of days to chill it to -4 F along with several DCUs and Sony stereos. Everything checked out OK in these conditions. I also froze EACH DCU and verified that they would program AND operate properly at -8 F before I shipped them. I also heated them to +160 F to verify that they would program and operate at this temp as well. I am quite certain that there are not temperature related problems.

As far as frequencies and beads are concerned, we are filtering for everything above 2 or 3 hundred kilohertz I think. My design engineer chose those parts, so I'm not certain of the values. Actually, NONE of the signals going in or out of the DCU are in the audio range. There is no on/off switched I/O with the DCU. The input is an analog voltage which varies only when a user presses a dash control. The output is a variable resistance that also varies only when the user presses a dash control. There is no external digital I/O in use. I'm glad you agree that speaker connections are the most likely culprit of crackling sounds at the speakers.

On the programming issues, I am confident that the problem I outlined above is the solution to most of these problems. If the 1/8" connector is not fully seated into the Sony, the commands sent to it may or may not be sent at the proper value since the pins may or may not be connected properly or even to each other. As I stated in the above post (and also in a first group buy e-mail) I recommend that EVERYONE fix this before using thier DCU. I'll be mailing out an adapter that will fix this to everyone in the first group buy, unless they e-mail me and tell me they fixed it on their own.

However, to address your comments on the noise, you are correct, noise has been a problem since the beginning of this project. There are hundreds of lines of code that deal with filtering out signals attributed to noise, the DCU does a lot of processing to determine which A-D samples represent a true 'button press' and which don't. The voltage regulator we are using was tested extensively at voltages ranging from 9V all the way up to 17 V and is very stable. Even in the car with the alternator running it is very stable and noise levels are low (As measured with O-scope).

The way the program in the DCU is designed, it is not possible for the data to be out of range. The range of each button press is 'expanded' so each button has a large 'window' in which the data is valid. Say No button pressed was 4 V, the 'mode' button was detected as 1 V, and the 'CH' button was detected at 2.5 V. If these were the only two buttons programmed, the DCU would 'expand' these values and make windows like: 0-1.75V = Mode. 1.76-3.0V = 'CH' and 3.1-5V = No button. This way the value from the A-D is able to be resolved into a particular button always, even if it is in transistion. This also helps stabilize operation during temperature and environmental changes, because even if the voltages read change a bit, they are still within the allowed 'window'.

For harmonics, you are correct. It seems possible to me that some of the harmonics of the frequency we are using could be causing some odd effects. Those types of problems are why I asked an electronic design engineer to do the final circuit design and PCB layout. I'll ask him about that and see what he thinks. I know he went to some amount of trouble to make sure that the digital and analogue portions of the circuit are isolated and have seperate power fields, etc... so I know thought went into preventing these types of problems. I would think that these types of problems would have shown up in the months of testing I did on the units before I shipped them to anyone.

In short, I am very confident in the design of the circuit and the programming. I think that the few problems we are seeing are related to the above post.

Thanks for your patience folks.

 

Byjove, I think he's got it!

I just got out the trusty Dremel (who could live without a Dremel!) and ground away some of the skirt on the jack plug. It now fits much deeper and firmly and I have to report that the functions 10 and 11 now work as programmed.

Thanks S2Kman. I can imagine this has been fairly stressful for you and you've probably consider joining a monastery but, I for one, am very happy with the result.

 

AusS2000:

I'm really glad to hear that fixed it for you. I'm really upset that this was a problem, I tried really hard to make sure that everything was perfect the first time. Glad to hear it worked for you, aren't dremels great!

Also: I tried to listen to AM on mine and couldn't pick up any stations (people really still listen to this buzzy AM radio?). If you wired your own stereo up, you need to make sure you connect the Sony's dark blue wire to the Honda harness pin 1, the Antenna amplifier. After I hooked up the wire, it was fine. Never encountered a car that had a built in antenna amplifier before.

Well, back to the grindstone (literally), I've got connectors to grind.

 

For anyone that hasn't received them yet here are the updated instructions.

I had a beotch of a time pulling out the dimmer pin and inserting it in position 1 (Antenna power) but managed it by cutting the top of the connector and then folding the plastic back with a jewelers screwdriver. I then took out the four rear speaker wires but this time managed to do it with just the screwdriver and lots of digging.

Whilst I had the stereo out I took the four rear speaker wire from the Sony harness and connected a four point plug to them. I then connected a couple of speaker wire sets to the mating plug and will run them to the rear speakers once I've cut out the panels.