Pigzmickey

Complete installation instructions including pictures are suppled.
Vehicle colour codes vary between model years, so these will be supplied with each individual unit sold.
The installation is completed under the rear seat at the roof ECU, no alarm connections are required. It took me approx 30 mins to complete my installation. As you can see, there are 5 wires to cut & 7 to splice. I've attached a schematic:

mebeJOE

Okay, I want one. However, because you are in UK and I am in USA, I guess I am concerned that if I get sent one of these modules, I might not be able to make it work -- and then I am stuck. I would like to see a post by one of the other board members after purchase and install.

bobterry99

A friend just bought a '90 300SL from me and wants the roof automated. I've designed a circuit to do this that I have simulated on a PC with PSPICE. It will be a few days before I get the parts to build it and try it out. Cost: less than $3 for a pair of 3-input NOR gates, some resistors and capacitors.

I will freely share this design with anyone who is interested and would welcome ideas on improvements. However, I don't think I am interested in building completed circuits to sell to others.

Pigzmickey

Remember:

Get it wrong, you better have deep pockets! Although roof ECU's are not rare @ MB dealerships, just expensive...

bobterry99

Unless someone is hopelessly negligent, it is impossible to damage the top controller with any of these modifications. The signals being interfaced to are designed and intended to be shorted to both ground and battery potentials and everything is current-limited.

I'm attaching the latest revision for fully auto operation with a 1-button remote. The diagram shows 7400-series gates, but the actual ones should be CMOS 4000. If Pigzmickey or anyone has suggestions, please pass them along.

mebeJOE

Hey, nice drawing. But I went to the electronics store, and they didn't have any circut chips shaped like the ones in your picture. The ones they have are all square or rectangles and stuff . . . and they have these metal legs coming out. So, I just bought a bunch of them -- some of them had some of the same numbers and letters you have in your drawing -- and I tried to figure out which legs to attach the wires in my car. Well, after making a couple of attempts, I smell something burning, and my top won't go up or down. I'm glad we have a real professor Einstien here to help us out.

99_SL_Sport

Hi. Several questions for bobterry99. What are these 4000 series gates? Where do you buy the parts? Where do you make the connections for the status and the switch and outputs to the controller? A friend of mine tinkers with electronics and can help me, but how much would you charge for one? And finally, I have a remote with lock, unlock, panic, and trunk buttons. How would I wire the circuit for this? I appreciate your time.

bobterry99

'Sport, the gates are 4011 and 4025. Depending on the manufactuer, there will be a couple of letters in a prefix and/or suffix. These are very common parts that are available at an electronics supply store or on the Internet at eBay, Mouser, Digikey, and Newark, among others.

As far as making connections, if you were just interested in having your top work automatically from the switch on the console, then you would make all of your connections right there at the switch. Otherwise, you would want to do everything back at the controller.

You asked about a price for this from me, and I'm sorry, but I won't be selling anything or constructing these for others. If I were to charge something, it would probably be comparable to what Mickey is asking, and there'd be no point in that. This is only being offered as an option for DIY people who want the feature and would like to save some money. Besides, anyone who has the ability to install it would surely have the ability to build it.

The circuit for the 4-button remote will be nearly identical to that of the 1-button remote. The likely difference will be that a status signal from the IRCL remote controller will substitute for the one from the top in compartment switch. I'm still waiting on parts for a prototype and will post back as matters develop.

rellik

what values should be used for the resistors and capacitors?

ill breadboard one up soon

bobterry99

That's a really good question. My version of PSPICE is an evalutation version that has no models for the 4025 and 4011 logic gates that are used in the design. So, I can't use a PC simulation to determine values.

I have the 4011 NAND chip and am waiting for the 4025 to be delivered from Digikey by mail before breadboarding a circuit that I will connect to the car. Meanwhile, experimenting with the NANDs and substituting one of them for a NOR to determine what R1&C1 and R2&C2 should be, it looks like 10K ohms for the resistor and 100uF for the capacitor work pretty well to give the 1-2 second delay wanted for actuation from the control switch. As for the 5-7 second delay on the status signal to allow for the windows to roll-up, that would depend on the output resistance of the signal from the controller. Rather than measure that and breadboard, I'm just going to wait for the 4011 and determine R3 and C3 through trial and error while connected to the car.

I'll mention something here that I mentioned on one of the other forums. I don't have a strong interest in any of these modifications for any of the cars I own, and I'm not interested in developing something I can sell. Therefore, I am really reluctant to modify wiring in one of my cars just to prove one of these designs. That said, there is nothing to lose at all by building a basic circuit that simply automates operation from the top switch. This is because the signals for that circuit are all available at the switch, and it, naturally, can be separated from its connector and you can interface the new circuits to the male pins of the switch and the female sockets of the connector without cutting any wiring. I'm currently looking at ways that the signals from the infrared remote control module can be interfaced to that do not require any wires to be cut.

tipds

You say this worked in simulation? Did you simulate using transient inputs? I suspect, based on the lack of latching mechanisms, that you used constant (voltage) supplies as inputs.

Previous comment not withstanding, I think you would have a more deterministic system if you employed timers of the 555 nature or delay relays.

Also, did you consider loading at the outputs? You may get a more efficient chip count if you perform boolean reduction... I could be wrong on that one, as I didn't do a thorough analysis of the circuit. I would be interested in considering the first two comments before persuing this analytic exercise.

All that said, if you do get it to work reliably on a real vehicle, I would be interested to hear abou it.

Tip DS

P.S. Are you an EE student? I'm guessing you are, based on your choice of simulator.

mebeJOE

Nothing to lose? How about freakin' time? That's all we have in this world. Has anyone purchased one of the INSTALLED AND TESTED models from Pigzmickey?

If so, please post a review. I'm waiting for some feedback before purchasing.

LeaUK

BT99

I made several comments on the parallel thread running over on the other forum, some have been repeated by tipds and others so I'll comment here too.

Please take this as positive criticism as we only want your circuit to evolve and be useful to everyone.

Design guidance:

1. Please show decoupling so we can advise

2. Avoid relying on unknown impedances to assist with timing

3. Use controlled methods to achieve accurate and repeatable timing ie. a 555 would be a significant improvement - but it's not a great selection!

4. Use ceramic decoupling (10nF + 100nF) for each IC

5. Don't forget power supply i/p decoupling - 10nF + 100nF + 47uF Tant >=30V (or equiv)

6. Avoid slow ramping signals into gates unless the gates have schmitt i/p

7. Use schmitts on ALL i/ps and buffers on all external o/ps (I use schmitts for o/p too)

8. Apply signal conditioning to all i/ps - ie overvoltage protection etc

9. Source and sink currents of 4000 series is only 2mA @ 12V

Just some food for thought. BTW I hav a full version of PSPICE should you wish to experiment further ;-) PM me if you're interested.

Lea

Bigdot

What The Hell Is Everyone Talking About!!! Im Soo Confused!

tipds

Very well said. Looks like great advice on all points.

Now that it's a new day, and I review my comments, I can see that they may (easily) come across as digs. Please understand that I wasn't trying to be an @$$. I have been told that I come across as arrogant, and I can see why. Believe me, I'm no better than the next Joe, and I know it. Please accept my humblest apologies for the (very very unintentional) tone.

Goose: You crack me up.

Lea: Did you used to use pspice for work? No "tone" intended, but... that's pretty old-school. Surely you use newer tools?

Tip DS

99_SL_Sport

Why don't you quit whining and actually make a postive contribution to this thread? You could be the first to buy a module from Pigzmickey and report to the rest of us. But of course you won't. Your too cheap.

tipds

Now, now. Let's play nicely. Despite my wife's comments about what a frivilous expense it is, I woul buy one in a heartbeat, if only I could get it installed by a factory qualified tech. (Did I just write a major run-on sentence?) My english teacher of a mother would be horrified.

Tip DS

LeaUK

For work I use Cadence CIS + PSPICE version 10 - doesn't get much more expensive than that - believe me! Although the Berkley PSPICE model regarded as a classic tool, it's still used heavily by our Design Engineers and is an excellent tool - just too expensive. If I had to carry out a new EDA search I would look at alternatives most definitely.

I carried out the departmental EDA tool chain search, justification and purchase some 5 or so years ago so it maybe a little outdated now

Back on-topic:

I personally think BT99 is carrying out a fine job of designing and actually doing the work and he's even prepared to test on his own car - he has got to be commended for his time and effort. Cheers Bt99 great stuff!

tipds - BTW I personally didn't think your comments were negative or dipict arrogancy, just good advice.

My last point - I have to say that I would be a little disappointed if BT99 doesn't take on board at least the primary issues faced by the current design and doesn't modify the circuit to reflect. I'd assist if requested but perhaps there's to much veroboarding to be done ;-) as no word for a couple of days now.


Lea

Pigzmickey

Do the R230 SL's do all what my module offers?

Only yesterday I was pulling out of a fuel station (which I frequent often having the V12) & started putting my roof down as I carried on along the road, I saw an SL350 driver giving me the evil eye.

I just wondered if the new models could do that as well?

P.S I have sold these modules on eBay this week.

bobterry99

The simulation was a transient analysis with switches to provide inputs that simulate pressing and releasing the top switch and activation of the status signal.

With regard to timing signals, the time constants are not at all critical -- the margin for error is greater than 100%. Therefore, I chose a simple R-C network.

Regarding loading, rather than measure input and output resistances of the Mercedes top controller and then choosing appropriate values, I decide to evaluate through trial and error while testing my circuit connected to a car.

Finally, I'm not an EE student. I chose PSPICE because it was free! If you can direct me to something better, I would appreciate it. Meanwhile, I will be bartering with Lea for a full version of PSPICE.

Your concern about capacitor discharge were valid, Lea. I put a diode in parallel with R3 for this. I don't think it is an issue with the the R-Cs, and besides, in the basic design that I am attaching here, I decided to omit them.

You mentioned some other changes which included protection of inputs and decoupling...My breadboard is pretty cluttered, and I'm going to hold-off on those for now while I evaluate the circuit.

They operate the top from the remote as well as the switch on the console. But for safety reasons, you still must hold the button and you have to be standing close to the car with the transmitter aimed at the door handle. When I had an R230 I never bothered with this feature -- it seemed like nothing more than a gimmick.

______________________________________________

Okay, now about the basic circuit which is currently working in my friend's 300SL (attached). It does not operate from a remote -- that design is still untested. Cost is less than $2 for the electrical components and about $2 for a small board to assemble them on.

The only anomaly I noticed was that if the car battery is disconnected and then reconnected, the top will automatically close immediately if open. I have a circuit modification to prevent this that will be incorporated into the design for operation with a remote. Without that mod, if you don't want the top to cycle when you reconnect the battery you can pull back on the top switch and cancel the action.

This circuit is essentially a simpler version of the others I have posted. The new schematic shows the connections at the top switch. Both the switch and its connector have pins that are clearly labeled 1-5 and 7 (they skipped 6).

If you want the ability to pause the top mid-cycle by momentarily touching the top switch, you can add a pair of resistors and capacitors. These were shown in an earlier schematic. Anyone who wants this feature can contact me for that simple modification.

The value of R3 and C3 set the time that the windows are rolled-up at the end of the cycle. If you have an older car with slow windows, you will want to increase one of these values. I'd suggest an R3 of 2.2K or a C3 of 1000uF.

If you decide to construct this circuit for yourself, there is a modification to your wiring that is not shown. Socket #2 of the connector is +12VDC that is switched-on with the key (Mercedes calls this "circuit 15"). But you need the top to work with the key off, of course, so you will need to bring unswitched 12-volts to this socket #2 (Mercedes calls this "circuit 30"). There are a number of places around the car where this can be done. If anyone needs assistance with that, then contact me.

One other point that was mentioned in an earlier post. I think the parts are available from M-B inexpensively to create a little harness with a male and female plugs that would allow you to just plug the new circuit in without cutting or splicing any of the existing wiring. I'll be looking into this.

Tomorrow I'm going to begin testing of the circuit for operation with a 1-button style remote.

LeaUK

I'm glad that your circuit proved fruitful but without adding the majority of suggestions made previously I would be personally concerned about operation across a broad range of environments and cars - just my observations.

As this circuit is a replacement for holding the button manually, am I correct in assuming that the top module will still obey the speed sensor and not over ride it?


Lea

Pigzmickey

Guys, I guess thats it. I won't be bothering to promote these modules via this forum anymore, I don't make enough from them to afford forum tax. .................................................. .................................................. ....
Pigzmickey:

This notice is a request that you cease any and all sales of products on MBWorld. Unless you are a paying sponsor, you are not allowed to conduct business on MBWorld and are in violation of our Terms Of Use Policy. Should you wish to become a sponsor of the community, please email advertise@mbworld.org. Otherwise, further violation of our terms will result in a suspension, up to and including a permanent ban from the forums.

We thank you for your prompt attention in this matter.

MBWorld Admin
__________________
Mr. Vanos | MBWorld Administrator

tipds

bobterry99: Kudos! Clearly _I_ am the one that needs to go back to school! That's not the way I would have done it, but I didn't have the initiative to do it at all. Great work.

One dem days you're gonna have to seel yer design.

Tip DS

LeaUK

I think we all agree that this project needs a uC (PIC 16F for example) and could be coded in a couple of hours, however this would mean increased expense as you would need to either 'purchase' the programmed uC or purchase a uC programmer - assuming the code was available.

I'm not one for rigorous cost cutting and so my preference and experience dictates a micro for but that's not to everone's taste or natural ability so we'll see how BT99 progresses - with eager interest.

Just to ensure admin are not confused over this thread's evolved structure and to reiterate, with regards to BT99s proposed design this is purely investigative and nothing is being sold . It's all free! This is a community of R129 owners with some of us enjoying the technical challenge of repair, design and construction so I hope it will remain.

Enjoy...

Lea

bobterry99

I talked to an EE friend of mine and he echoed exactly all of the suggestions that you made, Lea. He says whenever he designs a digital board he decouples all of the power pins. And he suggested gates with Shmitt triggers on the inputs for further noise immunits. It's just good engineering practice. For a retail product, he regarded them as a necessity, but for a DIY version he considered them optional.

The top module will still obey the speed sensors. Are any of us interested in a circuit to over-ride them?