Sign in to follow this  
KDirk

88/89 Ultimate radio upgrade - part 1

Recommended Posts

Ok, this is a long post with some large pictures. If you are on a marginal internet connection, you may want to skip this.

First, a little background. I have long bemoaned the stock 88/89 entertainment system in the Reatta for being a bit substandard. It was decent enough 26 years ago when it debuted, but I find it sorely lacking in 2014. The tuner and EQ are fine. The integral power amp, not so much. First, it is not the best sounding amp even when it works right. Second, after 25 years it is almost always noisier than a bowl of rice krispies (snap crackle pop). Third, it leaves no good way to use outboard devices and maintain both good sound quality and integration with the CRT.

An aftermarket head unit is not really a good option, as it sacrifices the CRT controls and needs to be clumsily mounted in the center console, glove box, or modified cassette compartment. All that said, I have been working for a while on a definitive solution that will give me the following:

- good fidelity and output power without the flaws and constraints of the factory radio module's power amp circuitry.

- the ability to use outboard sources with good sound quality (and proper signal level balance between auxiliary and radio) and not resort to quick and dirty additions of input jacks, or other bogus stuff like FM modulators.

- Retain the functionality of the CRT for the radio, EQ, balance, and reuse the tape controls for rudimentary control of outboard sources (i.e. mp3 player)

This first installment will focus on how I modified a RIM module to eliminate the integral power amp, and add pre-outs so an outboard power amp can be used. An added benefit of this mod is that the factory radio is no longer crippled by "dynamic distortion limiting" or DDL. DDL was designed into the RIM module by Delco as a means of preventing breakup and distortion in the rather mediocre paper cone speakers installed in the factory. In essence, it operates as a variable low-cut filter, and limits LF output from the power amp at higher gain (volume) levels to keep the cheap factory speakers from being over driven. This feature is implemented in the power amplifier IC's, Delco #DM165. There are two of these power IC's in the RIM module, one for the front two channels, and one for the rear.

In simple terms, I have gutted out the power amp stage of the RIM module and installed shielded cabling to 4 RCA jacks added to the housing of the RIM module going back to the appropriate points on the circuit board. Once completed, I can run 4 RCA cables, one for each of the four channels (front right and left, rear right and left) to a power amp of my choosing. This way, I get all the original functionality and control of the CRT based radio system, including volume, balance, fader, EQ, DNR and source selection (between radio and tape) while using a new, and much better, power amplifier.

On to the details. I disassembled the RIM module, removed the large aluminum heat sink, both circuit boards and used a vacuum desoldering rig (Hakko 808) to remove the following components:

2x Delco DM165 hybrid power amp IC's

6x small green radial (leads on bottom) electrolytic capacitors

4x brown axial (leads on either end) electrolytic capacitors

I then mapped out the location of the 4 channels audio signals after they left the control amp but before they entered the power amp IC's. Once I identified all four channels connection points, I took a decent quality stereo RCA cable and cut it into two short (3") lengths. I stripped the outer jacket, tinned the center (+) conductors and twisted and tinned the two outer shield (-) conductors at both ends. I color coded the cables using electrical tape based on this table:

GREEN front channels

VIOLET rear channels

RED right (one each for both front and rear)

WHITE left (one each for both front and rear)

Next, I soldered the center conductor of each short piece of shielded cable into the respective trough-hole on the circuit board for each of the four channels. I then soldered the shield conductors on that same end to the metal heat sink mounts that were still on the circuit board after I removed the two power amp IC's These can be soldered to directly with a good soldering rig. I was using a Weller variable temp rig with a digital readout.

Next, to put in the 4 new RCA jacks on a new end panel that will take the place of the heat sink, which can be dispensed with as it was only present for the original power amp IC's. I had a piece of old computer case steel that I cut, bent and drilled to make a new end filer panel for the RIM module. I then mounted 4 individual RCA jacks and soldered the remaining end of each of the 4 shielded cables to their respective jacks. The complete end panel was mounted to the RIM module housing and electrical tape was used to apply color coding to identify the output channel of each of the four jacks. Having done that, I tested it in car with an external power amp. Voila!

I will cover installation of the power amp and associated wiring, including an on-delay relay for the external power amp at startup (to prevent a spike to the speakers caused by the outboard amp powering up before the RIM module) in the second installment. That will be forthcoming soon.

Part three will deal with modifying a stock tape deck to eliminate the need to have a cassette loaded to make the CRT recognize it, and an improved means of adding an auxiliary input jack for an outboard device.

Part four will deal with modifying the CRTC firmware and gutting the electronics from a factory tape deck and reworking them to facilitate better control over an outboard device from the touchscreen controls. The tape deck electronics will essentially be used as a remote switch driver so that basic functions like play, stop, fwd, and rev can be "sent" to an outboard device (with some hardware hacking) so those basic functions can be controlled from the CRT, giving it a more integrated look and feel. This part will be a ways off yet, as I am still working through some of the details now.

Note that I modified two RIM modules at the same time, one for each of my 88 coupes. So, if you notice some differences in wire color or other minor details in the pictures, that is why. And now, the pictures and diagrams pertinent to part 1:

post-52795-143142705971_thumb.jpg

More to come!

KDirk

post-52795-143142705926_thumb.jpg

post-52795-143142705937_thumb.jpg

post-52795-143142705942_thumb.jpg

post-52795-143142705947_thumb.jpg

post-52795-143142705952_thumb.jpg

post-52795-143142705957_thumb.jpg

post-52795-143142705962_thumb.jpg

post-52795-143142705966_thumb.jpg

Edited by KDirk (see edit history)
  • Like 1

Share this post


Link to post
Share on other sites

Great Write up Kevin, This is one things I was going to start on soon also. My Handle I use here is because I spent 10 years plus working in the Car Stereo and Aftermarket electronic field. Was First Class Certified Installer. Very good work !!!

Share this post


Link to post
Share on other sites

Sound-wise, how does it compare with simply using the Bose module? (Or did the Bose module I sent you not work in the '88?)

Share this post


Link to post
Share on other sites

Walt,

I need to do a proper side by side comparison. I was forced into taking action on this last week as the stock RIM in my scrub 88 finally died (power amp gave its last). So I picked up an Alpine Class D amp KTP-445U - very compact so it fits in the center console - and went to work modding. Had usable sound system back in the car a day later.

Now that the rush to fix it is over I can play around with it a bit. I plan on comparing the Bose unit to what I have done above as time permits. I expect the results to be similar except that the Bose unit lacks the EQ, and has tone controls instead. I'd prefer to have the EQ as it permits slightly better tweaking of the sound. The Bose module may be a good option for those not willing to mod a RIM module themselves. I may do a writeup on that as well.

I am still needing to add an on delay relay to the Alpine amp. It powers up before the RIM module causing a pop at the speakers. Also need to tidy up some wires, presently the Alpine amp is powered by a fuse block tap-off terminal. I will wire it directly to the stock harness as it can use the existing 15A radio circuit feed (fuse #11) and then I can elminate the fuse tap terminal and have the amp working in RAP mode as well. All that will be covered in the second installment.

KDirk

Edited by KDirk (see edit history)

Share this post


Link to post
Share on other sites

Try the power antenna power trigger lead to power up the the amp to see if the pop goes away, unless the power ant is in the down mode when using tape deck input. I was messing with my radio this past weekend to see if tape worked and don't remember if the ant went down when switched to tape from radio.

Share this post


Link to post
Share on other sites

Dashmaster,

Power antenna ouput from the factory module is source sensitive (I.e. retracts when tape input is engaged) so is not a workable fix in this case. Were it not for that, tapping the antenna power lead might work. Not sure that would give enough delay though. I counted off from key in run to the pop, and it will take at least a 3 to 4 second delay to bring the outboard amp online after the RIM module has powered up, thus preventing an audible pop at the speakers. It isn't too bad, but it's bad enough to warrant a work around.

On-delay relays can be had cheap off Amazon and ebay, and a 555 timer based unit is easy enough to whip up if so inclined.

KDirk

Edited by KDirk (see edit history)

Share this post


Link to post
Share on other sites

Yes 555 timers are easy to do. Since I have not started any work on mine yet I was just speculating on if the power ant lead would work. You are far ahead of me on this for sure.

Share this post


Link to post
Share on other sites

For your Part 3, you may be interested that you can easily enable the AUX input pins on the back connector (pins 8, 9, and 10 - next to the old speaker out and power plugs.) These pins are used by the optional CD player in the '90-'91 Reatta (and Rivs). To turn it on, apply +12v to the "T-ON" pin on the connector on the other side of the box. T-ON might be labelled as either pin 13 or pin 12 depending on rev of the module. It is the same pin either way - 4th from the end on the larger 9-pin connector.

The cassette deck audio inputs, on both CRT and head unit cars, are on pins 9-11/8-10 on the big 9-pin connector. These seem to be switched via higher level communications with either the CRTC or the head unit.

Share this post


Link to post
Share on other sites

Walt,

Correct. My plan is to modify the cassette deck so that the 12vdc ordinarily supplied to the ff and rew solenoids in the deck will instead drive a solid state relay of some sort (perhaps as simple as the control inputs on a quad bilateral switch IC or the GPIO inputs on a microcontroller) so that they, when activated via the controls on the CRT, will drive a discrete switch closure.

These can then be used to drive the track up and down functions on an embedded type media player. Same with the play control, 12v is supplied to the motor and can be tapped to drive a switch closure, as well as the "T-ON" aux control on the RIM module. The eject "button" will be relabeled "tuner" so that pressing it will stop playback of the outboard source and switch back to the radio.

I may repurpose the CrO2 switch input on the deck to serve some other, as yet undermined, function to activate the on-screen indicator (which will be reprogrammed to show whatever is pertinent rather than CrO2). Could be DSP for an add on effects processor, or any number of things.

The activation of auxiliary input by the tape deck is handled by E&C bus signals - a 12 vdc pulse width modulated serial protocol unique to GM and very difficult to interface with - so using the discrete activation afforded by the T-ON terminal is much easier to use and assures a hard cut-over that mutes the radio. I have some rudimentary data on E&C implementation, but not enough to take a stab at developing a MCU that can control it reliably. Maybe someday, although that seems largely unnecessary.

What I will use as a media playback device is yet to be determined. I want something that can drive a VFD or TFT LCD at VGA resolution and can be "skinned" (much like Winamp ifyou are familiar) to match the look of the CRT or something else fitting to the cars instrumentation (i.e. not too wild and flashy) while being a dedicated embedded system that can start and shutdown automatically with the car and require minimal thought to operate.

The display would mount in lieu of the cassette deck and require no modification of the trim bezel so as to look stock-ish. It would display data on the currently playing song and playlists among other things. Could also show static images such as diagnostic code reference lists or almost anything the MCU is capable of. Could have a bluetooth interface to pull streaming audio from a smart phone as an option as well.

The beagle bone and Intel Galileo boards show some promise for this type of project, even platforms like the Altera Cyclone III FPGA could be used. Problem is cost (particularly the Altera, though it is very slick) and time to scratch develop it the way I want it. Raspberry Pi and Arduino would even be up to the job, though I view those more as hobbyist boards and not so well suited to a professional quality embedded system with high reliability. Of course, boot times and tolerance of the environmental extremes of a car have to be considered too.

But, I am getting ahead of myself and rambling now.

KDirk

Share this post


Link to post
Share on other sites

Good Idea. How did this work out for you. My 88 needs, I've been rebuilding for the past three years, need some help on the radio. I like the original look but want something behind the tape deck cover with updated technology. I understand your thinking on adapting the current electronic configuration and am comparing to a total retrofit but keeping the CRT in play.

Share this post


Link to post
Share on other sites
if you could tie steering wheel controls to this that would be awesome.ive often almost run off the road trying to change radio stations or volume.

If you can find a suitable steering wheel from a 87-95 GM, with the controls on it, then you can wire in the controls to the E&C data lines. The lines are one or two wires. I used them to control a DVD player a while back.

Share this post


Link to post
Share on other sites

Daniel,

That almost works. Here is the problem. The CRTC does not follow the E&C bus to reflect changes sent from a second "master". The bus master is the control head, in the case of an 88-89 that is the CRT via the CRTC interface to the radio module. The CRTC firmware is such that it only recognizes one master. If a second is added in parallel (like a tech 1 tool in E&C mode or a compatible control head from another GM model that used the same RIM module) that second master can send E&C commands that the RIM module will respond to, but the CRT display will no longer be in sync with the radio.

For example, if you use the CRT to set the radio to 94.7 fm, then use a second control head to change the station, the new frequency will be played by the RIM module but the CRT will still show 94.7 FM. I tried this both with a TECH 1 using the 88-89 body systems module radio diagnostics and with the control face removed from a 90-91 type stock radio paralleled onto the E&C bus. One will not follow the other in real time. This results in confusion as the CRT radio display will get out of sync with what the radio is actually doing. So, I don't consider this a proper solution. That's just me, others may find it an acceptable trade off if it gets them steering wheel controls.

Now, having stated all that, I have yet to try the steering wheel control system from a late 80's pre air-bag Bonneville with the factory sterring wheel control system. I have scrounged the parts from one (both the keypad and the E&C translator module). 88-89 Toronado had a similar setup as an option both with and without the VIC.

Maybe there is something different about this hardware that allows it to work properly in tandem. I intend to wire this in and see if perhaps it works correctly but I am not counting on it. If it does, that would be outstanding. If not, I am back at square one.

KDirk

Share this post


Link to post
Share on other sites

Again great work and a definite step forward. Correct my logic if I'm wrong, but now that you've proven some tandem signal riding piggyback on the E&C bus does work, yet albeit one directional where the second device signal steps on and over the originating signal, can't you now check for a reciprocal signal being sent back to the originating device, ie ca returning confirmatory signal from the radio? Somewhere I recalled reading somewhere about how the tape deck behaved that lead me to believe there was a challenge and response system of logic in place and being used. Like for instance, the radio wouldn't change inputs to the tape in if no signal was returned from the tape deck (such as a logic tree that checked if a tape was in and the deck was turning before activating that input). I'm thinking this is basic xor perhaps even xnor logic. If a reciprocating signal is getting hung up because there's an overriding signal, then shouldn't that be easy to prove with a double command from each source followed by a single command from the touchscreen? If you tune to one station with the touchscreen and then another with a translator module and then a third with the touchscreen, does the touchscreen reflect the first or third. If it reflects both, and I think it has a good chance, you then possibly could leverage the nand and invert the instructions being passed from one or the other device. Found this simple TTL NAND circuit at: http://www.allaboutcircuits.com/vol_4/chpt_3/5.html that looks like this may work for what you describe your encountering:imagepng

Share this post


Link to post
Share on other sites

Good info there. Here is my take. There is some strangeness to the interaction of the CRTC/RIM/casette deck. The CRTC queries the tape deck at bootup (car started) and if found with a tape loaded, enables the use of the on screen tape controls. If no tape is loaded, or the tape deck does not respond to a hail request (to verify is present) then the tape controls are locked out even though they appear on screen. Loading a tape (assuming the deck is otherwise functional) at any time enables the on screen controls.

The tape deck does send confirmation commands from it's own on-board logic to tell the CRTC and RIM the status of various parameters; whether the deck is in play, ff, rew, the direction of play (indicated by the PROG arrows on screen), whether the loaded tape is cr02or normal bias, when it has found the found the beginning of a track with the seek forward and back and thus resumed playback and probably a few others.

The tuner is an odd case. The RIM module has it's own internal memory for storing presets, the real time clock, EQ curve settings, volume level and presently tuned station. In the CRT system, it was apparently decided not to use this functionality and the CRTC RAM is used to store this data instead. This is part of the problem, as the RIM is not maintaining this data and feeding back to the CRTC for dispay, rather the CRTC is commanding the RIM by direct parameter input based on what the user has done on the touchscreen.

So, it functions as a partially bi-directional bus; that is only when it needs to. Otherwise the CRTC assumes that what it "knows" (what is presently indicated by stored variables in it's own RAM) to be the correct operating status. The CRTC does not poll the RIM at set intervals, nor does the RIM immediately push parameter changes to the CRTC, to insure that sync has been maintained. Thus the issue being found with E&C commmands not generated from the CRTC not being reflected back to the CRT.

I think the ideal solution to this is a modified CRTC that could have it's RAM selectively overwritten by an external microcontroller. This would force the screen to reflect the true status of the radio even if it had changed from some external controller. This would be a mess of work and fraught with the possibility of crashing the CRTC if anything went wrong in writing new data to the RAM - either bad data or bad timing of writes to memory. Trying to do this on an old embedded platform (dual CPU 8088/8052) without much documentation would be challenging, to say the least.

So my next idea is to inject simulated key presses on the touchscreen bus. An external micro could emulate this easily enough (touchscreen is a 5x5 matrix) so that when a command was issued from the steering wheel controls, it would change the CRT to the radio page, send the simulated keypress for the desired function, then switch back to whatever screen was previously displayed. The same could be done for basic climate control functions. The MCU would have to store which of the 6 main pages was displayed prior to switching to the radio page. All this would require a modified CRT with some I/O lines added. This would suffer from some system latency (the time it takes to change screens) so there would always be a slight delay but that's better than nothing.

For now, I really need to try the steering wheel control setup from the Bonneville and see if suffers the same problem. If it doesn't then we have a solution, providied the needed parts can be tracked down from a donor car. If It works then most of my ideas above become moot.

KDirk

Share this post


Link to post
Share on other sites

Kevin, GM used a couple of different steering wheel control (SWC) systems back then. According to the Delco "Sound Systems Diagnostics Guide", there was both an optical and an analog system. In the optical system, there is a E&C translation box hidden somewhere. In the analog system, sometimes there is a box, and sometimes it is built into the head unit. I also note that in the analog system, it is pretty important that the steering wheel buttons match the translation box. They document two different cases, Pontiac and Olds, where the resistance values for the switches have different values for the same function.

One thing I've wondered, for those of us who have '90 and '91 Reattas, is if the steering wheel from a later Buick will work. Buick mostly used the CDM radios up through '94, but the '95 Riviera also had a CDM radio. So one would hope the steering wheel controls would be compatible. The '93 Lesabre is especially interesting because they used both the two spoke wheels like the Reatta, and the four spoke wheels that were use through the rest of the '90s. And I suspect the air bag system may be compatible through at least '96. (In '97 they went to the SRS II system.) Though the air bag connector changed slightly somewhere around '94 or '95.

Share this post


Link to post
Share on other sites

The Pontiac Bonneville used a rotary contact collar hardwired system with a translator box. The Toronado and Trofeo used a series of IR transmitter and receivers in a ring configuration within the column to send the commands from the wheel mounted controls downstream to the translator box. The Bonneville system is more easily adapted to the Reatta where the Toro setup would pretty well require using the complete column from a donor car due to all the extra optocoupler parts involved.

Near as I can tell, the Pontiac translator box intercepts resistance ladder coded commands, converts these to their respective functions then directly transmits a corresponding command onto the E&C bus where the RIM module (and by extension the control head) act upon the command and update the display as well. There are also temp and fan speed controls as part of this setup, those commands being picked up by the ACP and ECC head in turn. Whatever the climate control interface is (usually dedicated panel as in the 90/91's for most GM's of that era) should update it's readout as well.

The crux of the problem is that the CRT system does not interface the same way the normal radio and climate control heads do, and so it doesn't appear to listen for and reflect parameter changes that are not initiated directly from the touchscreen. This is a shortcoming of the firmware, as the hardware is all there to accomodate this functionality, it just isn't programmed to do so. Trying to find a work around for this is the biggest problem I'm running into.

KDirk

Share this post


Link to post
Share on other sites

The optical system that your Olds parts apparently use are shown in the manual as having four components: The steering wheel/controls, the 'cancel cam' in the column, the turn signal switch - also obviously in the column, and the translation module. Bet you don't need a whole column. Just the cancel cam and turn signal switch.

If you are interested, I can scan the pages and send them to you.

Share this post


Link to post
Share on other sites

Thanks for the offer, I actually have original copies of the Toronado FSM for 89, 90 and 91 that I picked up as reseach material a while back. Also gave me some documentation on the VIC, which I'd still like to do something with someday. Of course, there is never enough time.

As far as grafting the Olds parts into a Reatta column, I loathe steering column work. Have had a few apart and try to avoid such entanglements to the greatest extent possible.

KDirk

Edited by KDirk (see edit history)

Share this post


Link to post
Share on other sites

Good reading guys. There really isn't enough time. I really wish I had more as I'm sure everyone does. I'm glad to have what I do. Used to be I'd have a "winter" project and a "summer" project. One year has run into another and then another and then another. I really think a CRT replacement is somehow possible, but the built in simplicity is a complex and elegant design to be sure. As many cars as I've owned or been in that have a touchscreen, I feel the unique set of pages the Reatta offers is among the best, or most intuitive. Thanks for putting in the time on this. Certainly, if history is left to the Facebook generation to write, this sort of intelligence will slip away. Good job!

Share this post


Link to post
Share on other sites

Awesome Kevin :)

 

I just installed a set of Alpine SPS-610's last night and middle of the week I bought a Kicker 11HS8 8" powered mini subwoofer unit. I didn't realize how bad the factory rear speakers were and bought the sub as a result to fill out the sound (haven't installed that yet).

 

In my testing last night, I found that DDL coming into play (blech). System sounds much better, but of course the damned left-rear-channel sizzling and half-volume-overall-during-sizzle syndrome has reared it's ugly head at times again today- ugh!

 

I dunno if you're interested, but I have a color scanned copy of the 27D-1988-3E tuner, 27D-1988-3F deck, and 27D-1990-3E tuner manuals. The first two are scanned color jpg's from the local Delco repair-house and the 3rd is a pdf found on the web....lemme know.

 

Dan Gibbs

Share this post


Link to post
Share on other sites

Dan,

I would be interested in those scans. I have some poorly done scans I found somewhere along the way that aided my efforts but if you have complete copies in good quality they'd be nice to have. I'll pm you my email. I need to get back to this project but work was stalled by my convertible and work I'm doing on that. I really want to decode the E&C bus protocol so I can interface a modern media player in place of the cassette deck. My initial efforts to that end show it will not be easy, the protocol is not at all documented and is a highly unusual specification, 1Kb/sec data rate with variable message length by way of a 12v pulse width modulated signal.

No currrent mainstream MCU of the hobbyist variety is suited to handling such a communications scheme which likely means using something like an 8051 to handle the translator interface. Not nearly as easy as programming an Arduino or the like.

KDirk

Share this post


Link to post
Share on other sites

PM me when you get a chance. I winrar-ed the color copies and they're around 94Mb each (ouch, problem for email with 10Mb limit). Another nice copy of the 88 Tape Deck is 350Mb (dbl-ouch). and the 1990 Tuner is a pdf at 3.2Mb

 

I'm sure there's another way to send these to you via internet, my mind is just going blank right now other than Dropbox.

 

Dan G.

Share this post


Link to post
Share on other sites

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
Sign in to follow this