Oh you should see the stuff I'm procrastinating on...

 

Alright, progress update. Made a decent amount of headway.

I'm almost 100% certain that the two controller chips are NEC uPD75108B maskrom microcontrollers. To replace the microcontroller, I need to address two functions.

1. Interfacing audio with existing DAC and filters. This part will be the more straightforward half. It will probably be easier to desolder and remove these items and place on a new pcb than have an adapter ribbon to install and then have to disable all the other components on the existing board. It would also simplify things a bunch. The audio connections I would need from the new microcontroller are the 3 standard I2S lines; data, l/r, and clock, along with two aperture lines to help with any excessive oscillations (from what I have found, these may not be needed with modern digital audio sources). I'd have the aperture lines hooked up as an option, and if they are not needed, they can be left out of the programming. There are two other lines that will need hooked up, the deemphasis and mute circuits. I'm still trying to understand these, but it looks like the deemphasis is to quiet down "noisy" signals, and the mute is engaged whenever the optical pickup is moving around, or the CD unit isn't playing music. I'll have to read more about how they work to implement these, but deemphasis and mute only need one circuit each to run. This should allow music to be fed from an I2S audio source, which a lot of available hobbyist microcontrollers seem to support. Good deal there. Also by removing all the extra circuitry, it frees up the 12v lines to be used for whatever. Currently just looking at bluetooth, but that might be able to supply enough power to enable other devices like a microphone or something for answering a call (feature for a later revision).

2. Integrating with controls. This is the difficult part, but should be a lot more possible since I believe I have ID'd the two control chips.
I have been looking to see where the Vdd for these chips comes from. It looks like all the main board and tape deck chips are all powered by a voltage regulator on the main PCB. Also the Vdd for the CD control chip on the CD pcb is also supplied with this Vdd. The power for all the other CD chips comes from a +5v regulator that only the CD controller chip can turn on and only powers stuff on the CD pcb. It keeps the CD chips powered down unless they are in use, probably to keep the thing from draining your battery from them running all the time. I think that both control chips are powered on when the unit receives +13v Batt power. It looks from the datasheets that the 75108 chips have a low power mode, and probably don't turn all the way on until it sees the signal from +13v ACC. This would make sense because if you unplug the unit, you lose your radio stations, but with battery power, it would keep your stations. Kinda like how your old nintendo cartridges have the coin cell batteries to keep the game saves in the static ram. If we got rid of the extra stuff in the CD portion, it frees up a bunch of power headroom that we could use at our discretion. The main power consumer at that point would be the amp units. I've just got to confirm the power-on status with a scope or voltmeter and that should be pretty straightforward. There is a reset line that runs to both controller chips that ensures both chips are reset to their starting point after a loss of power. I'm pretty sure it doesn't do anything after the initial power from +Batt. This would likely just be a status indicator input for the new microcontroller.
There is an LED output from the CD microcontroller that turns on the DISC SET in the upper left corner. Easy straight output from the microcontroller. Thinking this would work well as a bluetooth connected indicator.
That leaves 5 lines that I still need to figure out their function. From what I have seen, only 2 or 3 of these could be for the purposes of serial interface. I am trying to find out more information on this. It is referenced in the datasheets as NEC Standard Serial Bus Interface (SBI). If you guys could do some digging, that would be helpful. I haven't quite found the right thing. I have found SPI, and I2C interfaces, I wonder if SPI is what I am looking for. The lines I have are BRXEN, BRST, BDATA, BSCK, and BSRQ. BDATA and BSCK are fairly obvious as data and clock for the serial communications. BRXEN is an auxiliary enable output from the CD controller, purpose unknown. BRST Bus reset and BSRQ Bus request I also have to figure out. If we can tack these down, there's a pretty good shot at decoding it, or at least decoding what we need.

I am trying to get ahold of another professor at the university who would have grad students that could help me. It's more about who you know, than what you can do by yourself. I am still out of my depth, but I think I've made some big progress here. Let me know what you think, or if you guys find some relevant info.

 

Hope your project is going well... Only posting because I am also looking for the this bus protocol documentation or explanation.
Got here because you're the only mention of BSRQ on the internet.
I'm trying to adapt a bluetooth device to my own old radio (Pioneer KE-83) in a 1990 E32 BMW 750iL and have matched all the lines except the BSRQ line, and want to know if it's redundant for my needs.. https://maakbaas.com/bluetooth-cd-ch...-to-prototype/.

From what I can find the NEC SBI is also called the 3-line SIO or 3-wire SPI... hope it gets you in the right direction.

 

Welcome to the forum. I'm glad there are other people looking into this as well. Kinda felt like I was one of the only ones.

I've read that forum post. A lot of good information there. That guy went for a somewhat different use case, adapting a newer head unit to the cd changer. I think for our stuff it might be similar, but until I get a signal decoder like that, we won't know for sure. Can you tell me a little bit about your radio? Is it similar to ours, running a pair of 75108's for control?

Thank you for the SBI clue. I'm still looking through it. It's getting to be the busy season for me again, so updates are probably going to slow down more than they already have.

 

​​​​​​Yes I think it runs on a variant of the 75108. The schematic only shows 1 chip.
​​Good thing too, I've got enough duplicate computers with V12 as is.

Anyway, I briefly spoke to my Electrical engineer friend about this project and she said to just wire it up and test. (I am a little worried that pricey lil Bluetooth amp will get fried... but that's half the fun right?)
The SQR line
​​​might not be essential to function, but she also said it would be essentially impossible to manually spoof.
The mbus protocol notes I found say that the SQR line is only essential to tell the head unit that the peripheral is present and its ok to send data, since the head unit doesn't automatically search for peripherals on startup. The later KE-91 radio for which the Bluetooth adapter was designed does this automatically and has no the SQR pin.
​

 

Thank you very much for this info. It looks like we are at least in the same ballpark, if not pretty close in design. I'm sure this will be a big leap forward on this project. I haven't had a lot of time to look through, but from what I did review of the BMW project, BRXEN is likely the same thing as his ENBL pin. As far as BSRQ, I think it might be used as a sort of chip select. I haven't done anything to confirm it yet. On our Rx7 Pioneer units, we can boot the CD unit, which is built like a peripheral, into test mode. This pin might be used when the CD unit is in test mode to take over as the master chip (entirely speculation). Do your radios have a test mode so you can adjust and configure your CD changers?

 

BSRQ in our unit is used to search for a peripheral. Process goes like this (i think)... Power on - Headunit sends SRQ out - gets response from peripheral (CD changer) - sends BRXEN/ENBL signal, this is the "handshake", the data lines light up and the 2 units start talking.
The unit in MaakBaas project is more advanced and skips the SQR part, automatically initiating the BXRN/ENBL when switched into CD mode.

​​​​​​Afaik our units do not have a test mode. Couldn't find mention of one in the document.

When I set this up, I'm going to test with and without the SQR pin plugged in. Hopefully it's not needed. Just waiting on AliExpress parts to make the patch cables. ​

 

Hi,

Just thought I'd register to say this is an amazing piece of dedication to revive the old Mazda radio. I'm currently restoring a very similar unit (Pioneer FH-5071ZM from my 1993 Eunos Roadster, or the Mazda Miata as it's known in America) which shares it's architecture with the DEH-K4041ZM and is actually how I found your project.

With regards to your CD player not working where it briefly spins and then rejects the CD, I've managed to fix this issue on my two units very easily so they now read and play music again. Given how similar these units are I thought it would be worthwhile sharing with you to see if it helps you too. When you open up the CD player and look at the laser assembly, there is a tiny potentiometer on the side. I've turned this clockwise a tiny fraction, probably less than 1/16th of a turn, and the CD player now functions flawlessly. Please be aware that I've done no other testing than to power the unit up on a bench, hooking up a speaker and prove all CD functions work correctly (skip track, search, repeat etc.)

This potentiometer regulates the power to the laser diode. As the laser ages it degrades and eventually gets to a point it's too weak to send the correct signal to the control board. Increasing the power to the laser has brought back functionality but I expect it's potentially accelerated wear on other components.

Photo of the potentiometer location



And here's my radio on the bench. I'm sure you can see the similarities between this and the S5 radio

 

Here's a couple of before and after videos from adjusting the laser potentiometer on my second radio. Exactly the same procedure as I carried out on the first radio. Excuse the music choice, I had to borrow a CD from my neighbour to test as all mine are in storage

- the clicking you can hear when it tries spinning was sorted when cleaning the spindle top hat.

 

Welcome Rob! Thank you for the post. I'm happy you found success with this old tech. The pictures and video are great to have too. I expect that more people with similar radios will start drawing inspiration or using these posts as a starting point as people start cracking open these old units.

It does appear to be very similar architecture for the main unit. The AE unit may be a bit different, but I wouldn't expect it to be too dissimilar either. The laser revision is a bit newer judging by the CGY1015. The ribbon connector for that laser is a bit different than ours, even though I think the main laser is the same. What led you to adjust the potentiometer on the side of the laser? Do you have a manual for that radio that outlines that procedure? The manuals I have make no mention of ever touching that potentiometer that I can recall. Perhaps I would see more signal on the oscilloscope if I were to adjust this on the unit I have soldered back together. There should be a specification for setting this potentiometer, although I have no idea where to find that information. I'll have to look into this more now that you got me curious.

As an update, I do have a spare unit that I got earlier this spring. I haven't had a chance to test it to see if the CD drive works. The problem with all the units I have received is corrosion or shock breaking the suspension of the laser units which both suspend the laser lens, as well as carry the electrical current to the tracking and focus coils. One of my former students showed me that he has a laser cutter that can cut shim stock. He said he might be able to make those fine suspension pieces, but the problem then becomes how do you assemble them with any sense of precision in the laser assembly?

Here's where this project sits currently. I have some time this next month and plan on building a test bench for S4 and S5 RX7 stereos. If I can work on a few of them for small fixes, it might keep funding this project. If the CD unit works in this new unit, I would need to pick up a signal analyzer to intercept the microcontroller communications and hopefully confirm the information that beew provided. I haven't been able to find a programming guide for the microcontrollers in this radio, but I did find that another NEC microcontroller was used in the chassis CPU. Information on both of them suggest that the programming of both types of microcontrollers is similar, if not the same. I also need to learn how to design a flex cable that connects to the current flex cable and the CD unit connector. I think that a procedure to reverse engineer these microcontrollers might be another contribution in itself. If we can pull off something like that, I might have to write it up an get a second Ph.D. out of it.

Again, Rob thanks for the post. If you have ideas or contributions, I'd be happy to hear them.

 

Forgot to ask, Rob, would you happen to have a copy of the manual for that unit?

 

Hey guys,

made a little bit of progress on the research side of this project. I found the datasheets for the two NEC microcontrollers, uPD75801. I also found the booklet to program them using the old tools. However, I can't find the old tools. I have found some homemade new tools to program them. I am looking into a dissassembler program to be able to deconstruct the roms on these chips. I found a couple options briefly but I haven't had time to look into them. Been a busy summer for me. The programming instructions here are for the NEC 75xx programming family. It is also worth noting that the car alarm microcontroller in the CPU is a NEC uPD7507 microcontroller. It is like a smaller sibling to the microcontrollers in the radio and uses the NEC 75x instruction set. From what I have found, both instruction sets are close to identical. Might give us the opportunity to make some more cool stuff.

As far as physical progress, I am currently stalled. I have a bunch of structural repair work getting done on my house, and haven't had time or funds to work on this. My next hardware based steps in this project are to try to intercept and confirm the protocol commands posted above. If I am unable to get my newest unit to read a CD and read the commands, then reverse engineering the rom is the next alternative. With the interest that has been shown from other car communities in this post, I think writing up a procedure to reverse engineer NEC 75x and 75xx devices would help out a lot of people trying to keep their cars on the road.

I've got a bunch of work from here until October, so unless anything significant updates happen, I'll try to get back to this project then.

I know this is a post that just won't die, but I'm pretty sure it'll be really cool once we get it working.