MPC 4000 CPU's

I also have a source on the Intel Sharc, which actually is one family above the OEM chip used.

The origional chip is available in Australia, and would run about $24 on a bulk purchase of say 100 pieces.

The next step up is available in the US, and could be as cheap as $11 each.

I need a donor board (One that has a bad CPU) so I can test the newer Sharc

If anyone has a donor board, I will test it, and if it works, it would even be an upgrade, let alone a fix. Donor board would be returned either way.

By the way, the CPU replacement/Upgrade would include the following:

Remove old CPU
Install IC Socket (So the new CPU just slides into the socket, no soldering)
Installation of CPU into socket
Extra CPU for a back up (Self installable if you ever needed it)

Would this mean more power?

Yes it very likely would. Substantial increase IF the bus doesn't limit it.

Once I actually get to this I will find out for sure.

Here's the theory:

The mainboard isn't too far off from that of a normal PC motherboard. There are certain components, however, that may limit the entire bus speed, in which case, it (the 'cpu', nothing more than a microprocessor) would be limited by the 'slowest' 'component' in the bus.

I'm in the process of opening 2 more studios so I'm a little swamped

We need AKAI to release the OS.

i am dying to know this information, but dont have an extra board available.

if succeeding this means a 4k on steroids??????

yeah but come on, how many pins does one of those CPUs have?

It's probably dozen and dozens.

I very much doubt it could be done at home with a soldering iron! This is the sort of thing that needs factory
equipment to solder.

Sound's good. I think all that the forum needs is proof of it's reliability and viability and no doubt, given the already existing demand for more power in the 4000 and access to the code (for some emendations and updates), You know you'd have a fresh batch of clients, SMC. We all know that the 4000 is by far, the most intuitive drum machine/sampler/seqeuncer/synth/controller ever put into mass production and would love to see it through to it's logical contemporary conclusion. I don't know bout' ery bodee else but I'm a little tired of the proposals alone, no disrespect intended of course. My advice (although, in all fairness, you never asked for my advise-lol!) would be this : Get a working model of the 4000 with the 'Super Harvard....Blah, blah,blah.....' chip running in it. Burn some tests and results of the functionality under duration and heavy tasking. Give us an idea of the practical labor and costs involved to get it done right and hit us with the product, in hand, so to speak and I'm sure most of us will make the move. We've heard a lot of 'Bernie Madoff-like' Ideas on the forum. Some 'Reality MPC' would be in order right now. If you can substantiate the value of getting the forum involved in all of this, then make the move. Seriously, SMC, just get that product and it's all good, I'm down (within reason, of course). It's a good gesture though. Get at us with that .... Reub.

ReuBen KinKade wrote:Sound's good. I think all that the forum needs is proof of it's reliability and viability and no doubt, given the already existing demand for more power in the 4000 and access to the code (for some emendations and updates), You know you'd have a fresh batch of clients, SMC. We all know that the 4000 is by far, the most intuitive drum machine/sampler/seqeuncer/synth/controller ever put into mass production and would love to see it through to it's logical contemporary conclusion. I don't know bout' ery bodee else but I'm a little tired of the proposals alone, no disrespect intended of course. My advice (although, in all fairness, you never asked for my advise-lol!) would be this : Get a working model of the 4000 with the 'Super Harvard....Blah, blah,blah.....' chip running in it. Burn some tests and results of the functionality under duration and heavy tasking. Give us an idea of the practical labor and costs involved to get it done right and hit us with the product, in hand, so to speak and I'm sure most of us will make the move. We've heard a lot of 'Bernie Madoff-like' Ideas on the forum. Some 'Reality MPC' would be in order right now. If you can substantiate the value of getting the forum involved in all of this, then make the move. Seriously, SMC, just get that product and it's all good, I'm down (within reason, of course). It's a good gesture though. Get at us with that .... Reub.

you re right

feline1 wrote:yeah but come on, how many pins does one of those CPUs have?

It's probably dozen and dozens.

I very much doubt it could be done at home with a soldering iron! This is the sort of thing that needs factory
equipment to solder.

Yes, there's a lot. The trick is the IC soldering tool.

The one I have access to is a modular tool with a suction cup and a heated element surrounding it, that when in place, drops down and contacts ALL the pins on the microprocessor (Which is pretty standard configuration for IC's)

So, yes, I can do it 'at home' (Hey, check the 4K fix thread, I responded to help you with the manual thing)

Reub wrote:Sound's good. I think all that the forum needs is proof of it's reliability and viability and no doubt, given the already existing demand for more power in the 4000 and access to the code (for some emendations and updates), You know you'd have a fresh batch of clients, SMC. We all know that the 4000 is by far, the most intuitive drum machine/sampler/seqeuncer/synth/controller ever put into mass production and would love to see it through to it's logical contemporary conclusion. I don't know bout' ery bodee else but I'm a little tired of the proposals alone, no disrespect intended of course. My advice (although, in all fairness, you never asked for my advise-lol!) would be this : Get a working model of the 4000 with the 'Super Harvard....Blah, blah,blah.....' chip running in it. Burn some tests and results of the functionality under duration and heavy tasking. Give us an idea of the practical labor and costs involved to get it done right and hit us with the product, in hand, so to speak and I'm sure most of us will make the move. We've heard a lot of 'Bernie Madoff-like' Ideas on the forum. Some 'Reality MPC' would be in order right now. If you can substantiate the value of getting the forum involved in all of this, then make the move. Seriously, SMC, just get that product and it's all good, I'm down (within reason, of course). It's a good gesture though. Get at us with that .... Reub..

Sorry, I've been hella busy opening a second and third studio, 2 new HD rigs and all, buying a house, and modding my Control 24's (2) for all blue LED's, (EVEN METERS!!! IF POSSIBLE, not there yet) Modded Pre's, and all new Penny and Giles faders.

As far as this project goes, it is VERY feasible. I just need a board with a bad CPU, or a good board, as a donor.

I'll buy one, just haven't been able to find one. The other thing is, if it's bad I have to know it's the CPU so I don't have to spend countless hours finding another problem.

I have found a few, but with no guarantee that it was the CPU, and little spendy considering the unknowns.

Anyway. I would solder in a SOCKET to hold the replacement IC Microprocessor chip. So future replacements or possible upgrades could be done at home. This is a simple task for me given I have access to the equipment to do so.

Each mod would be accompanied by a second spare processor just in case (They are cheap)

Also, this would improve other things as well, such as heat dissipation, as there will be a nice space between the CPU and the Circuit Board. Less chance of overheating.

I was even thinking of going as far as tying in a small CPU fan to the existing case fan, just have to figure out how to mount it Just have to figure out how to get something in there for that. That'd be the last concern.

I don't screw around, I have a nicely designed Compact Flash mod, brand new card reader for the 2000XL, and am going to make a new bracket so it can mount in the 4K as well, I just need to crack mine open to test it. Should work, if there's not a seperate header it should work with a dual contact IDE ribbon steming off the hard drive, it's just not hot swappable is all. Might be handy anyway.

Also, I have some other things about to hit the market as well, just ironing out the patent end of things.



Help me help us all, I just need to find a donor board.

SMC wrote:I also have a source on the Intel Sharc, which actually is one family above the OEM chip used.

Intel made a Sharc ( news to me although I am prepared to be enlightened )

Intel made a StrongArm ...

Analog Devices made a Sharc ...

These CPU's are not even remotely similar.

The MPC4000 uses an Intel StrongArm.

Rohan.

mpc3000le wrote:

SMC wrote:I also have a source on the Intel Sharc, which actually is one family above the OEM chip used.

Intel made a Sharc ( news to me although I am prepared to be enlightened )

Intel made a StrongArm ...

Analog Devices made a Sharc ...

These CPU's are not even remotely similar.

The MPC4000 uses an Intel StrongArm.

Rohan.

My bad
working on some other mods using Sharc's

Yes, the 4K is a 110 family strongarm, and actually there are at least two others in the line which according to datasheets and tech docs should be compatible.

Sounds interesting man...anything that will keep my 4k running longer

keep us posted.

Well, VST has a couple CPU boards, but they can't confirm the actual problem with them.

Pretty cheap, but I don't want to spend a ton of time tryin to track down another problem....

Might just bite the bullet and grab one..... Just creates more work.... I have limited time.

I'd really rather find one that is confirmed as a bad CPU, or at least a freebie.... tracking down problems on circuit boards can be a real pain if you don't have the ARC's software/hardware......

You could always try and convince VST that theres a potential business for them. If you actually would succeed in doing this they could sell it as an upgrade for existing 4k users. That should be enough for them to throw you a couple free faulty boards.