A 'once and for all' declaration regarding L88M mask MC68040 CPUs

[VWGuy]

I'm just looking for a later .57 um mask at this point. The seller refunded my fake with no hesitation. They know what they are doing and don't care.

 

[jessenator]

I simply don't want the thread to derail. I want to keep this, as much as possible, about the thermal properties and debunking the bull$#!@ claims that have been parroted since time immem—no maybe the last 20-25 years or whatever the window is. There is a dedicated thread to fake chip ID, well, specifically 040s in this case… but at any rate, (clear) photos of fakes are welcome there for sure.

I know I sound like a broken record, but what I want is thermal imaging evidence/record of the properties of 040 packages, L88M most of all at this point.

From what I've seen and tested, minus the effects of the well-documented XC errata, is there any point in fawning over the L88M when another mask revision will perform just as well? I would say no. An E42K will perform just as well. Perhaps in situations where heatsinks aren't used there will (definitely) be issues, but IMO it's just flexing and !@#$ measuring at this point. Finding a genuine non-LC 040 in itself is a win anymore. With a proper heatsink, most that I've tested can go to 40MHz without adverse effects thermally. I think what I might do, if ever spare time wills itself into being, is do general system stability tests (in another thread).

Also, VWGuy, while warning about things like this is important, something that I find is valuable, it's a gray area as far as forum rules are concerned. While in the trading post section, it sure should be globally applicable, basically it's that we don't engage in ebay seller/buyer feedback or commentary—unless that rule has changed in my absence Anyway, I'm not a mod, but it's one of my neuroses, so apologies for calling out. It might be something to consider in the improvement/topics thread coryw set up.

 

[VWGuy]

What kind of heat sink are you running during these tests? My current chip (early 25 mhz) runs hot enough I can't barely touch the sink.

 

[rplacd]

Last I checked, Herb was out of stock.

If anyone has a spare they'd kindly let go? I am doing a scratch built Amiga 4000 from bare motherboard and accelerator card.

Herb just restocked! I purchased one... although there isn't many, Herb had to get the pins repaired. Hence why I posted here. Herb's an amazing detective for this stuff.

 

[rplacd]

Finding a genuine non-LC 040 in itself is a win anymore. With a proper heatsink, most that I've tested can go to 40MHz without adverse effects thermally. I think what I might do, if ever spare time wills itself into being, is do general system stability tests (in another thread).

I would love to hear your advice on proper heatsinks! I have an (evidently) non-genuine '040/40 that's unstable at 40 MHz. It reports to have both FPU and MMU, but I've stuck a cheap Amazon heatsink and it overheats and barfs. Would you have a link to a proper heatsink?

 

[David Cook]

I would love to hear your advice on proper heatsinks! I have an (evidently) non-genuine '040/40 that's unstable at 40 MHz. It reports to have both FPU and MMU, but I've stuck a cheap Amazon heatsink and it overheats and barfs. Would you have a link to a proper heatsink?

Double-check that the heatsink's adhesion is actually thermal. I've received heatsinks from Amazon sellers with ordinary double-sided sticky tape as well as 3M coated tissue tape (not thermal tape). That will act as an insulator, which is the opposite of what you want.



In the picture above, I peeled off the tissue tape, cleaned the part with isopropyl alcohol, and applied a ThermalRight thermal pad.

 

[rplacd]

Double-check that the heatsink's adhesion is actually thermal. I've received heatsinks from Amazon sellers with ordinary double-sided sticky tape as well as 3M coated tissue tape (not thermal tape). That will act as an insulator, which is the opposite of what you want.

View attachment 41870

In the picture above, I peeled off the tissue tape, cleaned the part with isopropyl alcohol, and applied a ThermalRight thermal pad.

Hey, thanks for the recommendation! I've just looked it up, and I've just noticed that you can get this stuff in multiple thicknesses. Would you have any recommendations/knowledge on why this is the case/what to get?

 

[jessenator]

What kind of heat sink are you running during these tests? My current chip (early 25 mhz) runs hot enough I can't barely touch the sink.

All the tests were with no heatsink at all.

This is not aimed at you, but you bring up a good point:
The myth-to-be-busted "cool to the touch", by implication is "it doesn't need a heatsink." If "cool to the touch" was intended to be applied to a thermal-managed CPU, it would be a moot point, IMO. Get a larger, more efficient heatsink: problem solved. (A resourceful individual could certainly find something even 25 years ago.) The thermal solution …solution isn't a qualifier here; the original intent of the saying was "this [L88M] is the coolest-running 040 there ever was; you could run it without a heatsink and it would run cool to the touch." regardless of over/under- clocking. If thermal management was a qualifier, I could just rig up a liquid cooling solution, or even some LN2 and say that even my D50D runs cooler than an L88M, and I'd be objectively correct!

Have we proved the first point? Yes, you can certainly run certain 040s (from the Wombat onward) without a heatsink and unless ambient temp is quite high (my tests it was in the 20s C somewhere), the package temps, save one, never reached 100C which was the Motorola thermal design test. In the one test further back in the thread, I cut it off before it hit 90C, and the D50D I couldn't get higher, because it refused to post after a certain clock speed.

Have we proved the second point? Absolutely not.


I hope I don't seem like I'm getting heated here. I'm not trying to call anyone out, either, I'm simply trying to stick to the premise, because I'm ruddy great pedant!

Do I think heatsinks are a good idea? Abso-FRIGGIN-lutely. Whatever fits in for the final application, use it! I just hold that heatsinks have nothing to do with the myth to be busted.

 

[MrFahrenheit]

For heat sinks for these, I do a border of thermal tape, avoid 1/4 inch, all the way around. Then I apply a pea size of thermal paste, like Noctua, in the center and the thermal tape holds the heat sink down.

By using only a border of tape, it can be removed if necessary, easily, and yet it holds down securely.

 

[VWGuy]

Good advise on the tape and paste. I like that method.

 

[MikeInMass]

Here's an eBay seller in China offering chips with mask 02E31F. Do these look legit?

Motorola 68040 XC68040RC33M MC68040RC 02E31F 33MHz Vintage CPU PGA | eBay

Find many great new & used options and get the best deals for Motorola 68040 XC68040RC33M MC68040RC 02E31F 33MHz Vintage CPU PGA at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[Phipli]

Here's an eBay seller in China offering chips with mask 02E31F. Do these look legit?

Motorola 68040 XC68040RC33M MC68040RC 02E31F 33MHz Vintage CPU PGA | eBay

Find many great new & used options and get the best deals for Motorola 68040 XC68040RC33M MC68040RC 02E31F 33MHz Vintage CPU PGA at the best online prices at eBay! Free shipping for many products!

www.ebay.com

It's impossible to tell, because there is no saying you'll get the one pictured.

 

[jessenator]

It's impossible to tell, because there is no saying you'll get the one pictured.

This… Also the shipping times are so extreme sometimes that you'll miss the feedback window (and sometimes the return window). Not talking about that seller specifically, just in general.

 

[Paralel]

I don't mean to go potentially "offtopic" on this thread, but I am curious, does anyone known if Moto (or its successors) ever made a full 68040 40 MHz QFP CPU, like the 33 MHz used on the 550c blackbird powerbook CPU daughtercards, rather than PGA?

 

[jessenator]

offtopic

I need to self-crit on this. I'm making strides to move beyond my weird obsessions. I honestly, now, couldn't care less I've apologized privately to the individual I went full passive-agressive on regarding the issue (which is/was my issue), and I don't want to let something that is made up of a load of pixels and energy get the best of me. </confession>

I haven't seen any QPFs beyond embedded iterations of the 040—but I am far from the expert there. Found this in the NXP documentation:

So unless it's 3.3V I don't think they did, sadly :/

 

[MrFahrenheit]

I compiled this chart last year:

 

[jessenator]

Okay, here's something weird I'm trying to understand:

MOTOROLA M68040 USER’S MANUAL C-1
APPENDIX C
MC68040V AND MC68EC040V
The MC68040V and MC68EC040V are Motorola’s 3.3 volt, static versions of the MC68040 third-generation, M68000-compatible, high-performance, 32-bit microprocessor. They require a 3.3V power supply providing over 50 percent reduction in power consumption compared to a 5.0V device. The maximum power used at 3.3 volts is 1.5 watts at an operating frequency of 33 MHz. They also have a low-power stop mode. Once in this state, both devices remain quiescent, consuming less than 330 μW of power. The low-power usage of these microprocessors makes them an ideal choice for portable computing and power constrained applications.

The MC68040V programming model, data formats and types, instruction set, caches, and MMUs are the same as those described for the MC68LC040 in Appendix A MC68LC040. The MC68EC040V programming model, data formats and types, and instruction set are the same as those described for the MC68EC040 in Appendix B MC68EC040. However, both devices contain additional features:
• For the MC68040V, all differences that exist between the MC68LC040 and the MC68040, as described in Appendix A MC68LC040, also apply to the MC68040V. For the MC68EC040V, all differences that exist between the MC68EC040 and the
MC68040, as described in Appendix B MC68EC040, also apply to the MC68EC040V.
• Both devices operate to 0 Hz and can accept 3.3V or 5V input.
• Both devices have a new processor status state, low-power stop mode, indicated when PST(3–0) = $6.
• There is no PCLK or TRST pin on either device.
• Both devices provide three new pins, system clock disable (SCD), low frequency operation (LFO), and loss of clock (LOC)

so the V versions can do either 3.3V or 5V? I was told they could only do 3.3V, which is why people weren't using them. Sorry if I've been parroting more anecdotal BS by saying they can't. Or maybe I'm reading that wrong, but the verbiage leads me to believe that they can indeed run with either 3.3V or 5V.

When looking at https://www.digikey.com/en/products/detail/nxp-usa-inc/MC68040FE40V/1167311 even though it links to the same NXP document, it shows 5.0V in the quick specs table.

 

[treellama]

That means 3.3V supply, but accepts 3.3V or 5V inputs. See table C.7.1 in the datasheet addendum. They cannot run on more than 3.6V supply.

 

[Fizzbinn]

Above my knowledge really but aren’t there ways of using 3v chips in 5v circuits, interposer boards/components that adapt the 3v volt part to work?

I’m thinking Power Mac l2 cache cards where earlier versions are 5v and later ones are 3v with extra components that make it work?

 

[Paralel]

Just for this information to be recorded somewhere, here is the data from the Powerbook 550c CPU's of which I am aware, including my own.

All of them are Motorola XC68040FE33M QFP184

All of them have the same mask & location of production: 02E31F Malaysia

Part Numbers: QEES9519A, QEEB9517B, QEBP9511A, QEBQ9511A