Diodes on SMPSU module outputs (FIRE RISK).

Help & news on accelerators from TF, Amiga, Atari, CD32 etc

Moderators: terriblefire, Terriblefire Moderator

terriblefire
Moderator Team
Moderator Team
Posts: 5368
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by terriblefire »

In the motorola 68060 manual. The problem statement...
11.2.1.2 INPUT SIGNALS DURING POWER-UP REQUIREMENT. The second issue
involves the requirement that during power-up, input signals to the MC68060 not exceed Vdd
by more than 4 V. This is achieved by ensuring that the 5-V supply not exceed the 3.3-V
supply by more than 4 V.
This is exactly the same as the XIlinx and will be for the same reasons.. As stated above ... in order for the the SMPSU output to be less than 1V when the 5V rail is completely up would need a startup current of 4amps or higher (without sagging the 5V line). So the diodes are completely unnecessary on an Amiga.

Even then you will be fine provided there is at least a 10 ohm input resistor in series per IO to limit the current draw and prevent damage. Just like the Xilinx.

The proposed example solution by motorola...
In any of the previously discussed DC-to-DC conversion solutions,
it is possible to add three diodes in series from the 5-V supply to the 3.3-V plane. During
power-up, the diodes forward bias and thus provide a current path between the 5-V source
and the 3.3-V plane. This solution provides no more than (0.7 * 3) = 2.1 V drop between the
5-V input and the 3.3-V plane. When the voltage regulator stabilizes, the difference of (5 –
3.3) = 1.7 V is insufficient to forward bias the three diodes, hence not dissipating any energy.
Both Motorola and Linear Technologies have indicated that the three diode shunt does not
adversely affect operation.
I disagree with this solution and i think i was cobbled together in a hurry. I think it's something that has been thrown into the docs without enough thought and are better solutions than this. It did not address the issue of the power rail being shorted or the safety needed in that instance.

Break a problem down and separate it from the solution. Decide is it actually a problem in your use case window. If you are building a VME machine with a 10 amp backplane power rail then yeah it could be an issue. If you are putting an accelerator in an Amiga which has a Choke on the input and caps that will take time to charge up then probably not. Don't be fooled into thinking that because X is true and Y follows right after it Y is also true.

All that Motorola has said here is that "it is possible to add three diodes in series from the 5-V supply to the 3.3-V plane". They haven't considered the wider implications of doing this in a consumer piece of electronics where users might accidentally short things.

The actual problem here is we cannot let input signals to the 060 be more than 4V above VDD. Well i'm telling you right now there is always a drop between the motherboard and the accelerator. Firstly its rare to see an Amiga running at true 5v. They're usually around 4.8v. Then there are drops across series resistors and contact resistance drops at the socket. So even if the PSU is missing it probably wont be much more than 4v difference.

Sure you can get guys like Jens who think the chip will explode if you go 0.1v over.. but thats not going to happen and honestly... in terms of outcomes. A flaming house or needing to replace an 060.. its not much of a choice is it?
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
chucky
Posts: 95
Joined: Sun Jan 13, 2019 8:42 pm

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by chucky »

In short: the risc of damages will happen with 5V on IO before 3.3V is up, is very VERY small.

compared to the danger of components going ABOVE 150degrees (as that's where my FLIR camera limits to) and I had it on like that for a LONG time to see when itwould fail at all, did not happen even in 1-2 minutes.. just smoke...

so the diodes on this DCDC card is simply put not good to have there. it is safer to have them totally removed. there is a extreme minimal risc of the cplds to die MAYBE.. a small price to pay
User avatar
rubber_jonnie
Site Admin
Site Admin
Posts: 10475
Joined: Thu Aug 17, 2017 7:40 pm
Location: Essex
Contact:

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by rubber_jonnie »

terriblefire wrote: Wed Feb 10, 2021 9:42 pm
adam wrote: Wed Feb 10, 2021 9:40 pm This topic begs for a stream:) just saying...
I am considering coming out of YT retirement to set something on fire.

EDIT: But lab is not what it used to be due to house sale
That's certainly a video i'd watch :)
Collector of many retro things!
800XL and 65XE both with Ultimate1MB,VBXL/XE & PokeyMax, SIDE3, SDrive Max, 2x 1010 cassette, 2x 1050 one with Happy mod, 3x 2600 Jr, 7800 and Lynx II
Approx 20 STs, including a 520 STM, 520 STFMs, 3x Mega ST, MSTE & 2x 32 Mhz boosted STEs
Plus the rest, totalling around 50 machines including a QL, 3x BBC Model B, Electron, Spectrums, ZX81 etc...
User avatar
stephen_usher
Posts: 5580
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by stephen_usher »

Sorry if I'm being thick but surely the most reliable way to make sure that the 5V rail doesn't go high before the 3.3V one is to make the 3.3V rail switch the 5V connection to the processor on once it reaches a certain voltage?
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
terriblefire
Moderator Team
Moderator Team
Posts: 5368
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by terriblefire »

stephen_usher wrote: Thu Feb 11, 2021 8:29 am Sorry if I'm being thick but surely the most reliable way to make sure that the 5V rail doesn't go high before the 3.3V one is to make the 3.3V rail switch the 5V connection to the processor on once it reaches a certain voltage?
The processor is 3.3v. Signals from the Amiga motherboard are 5v. The amiga motherboard comes up before accelerator since its powering the accelerator.

EDIT: The processor is 3.3v with 5v tolerant IO. But its only 5v tolerant (in theory) when the input signals are less than 4V higher than the PSU rail. In practice its much more forgiving than that.

EDIT2: My personal take on this is that those diodes are probably needed if you are designing VMEBUS machine which supports a hot swappable cards. Which is something Motorola was doing in the 90s. The backplanes on those machines had really heavy duty drivers so the data signals would always be up at a full 5v. Not the limp wristed drive strengths of the amiga chipset. Those machines lived in datacenters with freon fire suppression too.
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Higgy
Posts: 400
Joined: Tue Apr 23, 2019 8:05 pm
Location: Somerset

Re: Diodes on SMPSU module outputs (FIRE RISK).

Post by Higgy »

Is there any risk on the TF360? Do I need to mod the voltage reg circuit I have?
I have an IDE ribbon cable draped over the reg part and there are no heat marks I can see, so hopefully all is good.
Thanks.
Post Reply

Return to “Terriblefire's channel”