5VCPC

[Bryce]

Hi all,
     I thought I'd quickly introduce my latest project. Below is a picture of a very early prototype I built tonight. As the title implies I've called it the 5VCPC and that's more or less what it does. It's more of a solution to an issue rather than an expansion. The device solves two issues by doing one thing - creating a 12V supply from the 5V rail, so:

1) It allows a Classic CPC6128 to work without needing a 12V supply. Neat desk, no pigtail hanging out of the CPC, only one PSU.

2) It also allows you to use a standard 5V/12V PC drive or CPC 3in drive in a CPC6128+ without having to add a 12V PSU.

The plan is that it can be installed inside the CPC without any soldering or even needing any fixings. The PCB plugs directly onto the back of the disk drive and the normal power connector to the drive plugs into the 5VCPC PCB. This first prototype was to test dimensions and confirm the function, however, I messed up on the dimensions, so it's slightly too tall and doesn't fit in the CPC with the top in place, but that will be sorted on the next prototypes. This time around I just want to confirm that it works and test:

- How much power it pulls from the 5V rail.
- How hot it gets.
- That it doesn't disturb the 5V rail / make the CPC crash.
- That it doesn't transmit all sorts of shite that makes everything in the room crash or cause distortions on the CRT / Display.
- That the drive actually works when powered from it.
- The circuits efficiency.

I'll be doing the first tests this weekend and when it passes everything, I'll be sending one to Villain who has offered to use it for real and give further feedback from real use.

Bryce.

[||C|-|E||]

Very useful device. I will be getting one or two of the final version, probably two 

[robcfg]

That is awesome indeed! Very useful specially if you connect your CPC, say, to a 25" Sony Trinitron CRT [emoji4]


Sent from my iPhone using Tapatalk

[TotO]

I think that peoples using a CTM will have to continue to use +12V to avoid dim light when to FDD motor start.
But it is nice with an external +5V power supply instead of a dual voltage external PSU. (like provided with 3"1/2 HDD case)

[Bryce]

What 3.5in drive do you mean? Or did you mean the standard 3in plus drive?

A standard CPC6128 needs 960mA on the 5V rail in idle with no expansions attached. The 5VCPC will need about 700mA extra, so almost 1.7A for an idle machine. So the PSU will of course need to be powerful enough. 3A would be a good number to start with.
However, regarding the dimming of the screen. This may not be an issue of not enough amps. It could also be a problem with the power rail design not being able to react fast enough. ie: The PSU might have enough amps, but can't adapt fast enough to a sudden rise in required current. These are all things that I will be looking at before I ever release any final product. There are ways of improving or even completely removing this problem.

Bryce.

[TotO]

I have seen that on CPC using a 5V "Y" cable to power an external 3"1/2 PC floppy drive or with a hacked internal drive.
But I think it was mainly because the CTM was tired... I have fixed the problem with a 12V to 5V converter for the floppy drive.

[Bryce]

I'm not sure I fully understand you. Almost all PC 3.5in drives use the 12V rail for it's motors.

Bryce.

[TotO]

As I know, most PC floppy drives are +5V only. 

[PulkoMandy]

I never seen a PC drive using the +12V, either. Only old ones found in Amiga, Atari or Thomson machines.

And yes, the result when using those powered from the CTM +5V supply is that the CTM will be darker when the motor is spinning. Probably not a good idea to do so.

[Bryce]

That must be a French thing. I have 5 PC drives here and all of them use the 12V rail. Either way, as I said, the screen going dark during disk access is due to bad design, I will make sure my device has been tested for this.

Bryce.

[00WReX]

That is interesting, as I have a pile of 10, 3.5" drives from various manufacturers (Mitsumi, Sony, Panasonic...), and different years of manufacture, from mid 90's to mid 2000's and they are all 5v only.

Cheers,
Shane

[Bryce]

I'll check them all again later, but I'm pretty sure they are all 12V/5V. They are of course 720 drives, none of this fancy modern 1.44 gadgetry!
If they're not as common as I thought, then I only need to include one header on the PCB and go back to it only being for the Classic CPC.

Bryce.

[Gryzor]

This is really interesting and useful if it turns out it works (and if it does the question will be, of course, why didn't Amstrad do it in the first place?)

[CanonMan]

(and if it does the question will be, of course, why didn't Amstrad do it in the first place?)

Cost issues, of course!


The bottom line was the most important thing at Amstrad 

[Gryzor]

Yeah, it was mostly rhetorical

[PulkoMandy]

They did something similar, but decided to put it in the CTM rather than the CPC. For several possible reasons:


- Cost saving: having the 5V and 12V power supplies at the same place means they can share some hardware
- Lack of miniaturization: power supplies today are much smaller than they were in the 80s
- Power dissipation: anyone who saw a Thomson TO7 with the huge heatsink on the back knows what I mean

[1024MAK]

They did something similar, but decided to put it in the CTM rather than the CPC. For several possible reasons:


- Cost saving: having the 5V and 12V power supplies at the same place means they can share some hardware
- Lack of miniaturization: power supplies today are much smaller than they were in the 80s
- Power dissipation: anyone who saw a Thomson TO7 with the huge heatsink on the back knows what I mean

Err, yes and no.
The CRT monitor already includes a type of switch mode power supply. So why have another for the computer?
The technology did exist, yes, it would have been larger than today's technology. But there was space in the case if Amstrad had wanted to do it that way. Remember, the Acorn Electon has a low voltage switch mode power supply inside the computer case.

Mark

[Bryce]

There's no DC/DC converter in the CTM, they existed at the time, but they were extremely new (invented in 1977) and there were no dedicated ICs available, it would have had to be done with discrete components. That doesn't mean it would have been out of the question. The humble Spectrum 48K has a DC/DC Booster circuit to provide 12V from the 9V supply.

What always puzzled me, is why they didn't just have a 12V connection from the monitor and derive the 5V from that. It would be the identical components that they used in the monitor, just on a different PCB and they would have saved a cable and connector = cheaper.

Bryce.

[arnoldemu]

You're thinking about the CTM644.. what about the CTM640?

CTM644 is still useable on a 464 and because that never needed 12V I guess they added it as a second cable to keep compatibility?

[ikonsgr]

I'm not sure I fully understand you. Almost all PC 3.5in drives use the 12V rail for it's motors.Bryce.

Nice project Bryce, but i find you a bit "uninformed", regarding the power supply of 3.5" PC floppy drives   
For 20+ years now, ALL 3.5" floppy disk drives use ONLY 5volt power supply, for the board, the step motor and the plateau motor (btw, remember that 6128+ introduced on 1990, had even his 3" disk drive, with 5volt supply only?). I happen to know this because i've tested dozens of different 3.5" disk drives (some where 15-20 years old), and haven't found a single one that needed 12volt supply too. And that's why the Y power supply cable i make for powering Amstrad and 3.5" disk drive gives only 5volt( and the fact that after selling 100's of them, no one ever complained for a disk drive not working seems to "confirm" that  )  . I think that only VERY old 3.5" disk drives (from early 90's or older) might need 12volt supply too, but i doubt you can find many owners of such 3.5" disk drives any more....
Having said that, i wonder, how old might be the 3.5" disk drives you said you have, and need 12volt supply too? 

[1024MAK]

I've still got some 3½ drives with 12V motors. They are all single or double sided DD drives. I have not checked all my HD FDDs, but they may well all be 5V only.

Mark

[ikonsgr]

Single or DD drives you said? I suppose from mid 80's or something, right? 
And yes, you are probably right that perhaps all HD drives need  5volt only, as they were more recent than DD drives.
IMO, i tend to believe that almost all 3.5" disk drives manufactured after ~1990 needed 5volt only. Even with amstrad 3"  disk drives,i have noticed that early models (from 1985) had 12volt step and plateau motors but later 6128 models had 3" drives equipped with 5volt step motors instead! And also, don't forget that 6128+ released in 1990, had even his 3" drive with 5volt supply only.

[1024MAK]

The drives that I speak about are either ex-Atari ST drives, ex-Amiga drives or were obtained between 1988 and 1993 approximately. Note most were not retail "new". Some came from surplus electrical companies. Then there was a gap when I did not buy any. Then found that DD drives were no longer available, so got HD drives.

Mark

[Bryce]

I went back and checked the drives. 1 of them is a 5V only drive, the other 4 are all 5V/12V drives. They all came out of old PC's from the 80's that I dismantled a few years ago. I also found 3 other drives from more modern times and they are all 5V only drives, so there seems to be a cut-off date when drives went from being 12V/5V to being 5V only.
So yes, @ikongrs: seems like I am a little uninformed when it comes to 3.5in drives, I didn't know that drives became 5V only at some time, but then again, I don't have a lot to do with 3.5in drives and I don't think I've ever bothered fixing one. I just throw them away and get another one.
I'll change my 5VCPC schematic and layout to only handle 3in drives.

Bryce.

[Bryce]

A quick update on my 5VCPC project. I finally got around to finishing all the tests and here's the results:

DC/DC converter 5V side:
Uses 11mA with no load
Uses 119mA with disk drive in idle (an idle disk drive pulls 39mA from the 12V supply)
Reading/Writing to a disk requires 313mA (with a very old drive with dodgy motor it can have peaks of up to 500mA)

12V Side:
The normal 12V disk drive supply uses 39mA in idle and uses between 110mA up to 200mA on older worn drives.

The DC/DC converter is running at around 87% efficiency. I've designed it to supply up to 400mA so technically it would be able to supply two drives at the same time. It's only disappating a few mW, so even with constant usage it will never get even slightly warm.
Regarding the potential EMC/EMF issues mentioned above: No, it doesn't interfere with the CPC in any way, no crashes, no issues and no screen flickering during loading. All real-world tests were done with a Spectrum +3 2A 5V supply. The standard 6128 needs around 1.1Amps for the computer alone and with my DC/DC converter you should allow for an extra 500mA for the DC/DC 12V rail, so a 1.6Amp supply would be enough to power a 6128 with no further expansions.

Next steps: Despite my fancy design, I will need to change the layout and move away from the idea of mounting it on the back of the drive. It's very wobbly and there's a risk of shorting out against the metal drive parts. So the final design will be a small PCB with a flylead to the drive.

Bryce.


P.s. If anyone is interested in converting their 6128/664 to a 5V only device, then let me know here so that I can judge how many PCBs I should make in the initial batch.