Author Topic: How well would Floppy Disk film work as a micro-core transformer?  (Read 1507 times)

0 Members and 1 Guest are viewing this topic.

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Those of you well versed in the area of solid state magnetic technology might be able to piece together what I'm trying to accomplish, but I don't wanna make it too painfully obvious. ;)

I'm wondering if it's possible to make basically a microscopic iron-core toroidal transformer like so that can retain it's magnetized state without power:


I can't afford to buy a gazillion iron cores, but I do have a ton of (probably 300 or so) unused trash-picked floppy disks stuffed under my bed and am wondering if the film for those would work for what I need.
or would the film be too restrictive to carry a proper magnetic current??

thanks :)
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #1 on: December 30, 2019, 03:44:12 am »
I mean, that's what disks are, they're core memory where the core spins past the winding(s).  Can't you just use it as-is? :P

Good luck making discrete cores out of film, the layer is quite thin so mu_r and Ae will be very low.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #2 on: December 30, 2019, 04:05:20 am »
I mean, that's what disks are, they're core memory where the core spins past the winding(s).  Can't you just use it as-is? :P

that's the issue though, they need to be moving in order to even be charged...
at least I think anyways >.>
it's that motion that can be read.

in my case, I'm looking to induce a signal in the primary coil
and depending on the state of the toroid
if it works like I'd like it to
should induce a signal in the secondary coil on state change

but if it doesn't work like I'd like it to, it won't carry the field over to the secondary coil
if that's the case I may need to wind the primary coil over the secondary coil, and see if I can get a voltage spike on state change from a field collapse.

I don't have an oscilloscope to be able to test this
sure I have a linux PC, but I don't feel like DIYing a proper probe atm to plug into my sound card :P

EDIT:
note the test shouldn't be an immediate polarity switch

start with +- to set the primary coil with
then disable power
then enable power with -+ and see if you get a spike on the secondary coil

that is, if you have a floppy to rip apart anyways :P
« Last Edit: December 30, 2019, 04:14:12 am by Tcll »
 

Offline Alex Eisenhut

  • Super Contributor
  • ***
  • Posts: 3550
  • Country: ca
  • Place text here.
Hoarder of 8-bit Commodore relics and 1960s Tektronix 500-series stuff. Unconventional interior decorator.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #4 on: December 30, 2019, 05:03:08 am »
You can still read floppy bits statically, you just have to separate the self-inductance from the media inductance.  You wouldn't be able to use existing read amplifiers, of course.

Hmm, I wonder just how small of a difference that is.  It might be 1% or less, which would be difficult to resolve on an oscilloscope.

Anyway, what you would see is, when you apply a slow pulse (more of a ramp) to the head, there is a corresponding voltage pulse, due to self-inductance.  V = L dI/dt and all that.  Somewhere along that slope, the bit flips (or not), and an extra hump of voltage (flux actually, so volts*time) is generated.  You might put the head in a little H-bridge circuit (not much current is required, an analog switch I think would be okay?), so its polarity can be reversed easily, and that way you don't have to worry about positive/negative pulses being asymmetric, all your (new custom) R/W circuitry is consistent and repeatable.

You'd (obviously?) have a hard time also using the spindle motor as an indexer, so you'd need to replace that with a stepper to make single-bit seeks repeatable.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #5 on: December 30, 2019, 05:10:48 am »
You'd (obviously?) have a hard time also using the spindle motor as an indexer, so you'd need to replace that with a stepper to make single-bit seeks repeatable.
huh?

you're wrapping 2 coils around a small toroid of film (shown in the OP)
there's no motor involved...

I just need to know if the film will carry the field from the primary coil to the secondary coil.

EDIT:
to give you an idea of what I'm trying to achieve
the thickness of the toroid in the image would be the thickness of the floppy film
« Last Edit: December 30, 2019, 05:20:19 am by Tcll »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #6 on: December 30, 2019, 05:36:05 am »
I know what you're talking about, using the film as a core, and I already noted

Good luck making discrete cores out of film, the layer is quite thin so mu_r and Ae will be very low.

You're welcome to try it but I think you will find the geometry in the OP will be ineffective.  Low mu_r gives low coupling, and low Ae gives low core flux.  You may see some effect on self-inductance but it won't be nearly as clean of a "pop" as coercive force is exceeded, because of the nonuniformity of field around the film.  Also, expect to need many turns (or a ton of amperes), not a single turn like classic core memory.  Again, because of mu.

I'm musing on a potentially more practical form, using the original head (which potentially has a tighter magnetic loop to the media).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Tcll

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #7 on: December 30, 2019, 05:56:50 am »
lol I'm sorry, I don't mean to make you repeat yourself
I guess that means it's time for me to call it a night. XD

but yeah I was already expecting multiple windings (not sure how many exactly)
I was thinking 7 for the primary and 4 for the secondary, but I might need even more...
and more amps would probably create heat, which common sense says should probably be avoided ;)

but anyways, if I were going for your design, I'd probably go MRAM style and just have thousands of coils with film sandwiched in between 2 layers of them...
if the bottom coil is the read (secondary) coil, then there might be a noticable difference when the film polarity doesn't match the read op from the write (primary) coil.

but I'm still a noob when it comes to electronics
I'm just trying to do stuff cause I'm sick of waiting for the market to stop selling literal garbage
I'm still running spinning rust because of it (I refuse to use NAND outside of flash drives)

and yes, everything I know so far is all self-taught and figured out
I don't understand acronyms or material behavior properties yet, but I'll get there as I go. ;)
 

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #8 on: December 30, 2019, 01:14:33 pm »
yeah I definitely needed that

in simple terms for future DIYers like me who don't understand technology but know enough to make simple circuit boards,
no it would not work as expected, and would not transfer the magnetic field through the film (coupling and flux)

so yeah, I'm better off with sandwiching the film between 2 flat-coil matrix arrays
thank you for your time and helping me understand :)
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #9 on: December 30, 2019, 02:56:59 pm »
I mean, that makes things worse, the magnetic path length is even longer; and, I don't get the transmit-receive concept, when everything is reciprocal so it doesn't much matter which side you do it on.

Like, core memory uses multiple lines because of addressing, of course, and it's easier to just add one more (sense) line rather than trying to obtain the read from literally every possible row/column.

In some situations, independent windings can be useful to remove self-inductance from the reading (B-H curve plotting for example), but that shouldn't be necessary here at least for a PoC.

But yeah, as a beginner, you'll be quite confused by all this magnetic terminology.  This is something of an introduction/explanation:
https://www.seventransistorlabs.com/tmoranwms/Elec_Magnetics.html#magnet
but it is rather lacking in diagrams and you'll likely do well to find simpler materials.  You'll often see magnetic circuits drawn in terms of electric circuits, with voltage, current and resistance <--> MMF, flux and resistance respectively.  I don't personally find reluctance useful, but the circuit analogy is spot on and anywhere you can draw an accurate circuit and solve it, you have the magnetic solution too.

You'll also often see schematic illustrations of transformers with windings on opposite sides or legs, which is silly for the most part.  But it's hard to appreciate why, until you've figured out why "leakage" is what it is.  For the most part -- in practical transformers -- you want windings in the same place, not in opposite places.  Again, separate windings can be a good idea when you're doing things specifically with the magnetic core, like the B-H curve or core memory.


The design of a magnetic head, is a small "C", with the gap facing the media, and a winding on the back.  This shortens the magnetic path length, so that the media is a large fraction of the total magnetic path.  This solves the mu_r problem.  The fringing field around the gap is similarly small, meaning the field does not extend very far from the head, which may be annoying for arbitrary setups, but when that can be solved (transport system in tapes, "flying" heads in disks) it's rather useful for data storage, keeping the size of a bit domain small.

That's why I suggest sticking with the head, because a century ago they solved all the problems that you're creating for yourself. :P


On a related note, you can write to disk just like you'd write to tape, including with audio (the quality is just terrible).  If you overdrive it, adjacent tracks are written, too, giving a weird... "preverb" effect if you will.




Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: SeanB

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #10 on: December 30, 2019, 03:19:23 pm »
thanks for the resources, I'll definitely need them ;)

but yeah, I'm trying to go full solid state with no moving parts
the problem with old-style weaved core-memory is you can't write to one core without affecting others, making it super slow with all sorts of validation

so yeah, I'm trying to achieve MRAM style isolation with addressable bits
the problem is the average consumer can't go out and buy an MRAM SSD, let alone afford one.
(the benefits of MRAM over NAND is no write limit (endurance), and no validation (speed))
and capitalism doesn't want non-disposable drives.

so yeah, I'm trying to work on my own solution
I'll worry about the encoder once I have a decent array built
all that really needs to be done is to rewrite a 1 after reading it
(and even then I can probably just use a transistor on the read output coil to the write coil to achieve that)

getting back to topic though
so basically I'd suggested something previously that would work??
basically this with a microscopic film core instead of an iron core:


EDIT:
btw I'm not too worried about self-inductance (heat) cause I don't know of a drive that writes to the same bit more than 10 times a second
in fact afaik the file being written to is shuffled about on the drive, so usually you'd change the state of a single bit up to 3 times a day on average use. :P
(this doesn't factor in swap(file) or pagefile use)

the only time it becomes an issue is if it smokes after a single use.

EDIT2:
I'm gonna assume from the no response that my question is what you've been talking about all along. :P
initially I was actually thinking of the OP image as something like this:

where the inductance would carry from the primary to the secondary coil

but I see now that's not the case, and it would basically burst into magic smoke before the secondary coil would pick anything up
so answering my own question here, yeah, the previous image is what I'd want to go for in order for the secondary coil to pick up the state change of the film.

now all I need, if someone would please, is to test how many turns I would need for this to be effective at a microscopic level :)
cause I don't have any tools, and would have to build a specialized probe from salvaged components for sensitive measuring that I could plug into my PC soundcard
I'm sure you can understand why I would rather someone else test for me XD
« Last Edit: December 31, 2019, 04:01:45 pm by Tcll »
 

Offline TcllTopic starter

  • Newbie
  • Posts: 7
  • Country: us
Re: How well would Floppy Disk film work as a micro-core transformer?
« Reply #11 on: December 31, 2019, 04:23:39 pm »
btw, just to describe
That's why I suggest sticking with the head, because a century ago they solved all the problems that you're creating for yourself. :P

while I understand what you're getting at, the problem in and of itself is the moving parts:
1: the reason HDDs even at 10000+ RPM are super slow
2: the reason HDDs break down more often than SSDs
- WD has a problem with burning head amps
- Seagate has a problem with head sticking or the drive literally eating itself

technology has advanced enough to allow me to make microscopic heads
but I lack the knowledge to understand how wire densities affect mediums,
and how the properties of a medium affect current transfer through it...
but I know enough to build a 5v or 3v logic circuit out of basic transistors :P

I'm more of a programmer than an engineer, but that doesn't mean I'm incapable ;)
(unless you factor in the fact I have no tools and limited resource making me incapable T3T )
I've just recently had to build my own soldering iron and heat gun (nichrome wire based) just to get somewhere with my work... erg... x3x


btw, I checked out that resource, and I see it's just a reference (or rather more than) to help me understand you :)
thanks for that ;)
« Last Edit: January 02, 2020, 12:50:59 pm by Tcll »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf