Return to Table of Contents
Goal of this page - Discuss "Peaking Coils",
Inductors used to speed up logic level timing
Table of Contents
From Guy Fedorkow < email@example.com >
There was a question a couple weeks ago about the inductor in CTDL logic... I had an hour last week when I was should have been doing something else, but instead assembled a small Spice simulation of the gate with and without the inductor.
As guessed, the inductor sharpens the edge that's formed when the transistor turns off and the load resistor is in charge of switching levels.
I added 100 pF of load... I have no idea what the real loading is, but that seems like a possible value for a long wire and a couple of terminations.
If someone can stick a probe on a relatively clean logic signal and send me a photo, maybe we can calibrate the model a bit more carefully.
As usual, my first reaction was "an inductor makes it faster?? Can't be!" But of course that's what it does.
The peaking coils used in the SMS cards in the IBM 1401 system:
- are typically small in value ( 56uH ) and size ( 1/4 watt resistor )
- have been a high failure item in the German machine.
This is a bit of a surprise since the peaking coils seem basicly:
- maybe 50 turns of fine wire around a non-conducting slug
- two wire leads coming out the ends
- covered by epoxy and value identifying paint stripes
See photo below
On Dec 14, 2017 23:13, "Robert Garner" wrote: Robert, Thanks for dissolving away the outer shell or our failed inductor! Are the photos posted somewhere, or re-forward? Interesting that the failure point was where the external leads attach to the interior/inner wiring. Can you see if it was a solder connection? Perhaps heat expands the encapsulate, breaking the solder joint? Good idea, we should probably replace the core memory power inductors. (The entire 1401 may harbor over a 1000 inductors, so we won’t be replacing all of them. ;-)) Thanks again, - Robert
On Dec 14, 2017, at 10:58 PM, Robert Baruch
The sulfuric acid basically dissolved the outer shell, so the inner core was left alone. I can unwrap it if you'd like to see. The wiring does look somewhat messed up. Half of it actually looks nice, but the rest looks a bit wonky.
Inductor, used as Peaking Coil
- ExposedIt was also telling that during the procedure, the wire broke at both points where they join the leads. Definitely a weak point. You might consider replacing the other inductors which almost certainly will have the same problem at some point.
On Thu, Dec 14, 2017 at 10:39 PM Ken Shirriff
wrote: (Robert Baruch has been decapping chips and agreed to look at the bad inductor from the 1401. He runs project5474.org) Nice pictures. It's interesting to see what's inside the inductor. Thanks for dissolving it. The winding looks a bit haphazard - is that how it was, or did the wires shift around during processing? Is the core a different material or did you just take the inductor out before everything dissolved? Based on your previous comment that stressing the lead fixed the inductor, it seems like a bad connection between the lead and the internal wire would explain the problem. It seems like a random failure, not connected to the suspected power surge. Ken
Started Sept 6, 2015
Updated Dec 15, 2017