Value of L2 on SE-SE/30 Analog Board?

[CC_333]

Hi,

I have a couple of analog boards with an L2 width adjustment coil whose inner cores are missing and/or crumbling.

Is it possible to replace the coil outright, and if so, what is its value? If not, is there a way I can replace the core?

Thank you,

c

[nvdeynde]

L2 is a variable inductor. I have never seen this piece for sale on Mouser, Farnell, Digikey,...
Digikey sells variable inductors but as far as I can see, not one that can replace L2

If the plastic isn't broken, best chance is to look for a replacement coil.
However if you say that the coil has fallen out, the platic tube is probably damaged.

Best chance is look for a cheap/faulty Analog board from an 128K to SE30 and desolder the L2 variable inductor

[CC_333]

Best chance is look for a cheap/faulty Analog board from an 128K to SE30 and desolder the L2 variable inductor

That's what I thought.

Oh, well. I tried.

How come nobody makes this part, when everybody still makes almost everything else?

The plastic's intact, by the way. The core just crumbled inside (actually, the core is still mostly intact, but the center is stripped out so I can't turn it).

c

[trag]

These folks seem to have a dizzying array of old coils. I wouldn't be surprised if they have something compatible, but the problem is identifying it from the descriptions they use.

http://www.oldradioparts.com/
http://www.oldradioparts.com/2a6fla.txt

I don't know how significant it is, but from what I remember from a previous discussion on this topic, L2 is would from some kind of weird multi-stranded wire (miltz, mintz, mitchz wire, or some such), So, maybe we could find replacement ferrite cores and a plastic tube for it to turn inside, but finding that wire to wind around the thing could be a nightmare.

These folks have threaded cores, or rather, make them. Where you could actually buy them is another thing, but it's a place to start. Anyone know the dimensions (thread size and length) of the L2 core?
http://www.tscinternational.com/tscwebcat2007.html Scroll down about 3/4 of the way to the P1 threaded cores in various diameters/threads. It looks like the 1/4-28 series gets as long as 3/4".

[techknight]

Well see here is the thing. Anyone have an LCR meter handy? I sure dont. We need to take a known good coil and use a plastic adjustment tool and do an inductance test. both with the core in, and out.

This way, we know the value and be able to find a newer one that is electrically similar.

[CC_333]

We need to take a known good coil and use a plastic adjustment tool and do an inductance test.

What a great idea! Everyone can benefit, because these components do go bad from time to time, and the only other way to get a replacement is to steal one off another board.

Anyone have an LCR meter handy?

What's that?

c

[wally]

Some time ago prompted by some previous posts on jammed width control cores I started to experiment on a scrap analog board I just happened to have, and very quickly cracked and jammed my core quite badly into many fragments while using the correct size alignment tool. After some appropriate words I set it aside. Today I started examining the situation a bit more and here are some observations.

Removing a jammed core is easy if the core is mostly outside the part of the coil form covered by the litz wire coil winding. Carefully longitudinally slit the plastic coil form using a razor blade box cutter using precisely repeated shallow strokes until the blade cuts thru. Make the cut avoid the four inner ridges that provide the inner thread of the coil form, so there is a slight air space that the cutter encounters as it breaks thru, rather than core material. Now insert a jeweler's screwdriver blade into the slit near the coil and twist it a bit to widen the cut and relieve the grip the plastic coil form has on the core, or core fragments. It should be possible to back out the core, or at least the outer fragments of the core once the inner diameter of the coil form has been thus increased. Any fragments still remaining inside the coil proper might need some additional fragmentation to motivate them along.

After some partial reassembly my core fragments say they were once 1/4-28 x 1 inch long with teflon thread coating originally intended to provide some mechanical stability of position after initial factory adjustment. If you look under a microscope and play with this white material you will find it similar in consistency to teflon thread seal tape. Problem is that over the years, it balls up in clumps, incorporating any core wear particles until it jams the threads against the inside of the coil form. So once you get the core out, clean it completely free of teflon, and clean the inside of the coil form using a 1/4-28 threaded rod or a pulley tap.

My core fragments are black, more likely powdered iron rather than the gray of manganese-zinc ferrites associated with high frequency adjustable coils. So I went searching in my parts bins for adjustable coils and found a black core .76 inches long and another black one .38 inches long. There are many different grades/permeabilities of powdered iron but hey, this all I found. Then getting out the inductance bridge and the inductance meter as a cross check, I made the following measurements:
No core: 7 uH
.76 long core centered in winding: 33 uH
thread in additional .38 long core until it kisses against end of .76 core: 40 uH The inductance jumps up right at the contact point so cracks and gaps make for a significant decrease in inductance.

No claim on what's right for the Mac, and I'm not inclined to swap this in for a trial. I might try adhesive reassembly and measurement of the original core fragments, or open up a good unit but not today.

Some thoughts on long tight cores: if you use a plastic hex alignment tool to turn a tight core, the plastic twists until all the turning force is concentrated at the mouth of the core which tends to split it. A precisely ground metal hex tool would be better. But a poorly surfaced or plated metal hex tool might be worse because it could put all the turning force at a single high point anywhere inside the core and split it from there...

If a core does not need adjustment I would leave it. But if I find I need to do an adjustment on a teflon coated core, I would attempt to back it out completely and remove all teflon and wear particles from the threads inside and out while I still could. Repeated back and forth rotations are inviting thread jams while the core is deep inside the copper winding area.

[wally]

Measurements of an undisturbed, good and properly adjusted L2 in an SE-30 (with yoke plug disconnected, of course): 25 uH.
This unit has a black core, 1 inch long; .6 inches remain outside the copper winding area and .4 inches are inserted into the winding area that is about .550 inches long.

[trag]

Wow. Thank you for taking the actual measurement, Wally.

I did a little more searching after posting yesterday, and Litz wire, is apparently, not all that hard to find. Apparently, we might need to know things like how many strands in the sheath, and the strand gauge.

Anyway, it looks like, if we had something on which to wind the coil and into which to thread the core, we could probably make some of these if we really wanted to.

From your measurements, it appears that a 3/4" long slug might do the trick, unless other Macs need a higher inductance range.

Any idea at what frequency this component is operating in the circuit?

Okay, I emailed TSC to see if the 1/4-28 series threaded core is available. It doesn't appear to be present on their on-line store. I guess more to the point would be to check the first link I provided and see if any of those already made cores fit the inductance range you listed.

However, the total resistance probably matters too. While you have the equipment out, would you mind taking a DC resistance measurement of the coil as well. From what I remember the inductance is pretty much determined by the number of windings (does the diameter matter? I can't remember), and one chooses the gauge of wire to give the desired resistance for so many coils on such and such a diameter.

[wally]

.66 ohms, determined by four terminal measurement of .657 volts across coil component side solder joints while one amp was flowed thru PC solder side terminals.

Most people with this problem already have perfectly good copper windings on perfectly good coil forms, once that pesky but quite brittle stuck core is removed. One approach is to cut away the exposed form in various ways to allow an intact but stuck core to be grasped and twisted out to be cleaned and reused. Another approach better for already fragmented cores is further mechanical fragmentation by controlled violence. A nail with the end reshaped like a cold chisel should work, with the analog board removed well away from the CRT. If the fragments can be made small enough they could be shaken out and it may be unnecessary to do any plastic cutting at all. If you or others could find a suitable source for powdered iron cores of sufficient permeability, one or two could then be threaded in to the existing coil to repair it.

I don't know the sweep rate offhand but it's in the Guide to the Mac Family Hardware book. I think getting enough permeability will be the powdered iron core selection issue, not the core losses at the several harmonics of the sweep ramp (but it can't be solid or laminated transformer iron, and I am uncertain if ferrites of the manganese-zinc will saturate in this application).

[wally]

Major correction! This morning I read my last post again and decided to repeat my measurement of resistance which seemed to me to be too high given the wire size. Turns out in going to post yesterday I did a decimal point error. Sorry, here is the corrected post and value:

.066 ohms, determined by four terminal measurement of .0657 volts across coil component side solder joints while one amp was flowed thru PC solder side terminals.

[trag]

Thank you, Wally. This going to be handy in the long run, if we can remember to look in this thread. Perhaps we should add a page to the Wiki for the L2 coil specifications.

What do you think sufficient permeability would be?

From TSCI's document,which I linked to above, they sell threaded cores with permeabilities of 6, 7, 8, 8.5, 10, and 20. They list "recommended frequency ranges" of 40 - 250 MHz, 35 - 135 MHz, 10 - 100 MHz, 1 - 60 MHz, 100KHz - 25 MHz, and 50 KHz - 5 MHz respectively.

They have part numbers for the 1/4-28 X .75" threaded core at all six of those permeabilities. Do you think that someone replacing a disintegrated L2 core would need both a .75" core and a .25" core, or would the .75" core be sufficient? Or needs experimenting to tell for certain?

I've had an email reply from the Customer Service Rep. at TSCI, so if we can pick out a permeability, I can give her a specific part number that I'm interested in.

They have plain rods in .245"D X 1"L and .245"D X 1.25"L, but the threaded cores only go to .75" alas. They also mention that teflon coating is available on the threaded rods, but from what you wrote, I think we want to avoid that, yes?

[wally]

Needs some experimentation. If two cores are mated there is still a small gap plus chamfer taper of effective cross section at the meeting ends so we can't presume exact equivalence to a single seamless length of powdered iron.

The width control has to be able to compensate for variations in DC HV to the CRT (changes deflection factor), variations in flyback sweep voltage, and variations in the total of the 10 uH linearity coil plus the nominal 80 uH yoke (mine measured 80 uH). I suspect .75 inches long alone of the permeability 20 substance would work for me but is likely insufficient for accommodating all variations in the SE/SE/30 installed base.

Maybe a suitable repair kit would be two .75 inch long cores of the 20 permeability, one to replace the most instances of broken core, and the other, just in case a further boost is needed in inductance, or in case of mishap with the first core.

I'd avoid the teflon coating: we've had enough experience with clumping already! Some sweep linearity experimentation may be necessary between the permeability 20 and 10 choices, my hunch is that u=20 and .75 length are a starting point, with two cores on hand in case a second is necessary. Even if the second is way longer than needed, most of it will be so far from the coil I think core losses will not be an issue. Would they send you two free engineering samples of u=20 and two of u=10, all in length .75 inches?

[trag]

I forgot to update this at the time, but ultimately, the supplier I found, in a very round about way, indicated that they weren't interested in helping me at the volumes I was interested in. I don't understand the round about nature of their response though. I would be happy if they would tell me right at the outset, "Look, if you don't want to order 10,000 of them then we're not interested." It would save time. It would also be nice if they would state right out what quantities they consider too small.

Ah well. I'm still not 100% certain that was the final result. They were so wishy washy, that I'm tempted to email them and try again.

However, I wonder if we might just try getting a few of these:

http://www.allelectronics.com/make-a-store/item/RC-676/0.676-MICRO-HENRY-ADJUSTABLE-CHOKE/1.html

and rewind them with Litz wire.

Hmmm. Anyone have a definitely dead L2 coil? It might be useful to unwind it and carefully count the windings. I wonder if one can determine the number of strands in the Litz wire with a careful visual examination.